1.. SPDX-License-Identifier: BSD-3-Clause 2 Copyright(c) 2010-2016 Intel Corporation. 3 4.. _testpmd_runtime: 5 6Testpmd Runtime Functions 7========================= 8 9Where the testpmd application is started in interactive mode, (``-i|--interactive``), 10it displays a prompt that can be used to start and stop forwarding, 11configure the application, display statistics (including the extended NIC 12statistics aka xstats) , set the Flow Director and other tasks:: 13 14 testpmd> 15 16The testpmd prompt has some, limited, readline support. 17Common bash command-line functions such as ``Ctrl+a`` and ``Ctrl+e`` to go to the start and end of the prompt line are supported 18as well as access to the command history via the up-arrow. 19 20There is also support for tab completion. 21If you type a partial command and hit ``<TAB>`` you get a list of the available completions: 22 23.. code-block:: console 24 25 testpmd> show port <TAB> 26 27 info [Mul-choice STRING]: show|clear port info|stats|xstats|fdir|stat_qmap|dcb_tc|cap X 28 info [Mul-choice STRING]: show|clear port info|stats|xstats|fdir|stat_qmap|dcb_tc|cap all 29 stats [Mul-choice STRING]: show|clear port info|stats|xstats|fdir|stat_qmap|dcb_tc|cap X 30 stats [Mul-choice STRING]: show|clear port info|stats|xstats|fdir|stat_qmap|dcb_tc|cap all 31 ... 32 33 34.. note:: 35 36 Some examples in this document are too long to fit on one line are are shown wrapped at `"\\"` for display purposes:: 37 38 testpmd> set flow_ctrl rx (on|off) tx (on|off) (high_water) (low_water) \ 39 (pause_time) (send_xon) (port_id) 40 41In the real ``testpmd>`` prompt these commands should be on a single line. 42 43Help Functions 44-------------- 45 46The testpmd has on-line help for the functions that are available at runtime. 47These are divided into sections and can be accessed using help, help section or help all: 48 49.. code-block:: console 50 51 testpmd> help 52 53 help control : Start and stop forwarding. 54 help display : Displaying port, stats and config information. 55 help config : Configuration information. 56 help ports : Configuring ports. 57 help registers : Reading and setting port registers. 58 help filters : Filters configuration help. 59 help all : All of the above sections. 60 61 62Command File Functions 63---------------------- 64 65To facilitate loading large number of commands or to avoid cutting and pasting where not 66practical or possible testpmd supports alternative methods for executing commands. 67 68* If started with the ``--cmdline-file=FILENAME`` command line argument testpmd 69 will execute all CLI commands contained within the file immediately before 70 starting packet forwarding or entering interactive mode. 71 72.. code-block:: console 73 74 ./testpmd -n4 -r2 ... -- -i --cmdline-file=/home/ubuntu/flow-create-commands.txt 75 Interactive-mode selected 76 CLI commands to be read from /home/ubuntu/flow-create-commands.txt 77 Configuring Port 0 (socket 0) 78 Port 0: 7C:FE:90:CB:74:CE 79 Configuring Port 1 (socket 0) 80 Port 1: 7C:FE:90:CB:74:CA 81 Checking link statuses... 82 Port 0 Link Up - speed 10000 Mbps - full-duplex 83 Port 1 Link Up - speed 10000 Mbps - full-duplex 84 Done 85 Flow rule #0 created 86 Flow rule #1 created 87 ... 88 ... 89 Flow rule #498 created 90 Flow rule #499 created 91 Read all CLI commands from /home/ubuntu/flow-create-commands.txt 92 testpmd> 93 94 95* At run-time additional commands can be loaded in bulk by invoking the ``load FILENAME`` 96 command. 97 98.. code-block:: console 99 100 testpmd> load /home/ubuntu/flow-create-commands.txt 101 Flow rule #0 created 102 Flow rule #1 created 103 ... 104 ... 105 Flow rule #498 created 106 Flow rule #499 created 107 Read all CLI commands from /home/ubuntu/flow-create-commands.txt 108 testpmd> 109 110 111In all cases output from any included command will be displayed as standard output. 112Execution will continue until the end of the file is reached regardless of 113whether any errors occur. The end user must examine the output to determine if 114any failures occurred. 115 116 117Control Functions 118----------------- 119 120start 121~~~~~ 122 123Start packet forwarding with current configuration:: 124 125 testpmd> start 126 127start tx_first 128~~~~~~~~~~~~~~ 129 130Start packet forwarding with current configuration after sending specified number of bursts of packets:: 131 132 testpmd> start tx_first (""|burst_num) 133 134The default burst number is 1 when ``burst_num`` not presented. 135 136stop 137~~~~ 138 139Stop packet forwarding, and display accumulated statistics:: 140 141 testpmd> stop 142 143quit 144~~~~ 145 146Quit to prompt:: 147 148 testpmd> quit 149 150 151Display Functions 152----------------- 153 154The functions in the following sections are used to display information about the 155testpmd configuration or the NIC status. 156 157show port 158~~~~~~~~~ 159 160Display information for a given port or all ports:: 161 162 testpmd> show port (info|stats|xstats|fdir|stat_qmap|dcb_tc|cap) (port_id|all) 163 164The available information categories are: 165 166* ``info``: General port information such as MAC address. 167 168* ``stats``: RX/TX statistics. 169 170* ``xstats``: RX/TX extended NIC statistics. 171 172* ``fdir``: Flow Director information and statistics. 173 174* ``stat_qmap``: Queue statistics mapping. 175 176* ``dcb_tc``: DCB information such as TC mapping. 177 178* ``cap``: Supported offload capabilities. 179 180For example: 181 182.. code-block:: console 183 184 testpmd> show port info 0 185 186 ********************* Infos for port 0 ********************* 187 188 MAC address: XX:XX:XX:XX:XX:XX 189 Connect to socket: 0 190 memory allocation on the socket: 0 191 Link status: up 192 Link speed: 40000 Mbps 193 Link duplex: full-duplex 194 Promiscuous mode: enabled 195 Allmulticast mode: disabled 196 Maximum number of MAC addresses: 64 197 Maximum number of MAC addresses of hash filtering: 0 198 VLAN offload: 199 strip on 200 filter on 201 qinq(extend) off 202 Redirection table size: 512 203 Supported flow types: 204 ipv4-frag 205 ipv4-tcp 206 ipv4-udp 207 ipv4-sctp 208 ipv4-other 209 ipv6-frag 210 ipv6-tcp 211 ipv6-udp 212 ipv6-sctp 213 ipv6-other 214 l2_payload 215 port 216 vxlan 217 geneve 218 nvgre 219 220show port rss reta 221~~~~~~~~~~~~~~~~~~ 222 223Display the rss redirection table entry indicated by masks on port X:: 224 225 testpmd> show port (port_id) rss reta (size) (mask0, mask1...) 226 227size is used to indicate the hardware supported reta size 228 229show port rss-hash 230~~~~~~~~~~~~~~~~~~ 231 232Display the RSS hash functions and RSS hash key of a port:: 233 234 testpmd> show port (port_id) rss-hash ipv4|ipv4-frag|ipv4-tcp|ipv4-udp|ipv4-sctp|ipv4-other|ipv6|ipv6-frag|ipv6-tcp|ipv6-udp|ipv6-sctp|ipv6-other|l2-payload|ipv6-ex|ipv6-tcp-ex|ipv6-udp-ex [key] 235 236clear port 237~~~~~~~~~~ 238 239Clear the port statistics for a given port or for all ports:: 240 241 testpmd> clear port (info|stats|xstats|fdir|stat_qmap) (port_id|all) 242 243For example:: 244 245 testpmd> clear port stats all 246 247show (rxq|txq) 248~~~~~~~~~~~~~~ 249 250Display information for a given port's RX/TX queue:: 251 252 testpmd> show (rxq|txq) info (port_id) (queue_id) 253 254show config 255~~~~~~~~~~~ 256 257Displays the configuration of the application. 258The configuration comes from the command-line, the runtime or the application defaults:: 259 260 testpmd> show config (rxtx|cores|fwd|txpkts) 261 262The available information categories are: 263 264* ``rxtx``: RX/TX configuration items. 265 266* ``cores``: List of forwarding cores. 267 268* ``fwd``: Packet forwarding configuration. 269 270* ``txpkts``: Packets to TX configuration. 271 272For example: 273 274.. code-block:: console 275 276 testpmd> show config rxtx 277 278 io packet forwarding - CRC stripping disabled - packets/burst=16 279 nb forwarding cores=2 - nb forwarding ports=1 280 RX queues=1 - RX desc=128 - RX free threshold=0 281 RX threshold registers: pthresh=8 hthresh=8 wthresh=4 282 TX queues=1 - TX desc=512 - TX free threshold=0 283 TX threshold registers: pthresh=36 hthresh=0 wthresh=0 284 TX RS bit threshold=0 - TXQ flags=0x0 285 286set fwd 287~~~~~~~ 288 289Set the packet forwarding mode:: 290 291 testpmd> set fwd (io|mac|macswap|flowgen| \ 292 rxonly|txonly|csum|icmpecho) (""|retry) 293 294``retry`` can be specified for forwarding engines except ``rx_only``. 295 296The available information categories are: 297 298* ``io``: Forwards packets "as-is" in I/O mode. 299 This is the fastest possible forwarding operation as it does not access packets data. 300 This is the default mode. 301 302* ``mac``: Changes the source and the destination Ethernet addresses of packets before forwarding them. 303 Default application behaviour is to set source Ethernet address to that of the transmitting interface, and destination 304 address to a dummy value (set during init). The user may specify a target destination Ethernet address via the 'eth-peer' or 305 'eth-peer-configfile' command-line options. It is not currently possible to specify a specific source Ethernet address. 306 307* ``macswap``: MAC swap forwarding mode. 308 Swaps the source and the destination Ethernet addresses of packets before forwarding them. 309 310* ``flowgen``: Multi-flow generation mode. 311 Originates a number of flows (with varying destination IP addresses), and terminate receive traffic. 312 313* ``rxonly``: Receives packets but doesn't transmit them. 314 315* ``txonly``: Generates and transmits packets without receiving any. 316 317* ``csum``: Changes the checksum field with hardware or software methods depending on the offload flags on the packet. 318 319* ``icmpecho``: Receives a burst of packets, lookup for IMCP echo requests and, if any, send back ICMP echo replies. 320 321* ``ieee1588``: Demonstrate L2 IEEE1588 V2 PTP timestamping for RX and TX. Requires ``CONFIG_RTE_LIBRTE_IEEE1588=y``. 322 323* ``tm``: Traffic Management forwarding mode 324 Demonstrates the use of ethdev traffic management APIs and softnic PMD for 325 QoS traffic management. In this mode, 5-level hierarchical QoS scheduler is 326 available as an default option that can be enabled through CLI. The user can 327 also modify the default hierarchy or specify the new hierarchy through CLI for 328 implementing QoS scheduler. Requires ``CONFIG_RTE_LIBRTE_PMD_SOFTNIC=y`` ``CONFIG_RTE_LIBRTE_SCHED=y``. 329 330Example:: 331 332 testpmd> set fwd rxonly 333 334 Set rxonly packet forwarding mode 335 336 337read rxd 338~~~~~~~~ 339 340Display an RX descriptor for a port RX queue:: 341 342 testpmd> read rxd (port_id) (queue_id) (rxd_id) 343 344For example:: 345 346 testpmd> read rxd 0 0 4 347 0x0000000B - 0x001D0180 / 0x0000000B - 0x001D0180 348 349read txd 350~~~~~~~~ 351 352Display a TX descriptor for a port TX queue:: 353 354 testpmd> read txd (port_id) (queue_id) (txd_id) 355 356For example:: 357 358 testpmd> read txd 0 0 4 359 0x00000001 - 0x24C3C440 / 0x000F0000 - 0x2330003C 360 361ddp get list 362~~~~~~~~~~~~ 363 364Get loaded dynamic device personalization (DDP) package info list:: 365 366 testpmd> ddp get list (port_id) 367 368ddp get info 369~~~~~~~~~~~~ 370 371Display information about dynamic device personalization (DDP) profile:: 372 373 testpmd> ddp get info (profile_path) 374 375show vf stats 376~~~~~~~~~~~~~ 377 378Display VF statistics:: 379 380 testpmd> show vf stats (port_id) (vf_id) 381 382clear vf stats 383~~~~~~~~~~~~~~ 384 385Reset VF statistics:: 386 387 testpmd> clear vf stats (port_id) (vf_id) 388 389show port pctype mapping 390~~~~~~~~~~~~~~~~~~~~~~~~ 391 392List all items from the pctype mapping table:: 393 394 testpmd> show port (port_id) pctype mapping 395 396show rx offloading capabilities 397~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ 398 399List all per queue and per port Rx offloading capabilities of a port:: 400 401 testpmd> show port (port_id) rx_offload capabilities 402 403show rx offloading configuration 404~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ 405 406List port level and all queue level Rx offloading configuration:: 407 408 testpmd> show port (port_id) rx_offload configuration 409 410show tx offloading capabilities 411~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ 412 413List all per queue and per port Tx offloading capabilities of a port:: 414 415 testpmd> show port (port_id) tx_offload capabilities 416 417show tx offloading configuration 418~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ 419 420List port level and all queue level Tx offloading configuration:: 421 422 testpmd> show port (port_id) tx_offload configuration 423 424 425Configuration Functions 426----------------------- 427 428The testpmd application can be configured from the runtime as well as from the command-line. 429 430This section details the available configuration functions that are available. 431 432.. note:: 433 434 Configuration changes only become active when forwarding is started/restarted. 435 436set default 437~~~~~~~~~~~ 438 439Reset forwarding to the default configuration:: 440 441 testpmd> set default 442 443set verbose 444~~~~~~~~~~~ 445 446Set the debug verbosity level:: 447 448 testpmd> set verbose (level) 449 450Currently the only available levels are 0 (silent except for error) and 1 (fully verbose). 451 452set log 453~~~~~~~ 454 455Set the log level for a log type:: 456 457 testpmd> set log global|(type) (level) 458 459Where: 460 461* ``type`` is the log name. 462 463* ``level`` is the log level. 464 465For example, to change the global log level:: 466 testpmd> set log global (level) 467 468Regexes can also be used for type. To change log level of user1, user2 and user3:: 469 testpmd> set log user[1-3] (level) 470 471set nbport 472~~~~~~~~~~ 473 474Set the number of ports used by the application: 475 476set nbport (num) 477 478This is equivalent to the ``--nb-ports`` command-line option. 479 480set nbcore 481~~~~~~~~~~ 482 483Set the number of cores used by the application:: 484 485 testpmd> set nbcore (num) 486 487This is equivalent to the ``--nb-cores`` command-line option. 488 489.. note:: 490 491 The number of cores used must not be greater than number of ports used multiplied by the number of queues per port. 492 493set coremask 494~~~~~~~~~~~~ 495 496Set the forwarding cores hexadecimal mask:: 497 498 testpmd> set coremask (mask) 499 500This is equivalent to the ``--coremask`` command-line option. 501 502.. note:: 503 504 The master lcore is reserved for command line parsing only and cannot be masked on for packet forwarding. 505 506set portmask 507~~~~~~~~~~~~ 508 509Set the forwarding ports hexadecimal mask:: 510 511 testpmd> set portmask (mask) 512 513This is equivalent to the ``--portmask`` command-line option. 514 515set burst 516~~~~~~~~~ 517 518Set number of packets per burst:: 519 520 testpmd> set burst (num) 521 522This is equivalent to the ``--burst command-line`` option. 523 524When retry is enabled, the transmit delay time and number of retries can also be set:: 525 526 testpmd> set burst tx delay (microseconds) retry (num) 527 528set txpkts 529~~~~~~~~~~ 530 531Set the length of each segment of the TX-ONLY packets or length of packet for FLOWGEN mode:: 532 533 testpmd> set txpkts (x[,y]*) 534 535Where x[,y]* represents a CSV list of values, without white space. 536 537set txsplit 538~~~~~~~~~~~ 539 540Set the split policy for the TX packets, applicable for TX-ONLY and CSUM forwarding modes:: 541 542 testpmd> set txsplit (off|on|rand) 543 544Where: 545 546* ``off`` disable packet copy & split for CSUM mode. 547 548* ``on`` split outgoing packet into multiple segments. Size of each segment 549 and number of segments per packet is determined by ``set txpkts`` command 550 (see above). 551 552* ``rand`` same as 'on', but number of segments per each packet is a random value between 1 and total number of segments. 553 554set corelist 555~~~~~~~~~~~~ 556 557Set the list of forwarding cores:: 558 559 testpmd> set corelist (x[,y]*) 560 561For example, to change the forwarding cores: 562 563.. code-block:: console 564 565 testpmd> set corelist 3,1 566 testpmd> show config fwd 567 568 io packet forwarding - ports=2 - cores=2 - streams=2 - NUMA support disabled 569 Logical Core 3 (socket 0) forwards packets on 1 streams: 570 RX P=0/Q=0 (socket 0) -> TX P=1/Q=0 (socket 0) peer=02:00:00:00:00:01 571 Logical Core 1 (socket 0) forwards packets on 1 streams: 572 RX P=1/Q=0 (socket 0) -> TX P=0/Q=0 (socket 0) peer=02:00:00:00:00:00 573 574.. note:: 575 576 The cores are used in the same order as specified on the command line. 577 578set portlist 579~~~~~~~~~~~~ 580 581Set the list of forwarding ports:: 582 583 testpmd> set portlist (x[,y]*) 584 585For example, to change the port forwarding: 586 587.. code-block:: console 588 589 testpmd> set portlist 0,2,1,3 590 testpmd> show config fwd 591 592 io packet forwarding - ports=4 - cores=1 - streams=4 593 Logical Core 3 (socket 0) forwards packets on 4 streams: 594 RX P=0/Q=0 (socket 0) -> TX P=2/Q=0 (socket 0) peer=02:00:00:00:00:01 595 RX P=2/Q=0 (socket 0) -> TX P=0/Q=0 (socket 0) peer=02:00:00:00:00:00 596 RX P=1/Q=0 (socket 0) -> TX P=3/Q=0 (socket 0) peer=02:00:00:00:00:03 597 RX P=3/Q=0 (socket 0) -> TX P=1/Q=0 (socket 0) peer=02:00:00:00:00:02 598 599set tx loopback 600~~~~~~~~~~~~~~~ 601 602Enable/disable tx loopback:: 603 604 testpmd> set tx loopback (port_id) (on|off) 605 606set drop enable 607~~~~~~~~~~~~~~~ 608 609set drop enable bit for all queues:: 610 611 testpmd> set all queues drop (port_id) (on|off) 612 613set split drop enable (for VF) 614~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ 615 616set split drop enable bit for VF from PF:: 617 618 testpmd> set vf split drop (port_id) (vf_id) (on|off) 619 620set mac antispoof (for VF) 621~~~~~~~~~~~~~~~~~~~~~~~~~~ 622 623Set mac antispoof for a VF from the PF:: 624 625 testpmd> set vf mac antispoof (port_id) (vf_id) (on|off) 626 627set macsec offload 628~~~~~~~~~~~~~~~~~~ 629 630Enable/disable MACsec offload:: 631 632 testpmd> set macsec offload (port_id) on encrypt (on|off) replay-protect (on|off) 633 testpmd> set macsec offload (port_id) off 634 635set macsec sc 636~~~~~~~~~~~~~ 637 638Configure MACsec secure connection (SC):: 639 640 testpmd> set macsec sc (tx|rx) (port_id) (mac) (pi) 641 642.. note:: 643 644 The pi argument is ignored for tx. 645 Check the NIC Datasheet for hardware limits. 646 647set macsec sa 648~~~~~~~~~~~~~ 649 650Configure MACsec secure association (SA):: 651 652 testpmd> set macsec sa (tx|rx) (port_id) (idx) (an) (pn) (key) 653 654.. note:: 655 656 The IDX value must be 0 or 1. 657 Check the NIC Datasheet for hardware limits. 658 659set broadcast mode (for VF) 660~~~~~~~~~~~~~~~~~~~~~~~~~~~ 661 662Set broadcast mode for a VF from the PF:: 663 664 testpmd> set vf broadcast (port_id) (vf_id) (on|off) 665 666vlan set strip 667~~~~~~~~~~~~~~ 668 669Set the VLAN strip on a port:: 670 671 testpmd> vlan set strip (on|off) (port_id) 672 673vlan set stripq 674~~~~~~~~~~~~~~~ 675 676Set the VLAN strip for a queue on a port:: 677 678 testpmd> vlan set stripq (on|off) (port_id,queue_id) 679 680vlan set stripq (for VF) 681~~~~~~~~~~~~~~~~~~~~~~~~ 682 683Set VLAN strip for all queues in a pool for a VF from the PF:: 684 685 testpmd> set vf vlan stripq (port_id) (vf_id) (on|off) 686 687vlan set insert (for VF) 688~~~~~~~~~~~~~~~~~~~~~~~~ 689 690Set VLAN insert for a VF from the PF:: 691 692 testpmd> set vf vlan insert (port_id) (vf_id) (vlan_id) 693 694vlan set tag (for VF) 695~~~~~~~~~~~~~~~~~~~~~ 696 697Set VLAN tag for a VF from the PF:: 698 699 testpmd> set vf vlan tag (port_id) (vf_id) (on|off) 700 701vlan set antispoof (for VF) 702~~~~~~~~~~~~~~~~~~~~~~~~~~~ 703 704Set VLAN antispoof for a VF from the PF:: 705 706 testpmd> set vf vlan antispoof (port_id) (vf_id) (on|off) 707 708vlan set filter 709~~~~~~~~~~~~~~~ 710 711Set the VLAN filter on a port:: 712 713 testpmd> vlan set filter (on|off) (port_id) 714 715vlan set qinq 716~~~~~~~~~~~~~ 717 718Set the VLAN QinQ (extended queue in queue) on for a port:: 719 720 testpmd> vlan set qinq (on|off) (port_id) 721 722vlan set tpid 723~~~~~~~~~~~~~ 724 725Set the inner or outer VLAN TPID for packet filtering on a port:: 726 727 testpmd> vlan set (inner|outer) tpid (value) (port_id) 728 729.. note:: 730 731 TPID value must be a 16-bit number (value <= 65536). 732 733rx_vlan add 734~~~~~~~~~~~ 735 736Add a VLAN ID, or all identifiers, to the set of VLAN identifiers filtered by port ID:: 737 738 testpmd> rx_vlan add (vlan_id|all) (port_id) 739 740.. note:: 741 742 VLAN filter must be set on that port. VLAN ID < 4096. 743 Depending on the NIC used, number of vlan_ids may be limited to the maximum entries 744 in VFTA table. This is important if enabling all vlan_ids. 745 746rx_vlan rm 747~~~~~~~~~~ 748 749Remove a VLAN ID, or all identifiers, from the set of VLAN identifiers filtered by port ID:: 750 751 testpmd> rx_vlan rm (vlan_id|all) (port_id) 752 753rx_vlan add (for VF) 754~~~~~~~~~~~~~~~~~~~~ 755 756Add a VLAN ID, to the set of VLAN identifiers filtered for VF(s) for port ID:: 757 758 testpmd> rx_vlan add (vlan_id) port (port_id) vf (vf_mask) 759 760rx_vlan rm (for VF) 761~~~~~~~~~~~~~~~~~~~ 762 763Remove a VLAN ID, from the set of VLAN identifiers filtered for VF(s) for port ID:: 764 765 testpmd> rx_vlan rm (vlan_id) port (port_id) vf (vf_mask) 766 767tunnel_filter add 768~~~~~~~~~~~~~~~~~ 769 770Add a tunnel filter on a port:: 771 772 testpmd> tunnel_filter add (port_id) (outer_mac) (inner_mac) (ip_addr) \ 773 (inner_vlan) (vxlan|nvgre|ipingre) (imac-ivlan|imac-ivlan-tenid|\ 774 imac-tenid|imac|omac-imac-tenid|oip|iip) (tenant_id) (queue_id) 775 776The available information categories are: 777 778* ``vxlan``: Set tunnel type as VXLAN. 779 780* ``nvgre``: Set tunnel type as NVGRE. 781 782* ``ipingre``: Set tunnel type as IP-in-GRE. 783 784* ``imac-ivlan``: Set filter type as Inner MAC and VLAN. 785 786* ``imac-ivlan-tenid``: Set filter type as Inner MAC, VLAN and tenant ID. 787 788* ``imac-tenid``: Set filter type as Inner MAC and tenant ID. 789 790* ``imac``: Set filter type as Inner MAC. 791 792* ``omac-imac-tenid``: Set filter type as Outer MAC, Inner MAC and tenant ID. 793 794* ``oip``: Set filter type as Outer IP. 795 796* ``iip``: Set filter type as Inner IP. 797 798Example:: 799 800 testpmd> tunnel_filter add 0 68:05:CA:28:09:82 00:00:00:00:00:00 \ 801 192.168.2.2 0 ipingre oip 1 1 802 803 Set an IP-in-GRE tunnel on port 0, and the filter type is Outer IP. 804 805tunnel_filter remove 806~~~~~~~~~~~~~~~~~~~~ 807 808Remove a tunnel filter on a port:: 809 810 testpmd> tunnel_filter rm (port_id) (outer_mac) (inner_mac) (ip_addr) \ 811 (inner_vlan) (vxlan|nvgre|ipingre) (imac-ivlan|imac-ivlan-tenid|\ 812 imac-tenid|imac|omac-imac-tenid|oip|iip) (tenant_id) (queue_id) 813 814rx_vxlan_port add 815~~~~~~~~~~~~~~~~~ 816 817Add an UDP port for VXLAN packet filter on a port:: 818 819 testpmd> rx_vxlan_port add (udp_port) (port_id) 820 821rx_vxlan_port remove 822~~~~~~~~~~~~~~~~~~~~ 823 824Remove an UDP port for VXLAN packet filter on a port:: 825 826 testpmd> rx_vxlan_port rm (udp_port) (port_id) 827 828tx_vlan set 829~~~~~~~~~~~ 830 831Set hardware insertion of VLAN IDs in packets sent on a port:: 832 833 testpmd> tx_vlan set (port_id) vlan_id[, vlan_id_outer] 834 835For example, set a single VLAN ID (5) insertion on port 0:: 836 837 tx_vlan set 0 5 838 839Or, set double VLAN ID (inner: 2, outer: 3) insertion on port 1:: 840 841 tx_vlan set 1 2 3 842 843 844tx_vlan set pvid 845~~~~~~~~~~~~~~~~ 846 847Set port based hardware insertion of VLAN ID in packets sent on a port:: 848 849 testpmd> tx_vlan set pvid (port_id) (vlan_id) (on|off) 850 851tx_vlan reset 852~~~~~~~~~~~~~ 853 854Disable hardware insertion of a VLAN header in packets sent on a port:: 855 856 testpmd> tx_vlan reset (port_id) 857 858csum set 859~~~~~~~~ 860 861Select hardware or software calculation of the checksum when 862transmitting a packet using the ``csum`` forwarding engine:: 863 864 testpmd> csum set (ip|udp|tcp|sctp|outer-ip) (hw|sw) (port_id) 865 866Where: 867 868* ``ip|udp|tcp|sctp`` always relate to the inner layer. 869 870* ``outer-ip`` relates to the outer IP layer (only for IPv4) in the case where the packet is recognized 871 as a tunnel packet by the forwarding engine (vxlan, gre and ipip are 872 supported). See also the ``csum parse-tunnel`` command. 873 874.. note:: 875 876 Check the NIC Datasheet for hardware limits. 877 878RSS queue region 879~~~~~~~~~~~~~~~~ 880 881Set RSS queue region span on a port:: 882 883 testpmd> set port (port_id) queue-region region_id (value) \ 884 queue_start_index (value) queue_num (value) 885 886Set flowtype mapping on a RSS queue region on a port:: 887 888 testpmd> set port (port_id) queue-region region_id (value) flowtype (value) 889 890where: 891 892* For the flowtype(pctype) of packet,the specific index for each type has 893 been defined in file i40e_type.h as enum i40e_filter_pctype. 894 895Set user priority mapping on a RSS queue region on a port:: 896 897 testpmd> set port (port_id) queue-region UP (value) region_id (value) 898 899Flush all queue region related configuration on a port:: 900 901 testpmd> set port (port_id) queue-region flush (on|off) 902 903where: 904 905* "on"is just an enable function which server for other configuration, 906 it is for all configuration about queue region from up layer, 907 at first will only keep in DPDK softwarestored in driver, 908 only after "flush on", it commit all configuration to HW. 909 "off" is just clean all configuration about queue region just now, 910 and restore all to DPDK i40e driver default config when start up. 911 912Show all queue region related configuration info on a port:: 913 914 testpmd> show port (port_id) queue-region 915 916.. note:: 917 918 Queue region only support on PF by now, so these command is 919 only for configuration of queue region on PF port. 920 921csum parse-tunnel 922~~~~~~~~~~~~~~~~~ 923 924Define how tunneled packets should be handled by the csum forward 925engine:: 926 927 testpmd> csum parse-tunnel (on|off) (tx_port_id) 928 929If enabled, the csum forward engine will try to recognize supported 930tunnel headers (vxlan, gre, ipip). 931 932If disabled, treat tunnel packets as non-tunneled packets (a inner 933header is handled as a packet payload). 934 935.. note:: 936 937 The port argument is the TX port like in the ``csum set`` command. 938 939Example: 940 941Consider a packet in packet like the following:: 942 943 eth_out/ipv4_out/udp_out/vxlan/eth_in/ipv4_in/tcp_in 944 945* If parse-tunnel is enabled, the ``ip|udp|tcp|sctp`` parameters of ``csum set`` 946 command relate to the inner headers (here ``ipv4_in`` and ``tcp_in``), and the 947 ``outer-ip parameter`` relates to the outer headers (here ``ipv4_out``). 948 949* If parse-tunnel is disabled, the ``ip|udp|tcp|sctp`` parameters of ``csum set`` 950 command relate to the outer headers, here ``ipv4_out`` and ``udp_out``. 951 952csum show 953~~~~~~~~~ 954 955Display tx checksum offload configuration:: 956 957 testpmd> csum show (port_id) 958 959tso set 960~~~~~~~ 961 962Enable TCP Segmentation Offload (TSO) in the ``csum`` forwarding engine:: 963 964 testpmd> tso set (segsize) (port_id) 965 966.. note:: 967 968 Check the NIC datasheet for hardware limits. 969 970tso show 971~~~~~~~~ 972 973Display the status of TCP Segmentation Offload:: 974 975 testpmd> tso show (port_id) 976 977set port - gro 978~~~~~~~~~~~~~~ 979 980Enable or disable GRO in ``csum`` forwarding engine:: 981 982 testpmd> set port <port_id> gro on|off 983 984If enabled, the csum forwarding engine will perform GRO on the TCP/IPv4 985packets received from the given port. 986 987If disabled, packets received from the given port won't be performed 988GRO. By default, GRO is disabled for all ports. 989 990.. note:: 991 992 When enable GRO for a port, TCP/IPv4 packets received from the port 993 will be performed GRO. After GRO, all merged packets have bad 994 checksums, since the GRO library doesn't re-calculate checksums for 995 the merged packets. Therefore, if users want the merged packets to 996 have correct checksums, please select HW IP checksum calculation and 997 HW TCP checksum calculation for the port which the merged packets are 998 transmitted to. 999 1000show port - gro 1001~~~~~~~~~~~~~~~ 1002 1003Display GRO configuration for a given port:: 1004 1005 testpmd> show port <port_id> gro 1006 1007set gro flush 1008~~~~~~~~~~~~~ 1009 1010Set the cycle to flush the GROed packets from reassembly tables:: 1011 1012 testpmd> set gro flush <cycles> 1013 1014When enable GRO, the csum forwarding engine performs GRO on received 1015packets, and the GROed packets are stored in reassembly tables. Users 1016can use this command to determine when the GROed packets are flushed 1017from the reassembly tables. 1018 1019The ``cycles`` is measured in GRO operation times. The csum forwarding 1020engine flushes the GROed packets from the tables every ``cycles`` GRO 1021operations. 1022 1023By default, the value of ``cycles`` is 1, which means flush GROed packets 1024from the reassembly tables as soon as one GRO operation finishes. The value 1025of ``cycles`` should be in the range of 1 to ``GRO_MAX_FLUSH_CYCLES``. 1026 1027Please note that the large value of ``cycles`` may cause the poor TCP/IP 1028stack performance. Because the GROed packets are delayed to arrive the 1029stack, thus causing more duplicated ACKs and TCP retransmissions. 1030 1031set port - gso 1032~~~~~~~~~~~~~~ 1033 1034Toggle per-port GSO support in ``csum`` forwarding engine:: 1035 1036 testpmd> set port <port_id> gso on|off 1037 1038If enabled, the csum forwarding engine will perform GSO on supported IPv4 1039packets, transmitted on the given port. 1040 1041If disabled, packets transmitted on the given port will not undergo GSO. 1042By default, GSO is disabled for all ports. 1043 1044.. note:: 1045 1046 When GSO is enabled on a port, supported IPv4 packets transmitted on that 1047 port undergo GSO. Afterwards, the segmented packets are represented by 1048 multi-segment mbufs; however, the csum forwarding engine doesn't calculation 1049 of checksums for GSO'd segments in SW. As a result, if users want correct 1050 checksums in GSO segments, they should enable HW checksum calculation for 1051 GSO-enabled ports. 1052 1053 For example, HW checksum calculation for VxLAN GSO'd packets may be enabled 1054 by setting the following options in the csum forwarding engine: 1055 1056 testpmd> csum set outer_ip hw <port_id> 1057 1058 testpmd> csum set ip hw <port_id> 1059 1060 testpmd> csum set tcp hw <port_id> 1061 1062 UDP GSO is the same as IP fragmentation, which treats the UDP header 1063 as the payload and does not modify it during segmentation. That is, 1064 after UDP GSO, only the first output fragment has the original UDP 1065 header. Therefore, users need to enable HW IP checksum calculation 1066 and SW UDP checksum calculation for GSO-enabled ports, if they want 1067 correct checksums for UDP/IPv4 packets. 1068 1069set gso segsz 1070~~~~~~~~~~~~~ 1071 1072Set the maximum GSO segment size (measured in bytes), which includes the 1073packet header and the packet payload for GSO-enabled ports (global):: 1074 1075 testpmd> set gso segsz <length> 1076 1077show port - gso 1078~~~~~~~~~~~~~~~ 1079 1080Display the status of Generic Segmentation Offload for a given port:: 1081 1082 testpmd> show port <port_id> gso 1083 1084mac_addr add 1085~~~~~~~~~~~~ 1086 1087Add an alternative MAC address to a port:: 1088 1089 testpmd> mac_addr add (port_id) (XX:XX:XX:XX:XX:XX) 1090 1091mac_addr remove 1092~~~~~~~~~~~~~~~ 1093 1094Remove a MAC address from a port:: 1095 1096 testpmd> mac_addr remove (port_id) (XX:XX:XX:XX:XX:XX) 1097 1098mac_addr add (for VF) 1099~~~~~~~~~~~~~~~~~~~~~ 1100 1101Add an alternative MAC address for a VF to a port:: 1102 1103 testpmd> mac_add add port (port_id) vf (vf_id) (XX:XX:XX:XX:XX:XX) 1104 1105mac_addr set 1106~~~~~~~~~~~~ 1107 1108Set the default MAC address for a port:: 1109 1110 testpmd> mac_addr set (port_id) (XX:XX:XX:XX:XX:XX) 1111 1112mac_addr set (for VF) 1113~~~~~~~~~~~~~~~~~~~~~ 1114 1115Set the MAC address for a VF from the PF:: 1116 1117 testpmd> set vf mac addr (port_id) (vf_id) (XX:XX:XX:XX:XX:XX) 1118 1119set eth-peer 1120~~~~~~~~~~~~ 1121 1122Set the forwarding peer address for certain port:: 1123 1124 testpmd> set eth-peer (port_id) (perr_addr) 1125 1126This is equivalent to the ``--eth-peer`` command-line option. 1127 1128set port-uta 1129~~~~~~~~~~~~ 1130 1131Set the unicast hash filter(s) on/off for a port:: 1132 1133 testpmd> set port (port_id) uta (XX:XX:XX:XX:XX:XX|all) (on|off) 1134 1135set promisc 1136~~~~~~~~~~~ 1137 1138Set the promiscuous mode on for a port or for all ports. 1139In promiscuous mode packets are not dropped if they aren't for the specified MAC address:: 1140 1141 testpmd> set promisc (port_id|all) (on|off) 1142 1143set allmulti 1144~~~~~~~~~~~~ 1145 1146Set the allmulti mode for a port or for all ports:: 1147 1148 testpmd> set allmulti (port_id|all) (on|off) 1149 1150Same as the ifconfig (8) option. Controls how multicast packets are handled. 1151 1152set promisc (for VF) 1153~~~~~~~~~~~~~~~~~~~~ 1154 1155Set the unicast promiscuous mode for a VF from PF. 1156It's supported by Intel i40e NICs now. 1157In promiscuous mode packets are not dropped if they aren't for the specified MAC address:: 1158 1159 testpmd> set vf promisc (port_id) (vf_id) (on|off) 1160 1161set allmulticast (for VF) 1162~~~~~~~~~~~~~~~~~~~~~~~~~ 1163 1164Set the multicast promiscuous mode for a VF from PF. 1165It's supported by Intel i40e NICs now. 1166In promiscuous mode packets are not dropped if they aren't for the specified MAC address:: 1167 1168 testpmd> set vf allmulti (port_id) (vf_id) (on|off) 1169 1170set tx max bandwidth (for VF) 1171~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ 1172 1173Set TX max absolute bandwidth (Mbps) for a VF from PF:: 1174 1175 testpmd> set vf tx max-bandwidth (port_id) (vf_id) (max_bandwidth) 1176 1177set tc tx min bandwidth (for VF) 1178~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ 1179 1180Set all TCs' TX min relative bandwidth (%) for a VF from PF:: 1181 1182 testpmd> set vf tc tx min-bandwidth (port_id) (vf_id) (bw1, bw2, ...) 1183 1184set tc tx max bandwidth (for VF) 1185~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ 1186 1187Set a TC's TX max absolute bandwidth (Mbps) for a VF from PF:: 1188 1189 testpmd> set vf tc tx max-bandwidth (port_id) (vf_id) (tc_no) (max_bandwidth) 1190 1191set tc strict link priority mode 1192~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ 1193 1194Set some TCs' strict link priority mode on a physical port:: 1195 1196 testpmd> set tx strict-link-priority (port_id) (tc_bitmap) 1197 1198set tc tx min bandwidth 1199~~~~~~~~~~~~~~~~~~~~~~~ 1200 1201Set all TCs' TX min relative bandwidth (%) globally for all PF and VFs:: 1202 1203 testpmd> set tc tx min-bandwidth (port_id) (bw1, bw2, ...) 1204 1205set flow_ctrl rx 1206~~~~~~~~~~~~~~~~ 1207 1208Set the link flow control parameter on a port:: 1209 1210 testpmd> set flow_ctrl rx (on|off) tx (on|off) (high_water) (low_water) \ 1211 (pause_time) (send_xon) mac_ctrl_frame_fwd (on|off) \ 1212 autoneg (on|off) (port_id) 1213 1214Where: 1215 1216* ``high_water`` (integer): High threshold value to trigger XOFF. 1217 1218* ``low_water`` (integer): Low threshold value to trigger XON. 1219 1220* ``pause_time`` (integer): Pause quota in the Pause frame. 1221 1222* ``send_xon`` (0/1): Send XON frame. 1223 1224* ``mac_ctrl_frame_fwd``: Enable receiving MAC control frames. 1225 1226* ``autoneg``: Change the auto-negotiation parameter. 1227 1228set pfc_ctrl rx 1229~~~~~~~~~~~~~~~ 1230 1231Set the priority flow control parameter on a port:: 1232 1233 testpmd> set pfc_ctrl rx (on|off) tx (on|off) (high_water) (low_water) \ 1234 (pause_time) (priority) (port_id) 1235 1236Where: 1237 1238* ``high_water`` (integer): High threshold value. 1239 1240* ``low_water`` (integer): Low threshold value. 1241 1242* ``pause_time`` (integer): Pause quota in the Pause frame. 1243 1244* ``priority`` (0-7): VLAN User Priority. 1245 1246set stat_qmap 1247~~~~~~~~~~~~~ 1248 1249Set statistics mapping (qmapping 0..15) for RX/TX queue on port:: 1250 1251 testpmd> set stat_qmap (tx|rx) (port_id) (queue_id) (qmapping) 1252 1253For example, to set rx queue 2 on port 0 to mapping 5:: 1254 1255 testpmd>set stat_qmap rx 0 2 5 1256 1257set xstats-hide-zero 1258~~~~~~~~~~~~~~~~~~~~ 1259 1260Set the option to hide zero values for xstats display:: 1261 1262 testpmd> set xstats-hide-zero on|off 1263 1264.. note:: 1265 1266 By default, the zero values are displayed for xstats. 1267 1268set port - rx/tx (for VF) 1269~~~~~~~~~~~~~~~~~~~~~~~~~ 1270 1271Set VF receive/transmit from a port:: 1272 1273 testpmd> set port (port_id) vf (vf_id) (rx|tx) (on|off) 1274 1275set port - mac address filter (for VF) 1276~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ 1277 1278Add/Remove unicast or multicast MAC addr filter for a VF:: 1279 1280 testpmd> set port (port_id) vf (vf_id) (mac_addr) \ 1281 (exact-mac|exact-mac-vlan|hashmac|hashmac-vlan) (on|off) 1282 1283set port - rx mode(for VF) 1284~~~~~~~~~~~~~~~~~~~~~~~~~~ 1285 1286Set the VF receive mode of a port:: 1287 1288 testpmd> set port (port_id) vf (vf_id) \ 1289 rxmode (AUPE|ROPE|BAM|MPE) (on|off) 1290 1291The available receive modes are: 1292 1293* ``AUPE``: Accepts untagged VLAN. 1294 1295* ``ROPE``: Accepts unicast hash. 1296 1297* ``BAM``: Accepts broadcast packets. 1298 1299* ``MPE``: Accepts all multicast packets. 1300 1301set port - tx_rate (for Queue) 1302~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ 1303 1304Set TX rate limitation for a queue on a port:: 1305 1306 testpmd> set port (port_id) queue (queue_id) rate (rate_value) 1307 1308set port - tx_rate (for VF) 1309~~~~~~~~~~~~~~~~~~~~~~~~~~~ 1310 1311Set TX rate limitation for queues in VF on a port:: 1312 1313 testpmd> set port (port_id) vf (vf_id) rate (rate_value) queue_mask (queue_mask) 1314 1315set port - mirror rule 1316~~~~~~~~~~~~~~~~~~~~~~ 1317 1318Set pool or vlan type mirror rule for a port:: 1319 1320 testpmd> set port (port_id) mirror-rule (rule_id) \ 1321 (pool-mirror-up|pool-mirror-down|vlan-mirror) \ 1322 (poolmask|vlanid[,vlanid]*) dst-pool (pool_id) (on|off) 1323 1324Set link mirror rule for a port:: 1325 1326 testpmd> set port (port_id) mirror-rule (rule_id) \ 1327 (uplink-mirror|downlink-mirror) dst-pool (pool_id) (on|off) 1328 1329For example to enable mirror traffic with vlan 0,1 to pool 0:: 1330 1331 set port 0 mirror-rule 0 vlan-mirror 0,1 dst-pool 0 on 1332 1333reset port - mirror rule 1334~~~~~~~~~~~~~~~~~~~~~~~~ 1335 1336Reset a mirror rule for a port:: 1337 1338 testpmd> reset port (port_id) mirror-rule (rule_id) 1339 1340set flush_rx 1341~~~~~~~~~~~~ 1342 1343Set the flush on RX streams before forwarding. 1344The default is flush ``on``. 1345Mainly used with PCAP drivers to turn off the default behavior of flushing the first 512 packets on RX streams:: 1346 1347 testpmd> set flush_rx off 1348 1349set bypass mode 1350~~~~~~~~~~~~~~~ 1351 1352Set the bypass mode for the lowest port on bypass enabled NIC:: 1353 1354 testpmd> set bypass mode (normal|bypass|isolate) (port_id) 1355 1356set bypass event 1357~~~~~~~~~~~~~~~~ 1358 1359Set the event required to initiate specified bypass mode for the lowest port on a bypass enabled:: 1360 1361 testpmd> set bypass event (timeout|os_on|os_off|power_on|power_off) \ 1362 mode (normal|bypass|isolate) (port_id) 1363 1364Where: 1365 1366* ``timeout``: Enable bypass after watchdog timeout. 1367 1368* ``os_on``: Enable bypass when OS/board is powered on. 1369 1370* ``os_off``: Enable bypass when OS/board is powered off. 1371 1372* ``power_on``: Enable bypass when power supply is turned on. 1373 1374* ``power_off``: Enable bypass when power supply is turned off. 1375 1376 1377set bypass timeout 1378~~~~~~~~~~~~~~~~~~ 1379 1380Set the bypass watchdog timeout to ``n`` seconds where 0 = instant:: 1381 1382 testpmd> set bypass timeout (0|1.5|2|3|4|8|16|32) 1383 1384show bypass config 1385~~~~~~~~~~~~~~~~~~ 1386 1387Show the bypass configuration for a bypass enabled NIC using the lowest port on the NIC:: 1388 1389 testpmd> show bypass config (port_id) 1390 1391set link up 1392~~~~~~~~~~~ 1393 1394Set link up for a port:: 1395 1396 testpmd> set link-up port (port id) 1397 1398set link down 1399~~~~~~~~~~~~~ 1400 1401Set link down for a port:: 1402 1403 testpmd> set link-down port (port id) 1404 1405E-tag set 1406~~~~~~~~~ 1407 1408Enable E-tag insertion for a VF on a port:: 1409 1410 testpmd> E-tag set insertion on port-tag-id (value) port (port_id) vf (vf_id) 1411 1412Disable E-tag insertion for a VF on a port:: 1413 1414 testpmd> E-tag set insertion off port (port_id) vf (vf_id) 1415 1416Enable/disable E-tag stripping on a port:: 1417 1418 testpmd> E-tag set stripping (on|off) port (port_id) 1419 1420Enable/disable E-tag based forwarding on a port:: 1421 1422 testpmd> E-tag set forwarding (on|off) port (port_id) 1423 1424Add an E-tag forwarding filter on a port:: 1425 1426 testpmd> E-tag set filter add e-tag-id (value) dst-pool (pool_id) port (port_id) 1427 1428Delete an E-tag forwarding filter on a port:: 1429 testpmd> E-tag set filter del e-tag-id (value) port (port_id) 1430 1431ddp add 1432~~~~~~~ 1433 1434Load a dynamic device personalization (DDP) profile and store backup profile:: 1435 1436 testpmd> ddp add (port_id) (profile_path[,backup_profile_path]) 1437 1438ddp del 1439~~~~~~~ 1440 1441Delete a dynamic device personalization profile and restore backup profile:: 1442 1443 testpmd> ddp del (port_id) (backup_profile_path) 1444 1445ptype mapping 1446~~~~~~~~~~~~~ 1447 1448List all items from the ptype mapping table:: 1449 1450 testpmd> ptype mapping get (port_id) (valid_only) 1451 1452Where: 1453 1454* ``valid_only``: A flag indicates if only list valid items(=1) or all itemss(=0). 1455 1456Replace a specific or a group of software defined ptype with a new one:: 1457 1458 testpmd> ptype mapping replace (port_id) (target) (mask) (pkt_type) 1459 1460where: 1461 1462* ``target``: A specific software ptype or a mask to represent a group of software ptypes. 1463 1464* ``mask``: A flag indicate if "target" is a specific software ptype(=0) or a ptype mask(=1). 1465 1466* ``pkt_type``: The new software ptype to replace the old ones. 1467 1468Update hardware defined ptype to software defined packet type mapping table:: 1469 1470 testpmd> ptype mapping update (port_id) (hw_ptype) (sw_ptype) 1471 1472where: 1473 1474* ``hw_ptype``: hardware ptype as the index of the ptype mapping table. 1475 1476* ``sw_ptype``: software ptype as the value of the ptype mapping table. 1477 1478Reset ptype mapping table:: 1479 1480 testpmd> ptype mapping reset (port_id) 1481 1482config per port Rx offloading 1483~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ 1484 1485Enable or disable a per port Rx offloading on all Rx queues of a port:: 1486 1487 testpmd> port config (port_id) rx_offload (offloading) on|off 1488 1489* ``offloading``: can be any of these offloading capability: 1490 vlan_strip, ipv4_cksum, udp_cksum, tcp_cksum, tcp_lro, 1491 qinq_strip, outer_ipv4_cksum, macsec_strip, 1492 header_split, vlan_filter, vlan_extend, jumbo_frame, 1493 crc_strip, scatter, timestamp, security 1494 1495This command should be run when the port is stopped, or else it will fail. 1496 1497config per queue Rx offloading 1498~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ 1499 1500Enable or disable a per queue Rx offloading only on a specific Rx queue:: 1501 1502 testpmd> port (port_id) rxq (queue_id) rx_offload (offloading) on|off 1503 1504* ``offloading``: can be any of these offloading capability: 1505 vlan_strip, ipv4_cksum, udp_cksum, tcp_cksum, tcp_lro, 1506 qinq_strip, outer_ipv4_cksum, macsec_strip, 1507 header_split, vlan_filter, vlan_extend, jumbo_frame, 1508 crc_strip, scatter, timestamp, security 1509 1510This command should be run when the port is stopped, or else it will fail. 1511 1512config per port Tx offloading 1513~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ 1514 1515Enable or disable a per port Tx offloading on all Tx queues of a port:: 1516 1517 testpmd> port config (port_id) tx_offload (offloading) on|off 1518 1519* ``offloading``: can be any of these offloading capability: 1520 vlan_insert, ipv4_cksum, udp_cksum, tcp_cksum, 1521 sctp_cksum, tcp_tso, udp_tso, outer_ipv4_cksum, 1522 qinq_insert, vxlan_tnl_tso, gre_tnl_tso, 1523 ipip_tnl_tso, geneve_tnl_tso, macsec_insert, 1524 mt_lockfree, multi_segs, fast_free, security 1525 1526This command should be run when the port is stopped, or else it will fail. 1527 1528config per queue Tx offloading 1529~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ 1530 1531Enable or disable a per queue Tx offloading only on a specific Tx queue:: 1532 1533 testpmd> port (port_id) txq (queue_id) tx_offload (offloading) on|off 1534 1535* ``offloading``: can be any of these offloading capability: 1536 vlan_insert, ipv4_cksum, udp_cksum, tcp_cksum, 1537 sctp_cksum, tcp_tso, udp_tso, outer_ipv4_cksum, 1538 qinq_insert, vxlan_tnl_tso, gre_tnl_tso, 1539 ipip_tnl_tso, geneve_tnl_tso, macsec_insert, 1540 mt_lockfree, multi_segs, fast_free, security 1541 1542This command should be run when the port is stopped, or else it will fail. 1543 1544 1545Port Functions 1546-------------- 1547 1548The following sections show functions for configuring ports. 1549 1550.. note:: 1551 1552 Port configuration changes only become active when forwarding is started/restarted. 1553 1554port attach 1555~~~~~~~~~~~ 1556 1557Attach a port specified by pci address or virtual device args:: 1558 1559 testpmd> port attach (identifier) 1560 1561To attach a new pci device, the device should be recognized by kernel first. 1562Then it should be moved under DPDK management. 1563Finally the port can be attached to testpmd. 1564 1565For example, to move a pci device using ixgbe under DPDK management: 1566 1567.. code-block:: console 1568 1569 # Check the status of the available devices. 1570 ./usertools/dpdk-devbind.py --status 1571 1572 Network devices using DPDK-compatible driver 1573 ============================================ 1574 <none> 1575 1576 Network devices using kernel driver 1577 =================================== 1578 0000:0a:00.0 '82599ES 10-Gigabit' if=eth2 drv=ixgbe unused= 1579 1580 1581 # Bind the device to igb_uio. 1582 sudo ./usertools/dpdk-devbind.py -b igb_uio 0000:0a:00.0 1583 1584 1585 # Recheck the status of the devices. 1586 ./usertools/dpdk-devbind.py --status 1587 Network devices using DPDK-compatible driver 1588 ============================================ 1589 0000:0a:00.0 '82599ES 10-Gigabit' drv=igb_uio unused= 1590 1591To attach a port created by virtual device, above steps are not needed. 1592 1593For example, to attach a port whose pci address is 0000:0a:00.0. 1594 1595.. code-block:: console 1596 1597 testpmd> port attach 0000:0a:00.0 1598 Attaching a new port... 1599 EAL: PCI device 0000:0a:00.0 on NUMA socket -1 1600 EAL: probe driver: 8086:10fb rte_ixgbe_pmd 1601 EAL: PCI memory mapped at 0x7f83bfa00000 1602 EAL: PCI memory mapped at 0x7f83bfa80000 1603 PMD: eth_ixgbe_dev_init(): MAC: 2, PHY: 18, SFP+: 5 1604 PMD: eth_ixgbe_dev_init(): port 0 vendorID=0x8086 deviceID=0x10fb 1605 Port 0 is attached. Now total ports is 1 1606 Done 1607 1608For example, to attach a port created by pcap PMD. 1609 1610.. code-block:: console 1611 1612 testpmd> port attach net_pcap0 1613 Attaching a new port... 1614 PMD: Initializing pmd_pcap for net_pcap0 1615 PMD: Creating pcap-backed ethdev on numa socket 0 1616 Port 0 is attached. Now total ports is 1 1617 Done 1618 1619In this case, identifier is ``net_pcap0``. 1620This identifier format is the same as ``--vdev`` format of DPDK applications. 1621 1622For example, to re-attach a bonded port which has been previously detached, 1623the mode and slave parameters must be given. 1624 1625.. code-block:: console 1626 1627 testpmd> port attach net_bond_0,mode=0,slave=1 1628 Attaching a new port... 1629 EAL: Initializing pmd_bond for net_bond_0 1630 EAL: Create bonded device net_bond_0 on port 0 in mode 0 on socket 0. 1631 Port 0 is attached. Now total ports is 1 1632 Done 1633 1634 1635port detach 1636~~~~~~~~~~~ 1637 1638Detach a specific port:: 1639 1640 testpmd> port detach (port_id) 1641 1642Before detaching a port, the port should be stopped and closed. 1643 1644For example, to detach a pci device port 0. 1645 1646.. code-block:: console 1647 1648 testpmd> port stop 0 1649 Stopping ports... 1650 Done 1651 testpmd> port close 0 1652 Closing ports... 1653 Done 1654 1655 testpmd> port detach 0 1656 Detaching a port... 1657 EAL: PCI device 0000:0a:00.0 on NUMA socket -1 1658 EAL: remove driver: 8086:10fb rte_ixgbe_pmd 1659 EAL: PCI memory unmapped at 0x7f83bfa00000 1660 EAL: PCI memory unmapped at 0x7f83bfa80000 1661 Done 1662 1663 1664For example, to detach a virtual device port 0. 1665 1666.. code-block:: console 1667 1668 testpmd> port stop 0 1669 Stopping ports... 1670 Done 1671 testpmd> port close 0 1672 Closing ports... 1673 Done 1674 1675 testpmd> port detach 0 1676 Detaching a port... 1677 PMD: Closing pcap ethdev on numa socket 0 1678 Port 'net_pcap0' is detached. Now total ports is 0 1679 Done 1680 1681To remove a pci device completely from the system, first detach the port from testpmd. 1682Then the device should be moved under kernel management. 1683Finally the device can be removed using kernel pci hotplug functionality. 1684 1685For example, to move a pci device under kernel management: 1686 1687.. code-block:: console 1688 1689 sudo ./usertools/dpdk-devbind.py -b ixgbe 0000:0a:00.0 1690 1691 ./usertools/dpdk-devbind.py --status 1692 1693 Network devices using DPDK-compatible driver 1694 ============================================ 1695 <none> 1696 1697 Network devices using kernel driver 1698 =================================== 1699 0000:0a:00.0 '82599ES 10-Gigabit' if=eth2 drv=ixgbe unused=igb_uio 1700 1701To remove a port created by a virtual device, above steps are not needed. 1702 1703port start 1704~~~~~~~~~~ 1705 1706Start all ports or a specific port:: 1707 1708 testpmd> port start (port_id|all) 1709 1710port stop 1711~~~~~~~~~ 1712 1713Stop all ports or a specific port:: 1714 1715 testpmd> port stop (port_id|all) 1716 1717port close 1718~~~~~~~~~~ 1719 1720Close all ports or a specific port:: 1721 1722 testpmd> port close (port_id|all) 1723 1724port config - queue ring size 1725~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ 1726 1727Configure a rx/tx queue ring size:: 1728 1729 testpmd> port (port_id) (rxq|txq) (queue_id) ring_size (value) 1730 1731Only take effect after command that (re-)start the port or command that setup specific queue. 1732 1733port start/stop queue 1734~~~~~~~~~~~~~~~~~~~~~ 1735 1736Start/stop a rx/tx queue on a specific port:: 1737 1738 testpmd> port (port_id) (rxq|txq) (queue_id) (start|stop) 1739 1740port setup queue 1741~~~~~~~~~~~~~~~~~~~~~ 1742 1743Setup a rx/tx queue on a specific port:: 1744 1745 testpmd> port (port_id) (rxq|txq) (queue_id) setup 1746 1747Only take effect when port is started. 1748 1749port config - speed 1750~~~~~~~~~~~~~~~~~~~ 1751 1752Set the speed and duplex mode for all ports or a specific port:: 1753 1754 testpmd> port config (port_id|all) speed (10|100|1000|10000|25000|40000|50000|100000|auto) \ 1755 duplex (half|full|auto) 1756 1757port config - queues/descriptors 1758~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ 1759 1760Set number of queues/descriptors for rxq, txq, rxd and txd:: 1761 1762 testpmd> port config all (rxq|txq|rxd|txd) (value) 1763 1764This is equivalent to the ``--rxq``, ``--txq``, ``--rxd`` and ``--txd`` command-line options. 1765 1766port config - max-pkt-len 1767~~~~~~~~~~~~~~~~~~~~~~~~~ 1768 1769Set the maximum packet length:: 1770 1771 testpmd> port config all max-pkt-len (value) 1772 1773This is equivalent to the ``--max-pkt-len`` command-line option. 1774 1775port config - CRC Strip 1776~~~~~~~~~~~~~~~~~~~~~~~ 1777 1778Set hardware CRC stripping on or off for all ports:: 1779 1780 testpmd> port config all crc-strip (on|off) 1781 1782CRC stripping is on by default. 1783 1784The ``off`` option is equivalent to the ``--disable-crc-strip`` command-line option. 1785 1786port config - scatter 1787~~~~~~~~~~~~~~~~~~~~~~~ 1788 1789Set RX scatter mode on or off for all ports:: 1790 1791 testpmd> port config all scatter (on|off) 1792 1793RX scatter mode is off by default. 1794 1795The ``on`` option is equivalent to the ``--enable-scatter`` command-line option. 1796 1797port config - RX Checksum 1798~~~~~~~~~~~~~~~~~~~~~~~~~ 1799 1800Set hardware RX checksum offload to on or off for all ports:: 1801 1802 testpmd> port config all rx-cksum (on|off) 1803 1804Checksum offload is off by default. 1805 1806The ``on`` option is equivalent to the ``--enable-rx-cksum`` command-line option. 1807 1808port config - VLAN 1809~~~~~~~~~~~~~~~~~~ 1810 1811Set hardware VLAN on or off for all ports:: 1812 1813 testpmd> port config all hw-vlan (on|off) 1814 1815Hardware VLAN is off by default. 1816 1817The ``on`` option is equivalent to the ``--enable-hw-vlan`` command-line option. 1818 1819port config - VLAN filter 1820~~~~~~~~~~~~~~~~~~~~~~~~~ 1821 1822Set hardware VLAN filter on or off for all ports:: 1823 1824 testpmd> port config all hw-vlan-filter (on|off) 1825 1826Hardware VLAN filter is off by default. 1827 1828The ``on`` option is equivalent to the ``--enable-hw-vlan-filter`` command-line option. 1829 1830port config - VLAN strip 1831~~~~~~~~~~~~~~~~~~~~~~~~ 1832 1833Set hardware VLAN strip on or off for all ports:: 1834 1835 testpmd> port config all hw-vlan-strip (on|off) 1836 1837Hardware VLAN strip is off by default. 1838 1839The ``on`` option is equivalent to the ``--enable-hw-vlan-strip`` command-line option. 1840 1841port config - VLAN extend 1842~~~~~~~~~~~~~~~~~~~~~~~~~ 1843 1844Set hardware VLAN extend on or off for all ports:: 1845 1846 testpmd> port config all hw-vlan-extend (on|off) 1847 1848Hardware VLAN extend is off by default. 1849 1850The ``on`` option is equivalent to the ``--enable-hw-vlan-extend`` command-line option. 1851 1852port config - Drop Packets 1853~~~~~~~~~~~~~~~~~~~~~~~~~~ 1854 1855Set packet drop for packets with no descriptors on or off for all ports:: 1856 1857 testpmd> port config all drop-en (on|off) 1858 1859Packet dropping for packets with no descriptors is off by default. 1860 1861The ``on`` option is equivalent to the ``--enable-drop-en`` command-line option. 1862 1863port config - RSS 1864~~~~~~~~~~~~~~~~~ 1865 1866Set the RSS (Receive Side Scaling) mode on or off:: 1867 1868 testpmd> port config all rss (all|default|ip|tcp|udp|sctp|ether|port|vxlan|geneve|nvgre|none) 1869 1870RSS is on by default. 1871 1872The ``all`` option is equivalent to ip|tcp|udp|sctp|ether. 1873The ``default`` option enables all supported RSS types reported by device info. 1874The ``none`` option is equivalent to the ``--disable-rss`` command-line option. 1875 1876port config - RSS Reta 1877~~~~~~~~~~~~~~~~~~~~~~ 1878 1879Set the RSS (Receive Side Scaling) redirection table:: 1880 1881 testpmd> port config all rss reta (hash,queue)[,(hash,queue)] 1882 1883port config - DCB 1884~~~~~~~~~~~~~~~~~ 1885 1886Set the DCB mode for an individual port:: 1887 1888 testpmd> port config (port_id) dcb vt (on|off) (traffic_class) pfc (on|off) 1889 1890The traffic class should be 4 or 8. 1891 1892port config - Burst 1893~~~~~~~~~~~~~~~~~~~ 1894 1895Set the number of packets per burst:: 1896 1897 testpmd> port config all burst (value) 1898 1899This is equivalent to the ``--burst`` command-line option. 1900 1901port config - Threshold 1902~~~~~~~~~~~~~~~~~~~~~~~ 1903 1904Set thresholds for TX/RX queues:: 1905 1906 testpmd> port config all (threshold) (value) 1907 1908Where the threshold type can be: 1909 1910* ``txpt:`` Set the prefetch threshold register of the TX rings, 0 <= value <= 255. 1911 1912* ``txht:`` Set the host threshold register of the TX rings, 0 <= value <= 255. 1913 1914* ``txwt:`` Set the write-back threshold register of the TX rings, 0 <= value <= 255. 1915 1916* ``rxpt:`` Set the prefetch threshold register of the RX rings, 0 <= value <= 255. 1917 1918* ``rxht:`` Set the host threshold register of the RX rings, 0 <= value <= 255. 1919 1920* ``rxwt:`` Set the write-back threshold register of the RX rings, 0 <= value <= 255. 1921 1922* ``txfreet:`` Set the transmit free threshold of the TX rings, 0 <= value <= txd. 1923 1924* ``rxfreet:`` Set the transmit free threshold of the RX rings, 0 <= value <= rxd. 1925 1926* ``txrst:`` Set the transmit RS bit threshold of TX rings, 0 <= value <= txd. 1927 1928These threshold options are also available from the command-line. 1929 1930port config - E-tag 1931~~~~~~~~~~~~~~~~~~~ 1932 1933Set the value of ether-type for E-tag:: 1934 1935 testpmd> port config (port_id|all) l2-tunnel E-tag ether-type (value) 1936 1937Enable/disable the E-tag support:: 1938 1939 testpmd> port config (port_id|all) l2-tunnel E-tag (enable|disable) 1940 1941port config pctype mapping 1942~~~~~~~~~~~~~~~~~~~~~~~~~~ 1943 1944Reset pctype mapping table:: 1945 1946 testpmd> port config (port_id) pctype mapping reset 1947 1948Update hardware defined pctype to software defined flow type mapping table:: 1949 1950 testpmd> port config (port_id) pctype mapping update (pctype_id_0[,pctype_id_1]*) (flow_type_id) 1951 1952where: 1953 1954* ``pctype_id_x``: hardware pctype id as index of bit in bitmask value of the pctype mapping table. 1955 1956* ``flow_type_id``: software flow type id as the index of the pctype mapping table. 1957 1958port config input set 1959~~~~~~~~~~~~~~~~~~~~~ 1960 1961Config RSS/FDIR/FDIR flexible payload input set for some pctype:: 1962 testpmd> port config (port_id) pctype (pctype_id) \ 1963 (hash_inset|fdir_inset|fdir_flx_inset) \ 1964 (get|set|clear) field (field_idx) 1965 1966Clear RSS/FDIR/FDIR flexible payload input set for some pctype:: 1967 testpmd> port config (port_id) pctype (pctype_id) \ 1968 (hash_inset|fdir_inset|fdir_flx_inset) clear all 1969 1970where: 1971 1972* ``pctype_id``: hardware packet classification types. 1973* ``field_idx``: hardware field index. 1974 1975port config udp_tunnel_port 1976~~~~~~~~~~~~~~~~~~~~~~~~~~~ 1977 1978Add/remove UDP tunnel port for VXLAN/GENEVE tunneling protocols:: 1979 testpmd> port config (port_id) udp_tunnel_port add|rm vxlan|geneve (udp_port) 1980 1981Link Bonding Functions 1982---------------------- 1983 1984The Link Bonding functions make it possible to dynamically create and 1985manage link bonding devices from within testpmd interactive prompt. 1986 1987create bonded device 1988~~~~~~~~~~~~~~~~~~~~ 1989 1990Create a new bonding device:: 1991 1992 testpmd> create bonded device (mode) (socket) 1993 1994For example, to create a bonded device in mode 1 on socket 0:: 1995 1996 testpmd> create bonded 1 0 1997 created new bonded device (port X) 1998 1999add bonding slave 2000~~~~~~~~~~~~~~~~~ 2001 2002Adds Ethernet device to a Link Bonding device:: 2003 2004 testpmd> add bonding slave (slave id) (port id) 2005 2006For example, to add Ethernet device (port 6) to a Link Bonding device (port 10):: 2007 2008 testpmd> add bonding slave 6 10 2009 2010 2011remove bonding slave 2012~~~~~~~~~~~~~~~~~~~~ 2013 2014Removes an Ethernet slave device from a Link Bonding device:: 2015 2016 testpmd> remove bonding slave (slave id) (port id) 2017 2018For example, to remove Ethernet slave device (port 6) to a Link Bonding device (port 10):: 2019 2020 testpmd> remove bonding slave 6 10 2021 2022set bonding mode 2023~~~~~~~~~~~~~~~~ 2024 2025Set the Link Bonding mode of a Link Bonding device:: 2026 2027 testpmd> set bonding mode (value) (port id) 2028 2029For example, to set the bonding mode of a Link Bonding device (port 10) to broadcast (mode 3):: 2030 2031 testpmd> set bonding mode 3 10 2032 2033set bonding primary 2034~~~~~~~~~~~~~~~~~~~ 2035 2036Set an Ethernet slave device as the primary device on a Link Bonding device:: 2037 2038 testpmd> set bonding primary (slave id) (port id) 2039 2040For example, to set the Ethernet slave device (port 6) as the primary port of a Link Bonding device (port 10):: 2041 2042 testpmd> set bonding primary 6 10 2043 2044set bonding mac 2045~~~~~~~~~~~~~~~ 2046 2047Set the MAC address of a Link Bonding device:: 2048 2049 testpmd> set bonding mac (port id) (mac) 2050 2051For example, to set the MAC address of a Link Bonding device (port 10) to 00:00:00:00:00:01:: 2052 2053 testpmd> set bonding mac 10 00:00:00:00:00:01 2054 2055set bonding xmit_balance_policy 2056~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ 2057 2058Set the transmission policy for a Link Bonding device when it is in Balance XOR mode:: 2059 2060 testpmd> set bonding xmit_balance_policy (port_id) (l2|l23|l34) 2061 2062For example, set a Link Bonding device (port 10) to use a balance policy of layer 3+4 (IP addresses & UDP ports):: 2063 2064 testpmd> set bonding xmit_balance_policy 10 l34 2065 2066 2067set bonding mon_period 2068~~~~~~~~~~~~~~~~~~~~~~ 2069 2070Set the link status monitoring polling period in milliseconds for a bonding device. 2071 2072This adds support for PMD slave devices which do not support link status interrupts. 2073When the mon_period is set to a value greater than 0 then all PMD's which do not support 2074link status ISR will be queried every polling interval to check if their link status has changed:: 2075 2076 testpmd> set bonding mon_period (port_id) (value) 2077 2078For example, to set the link status monitoring polling period of bonded device (port 5) to 150ms:: 2079 2080 testpmd> set bonding mon_period 5 150 2081 2082 2083set bonding lacp dedicated_queue 2084~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ 2085 2086Enable dedicated tx/rx queues on bonding devices slaves to handle LACP control plane traffic 2087when in mode 4 (link-aggregration-802.3ad):: 2088 2089 testpmd> set bonding lacp dedicated_queues (port_id) (enable|disable) 2090 2091 2092set bonding agg_mode 2093~~~~~~~~~~~~~~~~~~~~ 2094 2095Enable one of the specific aggregators mode when in mode 4 (link-aggregration-802.3ad):: 2096 2097 testpmd> set bonding agg_mode (port_id) (bandwidth|count|stable) 2098 2099 2100show bonding config 2101~~~~~~~~~~~~~~~~~~~ 2102 2103Show the current configuration of a Link Bonding device:: 2104 2105 testpmd> show bonding config (port id) 2106 2107For example, 2108to show the configuration a Link Bonding device (port 9) with 3 slave devices (1, 3, 4) 2109in balance mode with a transmission policy of layer 2+3:: 2110 2111 testpmd> show bonding config 9 2112 Bonding mode: 2 2113 Balance Xmit Policy: BALANCE_XMIT_POLICY_LAYER23 2114 Slaves (3): [1 3 4] 2115 Active Slaves (3): [1 3 4] 2116 Primary: [3] 2117 2118 2119Register Functions 2120------------------ 2121 2122The Register Functions can be used to read from and write to registers on the network card referenced by a port number. 2123This is mainly useful for debugging purposes. 2124Reference should be made to the appropriate datasheet for the network card for details on the register addresses 2125and fields that can be accessed. 2126 2127read reg 2128~~~~~~~~ 2129 2130Display the value of a port register:: 2131 2132 testpmd> read reg (port_id) (address) 2133 2134For example, to examine the Flow Director control register (FDIRCTL, 0x0000EE000) on an Intel 82599 10 GbE Controller:: 2135 2136 testpmd> read reg 0 0xEE00 2137 port 0 PCI register at offset 0xEE00: 0x4A060029 (1241907241) 2138 2139read regfield 2140~~~~~~~~~~~~~ 2141 2142Display a port register bit field:: 2143 2144 testpmd> read regfield (port_id) (address) (bit_x) (bit_y) 2145 2146For example, reading the lowest two bits from the register in the example above:: 2147 2148 testpmd> read regfield 0 0xEE00 0 1 2149 port 0 PCI register at offset 0xEE00: bits[0, 1]=0x1 (1) 2150 2151read regbit 2152~~~~~~~~~~~ 2153 2154Display a single port register bit:: 2155 2156 testpmd> read regbit (port_id) (address) (bit_x) 2157 2158For example, reading the lowest bit from the register in the example above:: 2159 2160 testpmd> read regbit 0 0xEE00 0 2161 port 0 PCI register at offset 0xEE00: bit 0=1 2162 2163write reg 2164~~~~~~~~~ 2165 2166Set the value of a port register:: 2167 2168 testpmd> write reg (port_id) (address) (value) 2169 2170For example, to clear a register:: 2171 2172 testpmd> write reg 0 0xEE00 0x0 2173 port 0 PCI register at offset 0xEE00: 0x00000000 (0) 2174 2175write regfield 2176~~~~~~~~~~~~~~ 2177 2178Set bit field of a port register:: 2179 2180 testpmd> write regfield (port_id) (address) (bit_x) (bit_y) (value) 2181 2182For example, writing to the register cleared in the example above:: 2183 2184 testpmd> write regfield 0 0xEE00 0 1 2 2185 port 0 PCI register at offset 0xEE00: 0x00000002 (2) 2186 2187write regbit 2188~~~~~~~~~~~~ 2189 2190Set single bit value of a port register:: 2191 2192 testpmd> write regbit (port_id) (address) (bit_x) (value) 2193 2194For example, to set the high bit in the register from the example above:: 2195 2196 testpmd> write regbit 0 0xEE00 31 1 2197 port 0 PCI register at offset 0xEE00: 0x8000000A (2147483658) 2198 2199Traffic Metering and Policing 2200----------------------------- 2201 2202The following section shows functions for configuring traffic metering and 2203policing on the ethernet device through the use of generic ethdev API. 2204 2205show port traffic management capability 2206~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ 2207 2208Show traffic metering and policing capability of the port:: 2209 2210 testpmd> show port meter cap (port_id) 2211 2212add port meter profile (srTCM rfc2967) 2213~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ 2214 2215Add meter profile (srTCM rfc2697) to the ethernet device:: 2216 2217 testpmd> add port meter profile srtcm_rfc2697 (port_id) (profile_id) \ 2218 (cir) (cbs) (ebs) 2219 2220where: 2221 2222* ``profile_id``: ID for the meter profile. 2223* ``cir``: Committed Information Rate (CIR) (bytes/second). 2224* ``cbs``: Committed Burst Size (CBS) (bytes). 2225* ``ebs``: Excess Burst Size (EBS) (bytes). 2226 2227add port meter profile (trTCM rfc2968) 2228~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ 2229 2230Add meter profile (srTCM rfc2698) to the ethernet device:: 2231 2232 testpmd> add port meter profile trtcm_rfc2698 (port_id) (profile_id) \ 2233 (cir) (pir) (cbs) (pbs) 2234 2235where: 2236 2237* ``profile_id``: ID for the meter profile. 2238* ``cir``: Committed information rate (bytes/second). 2239* ``pir``: Peak information rate (bytes/second). 2240* ``cbs``: Committed burst size (bytes). 2241* ``pbs``: Peak burst size (bytes). 2242 2243add port meter profile (trTCM rfc4115) 2244~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ 2245 2246Add meter profile (trTCM rfc4115) to the ethernet device:: 2247 2248 testpmd> add port meter profile trtcm_rfc4115 (port_id) (profile_id) \ 2249 (cir) (eir) (cbs) (ebs) 2250 2251where: 2252 2253* ``profile_id``: ID for the meter profile. 2254* ``cir``: Committed information rate (bytes/second). 2255* ``eir``: Excess information rate (bytes/second). 2256* ``cbs``: Committed burst size (bytes). 2257* ``ebs``: Excess burst size (bytes). 2258 2259delete port meter profile 2260~~~~~~~~~~~~~~~~~~~~~~~~~ 2261 2262Delete meter profile from the ethernet device:: 2263 2264 testpmd> del port meter profile (port_id) (profile_id) 2265 2266create port meter 2267~~~~~~~~~~~~~~~~~ 2268 2269Create new meter object for the ethernet device:: 2270 2271 testpmd> create port meter (port_id) (mtr_id) (profile_id) \ 2272 (meter_enable) (g_action) (y_action) (r_action) (stats_mask) (shared) \ 2273 (use_pre_meter_color) [(dscp_tbl_entry0) (dscp_tbl_entry1)...\ 2274 (dscp_tbl_entry63)] 2275 2276where: 2277 2278* ``mtr_id``: meter object ID. 2279* ``profile_id``: ID for the meter profile. 2280* ``meter_enable``: When this parameter has a non-zero value, the meter object 2281 gets enabled at the time of creation, otherwise remains disabled. 2282* ``g_action``: Policer action for the packet with green color. 2283* ``y_action``: Policer action for the packet with yellow color. 2284* ``r_action``: Policer action for the packet with red color. 2285* ``stats_mask``: Mask of statistics counter types to be enabled for the 2286 meter object. 2287* ``shared``: When this parameter has a non-zero value, the meter object is 2288 shared by multiple flows. Otherwise, meter object is used by single flow. 2289* ``use_pre_meter_color``: When this parameter has a non-zero value, the 2290 input color for the current meter object is determined by the latest meter 2291 object in the same flow. Otherwise, the current meter object uses the 2292 *dscp_table* to determine the input color. 2293* ``dscp_tbl_entryx``: DSCP table entry x providing meter providing input 2294 color, 0 <= x <= 63. 2295 2296enable port meter 2297~~~~~~~~~~~~~~~~~ 2298 2299Enable meter for the ethernet device:: 2300 2301 testpmd> enable port meter (port_id) (mtr_id) 2302 2303disable port meter 2304~~~~~~~~~~~~~~~~~~ 2305 2306Disable meter for the ethernet device:: 2307 2308 testpmd> disable port meter (port_id) (mtr_id) 2309 2310delete port meter 2311~~~~~~~~~~~~~~~~~ 2312 2313Delete meter for the ethernet device:: 2314 2315 testpmd> del port meter (port_id) (mtr_id) 2316 2317Set port meter profile 2318~~~~~~~~~~~~~~~~~~~~~~ 2319 2320Set meter profile for the ethernet device:: 2321 2322 testpmd> set port meter profile (port_id) (mtr_id) (profile_id) 2323 2324set port meter dscp table 2325~~~~~~~~~~~~~~~~~~~~~~~~~ 2326 2327Set meter dscp table for the ethernet device:: 2328 2329 testpmd> set port meter dscp table (port_id) (mtr_id) [(dscp_tbl_entry0) \ 2330 (dscp_tbl_entry1)...(dscp_tbl_entry63)] 2331 2332set port meter policer action 2333~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ 2334 2335Set meter policer action for the ethernet device:: 2336 2337 testpmd> set port meter policer action (port_id) (mtr_id) (action_mask) \ 2338 (action0) [(action1) (action1)] 2339 2340where: 2341 2342* ``action_mask``: Bit mask indicating which policer actions need to be 2343 updated. One or more policer actions can be updated in a single function 2344 invocation. To update the policer action associated with color C, bit 2345 (1 << C) needs to be set in *action_mask* and element at position C 2346 in the *actions* array needs to be valid. 2347* ``actionx``: Policer action for the color x, 2348 RTE_MTR_GREEN <= x < RTE_MTR_COLORS 2349 2350set port meter stats mask 2351~~~~~~~~~~~~~~~~~~~~~~~~~ 2352 2353Set meter stats mask for the ethernet device:: 2354 2355 testpmd> set port meter stats mask (port_id) (mtr_id) (stats_mask) 2356 2357where: 2358 2359* ``stats_mask``: Bit mask indicating statistics counter types to be enabled. 2360 2361show port meter stats 2362~~~~~~~~~~~~~~~~~~~~~ 2363 2364Show meter stats of the ethernet device:: 2365 2366 testpmd> show port meter stats (port_id) (mtr_id) (clear) 2367 2368where: 2369 2370* ``clear``: Flag that indicates whether the statistics counters should 2371 be cleared (i.e. set to zero) immediately after they have been read or not. 2372 2373Traffic Management 2374------------------ 2375 2376The following section shows functions for configuring traffic management on 2377on the ethernet device through the use of generic TM API. 2378 2379show port traffic management capability 2380~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ 2381 2382Show traffic management capability of the port:: 2383 2384 testpmd> show port tm cap (port_id) 2385 2386show port traffic management capability (hierarchy level) 2387~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ 2388 2389Show traffic management hierarchy level capability of the port:: 2390 2391 testpmd> show port tm level cap (port_id) (level_id) 2392 2393show port traffic management capability (hierarchy node level) 2394~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ 2395 2396Show the traffic management hierarchy node capability of the port:: 2397 2398 testpmd> show port tm node cap (port_id) (node_id) 2399 2400show port traffic management hierarchy node type 2401~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ 2402 2403Show the port traffic management hierarchy node type:: 2404 2405 testpmd> show port tm node type (port_id) (node_id) 2406 2407show port traffic management hierarchy node stats 2408~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ 2409 2410Show the port traffic management hierarchy node statistics:: 2411 2412 testpmd> show port tm node stats (port_id) (node_id) (clear) 2413 2414where: 2415 2416* ``clear``: When this parameter has a non-zero value, the statistics counters 2417 are cleared (i.e. set to zero) immediately after they have been read, 2418 otherwise the statistics counters are left untouched. 2419 2420Add port traffic management private shaper profile 2421~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ 2422 2423Add the port traffic management private shaper profile:: 2424 2425 testpmd> add port tm node shaper profile (port_id) (shaper_profile_id) \ 2426 (tb_rate) (tb_size) (packet_length_adjust) 2427 2428where: 2429 2430* ``shaper_profile id``: Shaper profile ID for the new profile. 2431* ``tb_rate``: Token bucket rate (bytes per second). 2432* ``tb_size``: Token bucket size (bytes). 2433* ``packet_length_adjust``: The value (bytes) to be added to the length of 2434 each packet for the purpose of shaping. This parameter value can be used to 2435 correct the packet length with the framing overhead bytes that are consumed 2436 on the wire. 2437 2438Delete port traffic management private shaper profile 2439~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ 2440 2441Delete the port traffic management private shaper:: 2442 2443 testpmd> del port tm node shaper profile (port_id) (shaper_profile_id) 2444 2445where: 2446 2447* ``shaper_profile id``: Shaper profile ID that needs to be deleted. 2448 2449Add port traffic management shared shaper 2450~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ 2451 2452Create the port traffic management shared shaper:: 2453 2454 testpmd> add port tm node shared shaper (port_id) (shared_shaper_id) \ 2455 (shaper_profile_id) 2456 2457where: 2458 2459* ``shared_shaper_id``: Shared shaper ID to be created. 2460* ``shaper_profile id``: Shaper profile ID for shared shaper. 2461 2462Set port traffic management shared shaper 2463~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ 2464 2465Update the port traffic management shared shaper:: 2466 2467 testpmd> set port tm node shared shaper (port_id) (shared_shaper_id) \ 2468 (shaper_profile_id) 2469 2470where: 2471 2472* ``shared_shaper_id``: Shared shaper ID to be update. 2473* ``shaper_profile id``: Shaper profile ID for shared shaper. 2474 2475Delete port traffic management shared shaper 2476~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ 2477 2478Delete the port traffic management shared shaper:: 2479 2480 testpmd> del port tm node shared shaper (port_id) (shared_shaper_id) 2481 2482where: 2483 2484* ``shared_shaper_id``: Shared shaper ID to be deleted. 2485 2486Set port traffic management hiearchy node private shaper 2487~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ 2488 2489set the port traffic management hierarchy node private shaper:: 2490 2491 testpmd> set port tm node shaper profile (port_id) (node_id) \ 2492 (shaper_profile_id) 2493 2494where: 2495 2496* ``shaper_profile id``: Private shaper profile ID to be enabled on the 2497 hierarchy node. 2498 2499Add port traffic management WRED profile 2500~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ 2501 2502Create a new WRED profile:: 2503 2504 testpmd> add port tm node wred profile (port_id) (wred_profile_id) \ 2505 (color_g) (min_th_g) (max_th_g) (maxp_inv_g) (wq_log2_g) \ 2506 (color_y) (min_th_y) (max_th_y) (maxp_inv_y) (wq_log2_y) \ 2507 (color_r) (min_th_r) (max_th_r) (maxp_inv_r) (wq_log2_r) 2508 2509where: 2510 2511* ``wred_profile id``: Identifier for the newly create WRED profile 2512* ``color_g``: Packet color (green) 2513* ``min_th_g``: Minimum queue threshold for packet with green color 2514* ``max_th_g``: Minimum queue threshold for packet with green color 2515* ``maxp_inv_g``: Inverse of packet marking probability maximum value (maxp) 2516* ``wq_log2_g``: Negated log2 of queue weight (wq) 2517* ``color_y``: Packet color (yellow) 2518* ``min_th_y``: Minimum queue threshold for packet with yellow color 2519* ``max_th_y``: Minimum queue threshold for packet with yellow color 2520* ``maxp_inv_y``: Inverse of packet marking probability maximum value (maxp) 2521* ``wq_log2_y``: Negated log2 of queue weight (wq) 2522* ``color_r``: Packet color (red) 2523* ``min_th_r``: Minimum queue threshold for packet with yellow color 2524* ``max_th_r``: Minimum queue threshold for packet with yellow color 2525* ``maxp_inv_r``: Inverse of packet marking probability maximum value (maxp) 2526* ``wq_log2_r``: Negated log2 of queue weight (wq) 2527 2528Delete port traffic management WRED profile 2529~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ 2530 2531Delete the WRED profile:: 2532 2533 testpmd> del port tm node wred profile (port_id) (wred_profile_id) 2534 2535Add port traffic management hierarchy nonleaf node 2536~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ 2537 2538Add nonleaf node to port traffic management hiearchy:: 2539 2540 testpmd> add port tm nonleaf node (port_id) (node_id) (parent_node_id) \ 2541 (priority) (weight) (level_id) (shaper_profile_id) \ 2542 (n_sp_priorities) (stats_mask) (n_shared_shapers) \ 2543 [(shared_shaper_0) (shared_shaper_1) ...] \ 2544 2545where: 2546 2547* ``parent_node_id``: Node ID of the parent. 2548* ``priority``: Node priority (highest node priority is zero). This is used by 2549 the SP algorithm running on the parent node for scheduling this node. 2550* ``weight``: Node weight (lowest weight is one). The node weight is relative 2551 to the weight sum of all siblings that have the same priority. It is used by 2552 the WFQ algorithm running on the parent node for scheduling this node. 2553* ``level_id``: Hiearchy level of the node. 2554* ``shaper_profile_id``: Shaper profile ID of the private shaper to be used by 2555 the node. 2556* ``n_sp_priorities``: Number of strict priorities. 2557* ``stats_mask``: Mask of statistics counter types to be enabled for this node. 2558* ``n_shared_shapers``: Number of shared shapers. 2559* ``shared_shaper_id``: Shared shaper id. 2560 2561Add port traffic management hierarchy leaf node 2562~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ 2563 2564Add leaf node to port traffic management hiearchy:: 2565 2566 testpmd> add port tm leaf node (port_id) (node_id) (parent_node_id) \ 2567 (priority) (weight) (level_id) (shaper_profile_id) \ 2568 (cman_mode) (wred_profile_id) (stats_mask) (n_shared_shapers) \ 2569 [(shared_shaper_id) (shared_shaper_id) ...] \ 2570 2571where: 2572 2573* ``parent_node_id``: Node ID of the parent. 2574* ``priority``: Node priority (highest node priority is zero). This is used by 2575 the SP algorithm running on the parent node for scheduling this node. 2576* ``weight``: Node weight (lowest weight is one). The node weight is relative 2577 to the weight sum of all siblings that have the same priority. It is used by 2578 the WFQ algorithm running on the parent node for scheduling this node. 2579* ``level_id``: Hiearchy level of the node. 2580* ``shaper_profile_id``: Shaper profile ID of the private shaper to be used by 2581 the node. 2582* ``cman_mode``: Congestion management mode to be enabled for this node. 2583* ``wred_profile_id``: WRED profile id to be enabled for this node. 2584* ``stats_mask``: Mask of statistics counter types to be enabled for this node. 2585* ``n_shared_shapers``: Number of shared shapers. 2586* ``shared_shaper_id``: Shared shaper id. 2587 2588Delete port traffic management hierarchy node 2589~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ 2590 2591Delete node from port traffic management hiearchy:: 2592 2593 testpmd> del port tm node (port_id) (node_id) 2594 2595Update port traffic management hierarchy parent node 2596~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ 2597 2598Update port traffic management hierarchy parent node:: 2599 2600 testpmd> set port tm node parent (port_id) (node_id) (parent_node_id) \ 2601 (priority) (weight) 2602 2603This function can only be called after the hierarchy commit invocation. Its 2604success depends on the port support for this operation, as advertised through 2605the port capability set. This function is valid for all nodes of the traffic 2606management hierarchy except root node. 2607 2608Suspend port traffic management hierarchy node 2609~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ 2610 2611 testpmd> suspend port tm node (port_id) (node_id) 2612 2613Resume port traffic management hierarchy node 2614~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ 2615 2616 testpmd> resume port tm node (port_id) (node_id) 2617 2618Commit port traffic management hierarchy 2619~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ 2620 2621Commit the traffic management hierarchy on the port:: 2622 2623 testpmd> port tm hierarchy commit (port_id) (clean_on_fail) 2624 2625where: 2626 2627* ``clean_on_fail``: When set to non-zero, hierarchy is cleared on function 2628 call failure. On the other hand, hierarchy is preserved when this parameter 2629 is equal to zero. 2630 2631Set port traffic management default hierarchy (tm forwarding mode) 2632~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ 2633 2634set the traffic management default hierarchy on the port:: 2635 2636 testpmd> set port tm hierarchy default (port_id) 2637 2638Filter Functions 2639---------------- 2640 2641This section details the available filter functions that are available. 2642 2643Note these functions interface the deprecated legacy filtering framework, 2644superseded by *rte_flow*. See `Flow rules management`_. 2645 2646ethertype_filter 2647~~~~~~~~~~~~~~~~~~~~ 2648 2649Add or delete a L2 Ethertype filter, which identify packets by their L2 Ethertype mainly assign them to a receive queue:: 2650 2651 ethertype_filter (port_id) (add|del) (mac_addr|mac_ignr) (mac_address) \ 2652 ethertype (ether_type) (drop|fwd) queue (queue_id) 2653 2654The available information parameters are: 2655 2656* ``port_id``: The port which the Ethertype filter assigned on. 2657 2658* ``mac_addr``: Compare destination mac address. 2659 2660* ``mac_ignr``: Ignore destination mac address match. 2661 2662* ``mac_address``: Destination mac address to match. 2663 2664* ``ether_type``: The EtherType value want to match, 2665 for example 0x0806 for ARP packet. 0x0800 (IPv4) and 0x86DD (IPv6) are invalid. 2666 2667* ``queue_id``: The receive queue associated with this EtherType filter. 2668 It is meaningless when deleting or dropping. 2669 2670Example, to add/remove an ethertype filter rule:: 2671 2672 testpmd> ethertype_filter 0 add mac_ignr 00:11:22:33:44:55 \ 2673 ethertype 0x0806 fwd queue 3 2674 2675 testpmd> ethertype_filter 0 del mac_ignr 00:11:22:33:44:55 \ 2676 ethertype 0x0806 fwd queue 3 2677 26782tuple_filter 2679~~~~~~~~~~~~~~~~~ 2680 2681Add or delete a 2-tuple filter, 2682which identifies packets by specific protocol and destination TCP/UDP port 2683and forwards packets into one of the receive queues:: 2684 2685 2tuple_filter (port_id) (add|del) dst_port (dst_port_value) \ 2686 protocol (protocol_value) mask (mask_value) \ 2687 tcp_flags (tcp_flags_value) priority (prio_value) \ 2688 queue (queue_id) 2689 2690The available information parameters are: 2691 2692* ``port_id``: The port which the 2-tuple filter assigned on. 2693 2694* ``dst_port_value``: Destination port in L4. 2695 2696* ``protocol_value``: IP L4 protocol. 2697 2698* ``mask_value``: Participates in the match or not by bit for field above, 1b means participate. 2699 2700* ``tcp_flags_value``: TCP control bits. The non-zero value is invalid, when the pro_value is not set to 0x06 (TCP). 2701 2702* ``prio_value``: Priority of this filter. 2703 2704* ``queue_id``: The receive queue associated with this 2-tuple filter. 2705 2706Example, to add/remove an 2tuple filter rule:: 2707 2708 testpmd> 2tuple_filter 0 add dst_port 32 protocol 0x06 mask 0x03 \ 2709 tcp_flags 0x02 priority 3 queue 3 2710 2711 testpmd> 2tuple_filter 0 del dst_port 32 protocol 0x06 mask 0x03 \ 2712 tcp_flags 0x02 priority 3 queue 3 2713 27145tuple_filter 2715~~~~~~~~~~~~~~~~~ 2716 2717Add or delete a 5-tuple filter, 2718which consists of a 5-tuple (protocol, source and destination IP addresses, source and destination TCP/UDP/SCTP port) 2719and routes packets into one of the receive queues:: 2720 2721 5tuple_filter (port_id) (add|del) dst_ip (dst_address) src_ip \ 2722 (src_address) dst_port (dst_port_value) \ 2723 src_port (src_port_value) protocol (protocol_value) \ 2724 mask (mask_value) tcp_flags (tcp_flags_value) \ 2725 priority (prio_value) queue (queue_id) 2726 2727The available information parameters are: 2728 2729* ``port_id``: The port which the 5-tuple filter assigned on. 2730 2731* ``dst_address``: Destination IP address. 2732 2733* ``src_address``: Source IP address. 2734 2735* ``dst_port_value``: TCP/UDP destination port. 2736 2737* ``src_port_value``: TCP/UDP source port. 2738 2739* ``protocol_value``: L4 protocol. 2740 2741* ``mask_value``: Participates in the match or not by bit for field above, 1b means participate 2742 2743* ``tcp_flags_value``: TCP control bits. The non-zero value is invalid, when the protocol_value is not set to 0x06 (TCP). 2744 2745* ``prio_value``: The priority of this filter. 2746 2747* ``queue_id``: The receive queue associated with this 5-tuple filter. 2748 2749Example, to add/remove an 5tuple filter rule:: 2750 2751 testpmd> 5tuple_filter 0 add dst_ip 2.2.2.5 src_ip 2.2.2.4 \ 2752 dst_port 64 src_port 32 protocol 0x06 mask 0x1F \ 2753 flags 0x0 priority 3 queue 3 2754 2755 testpmd> 5tuple_filter 0 del dst_ip 2.2.2.5 src_ip 2.2.2.4 \ 2756 dst_port 64 src_port 32 protocol 0x06 mask 0x1F \ 2757 flags 0x0 priority 3 queue 3 2758 2759syn_filter 2760~~~~~~~~~~ 2761 2762Using the SYN filter, TCP packets whose *SYN* flag is set can be forwarded to a separate queue:: 2763 2764 syn_filter (port_id) (add|del) priority (high|low) queue (queue_id) 2765 2766The available information parameters are: 2767 2768* ``port_id``: The port which the SYN filter assigned on. 2769 2770* ``high``: This SYN filter has higher priority than other filters. 2771 2772* ``low``: This SYN filter has lower priority than other filters. 2773 2774* ``queue_id``: The receive queue associated with this SYN filter 2775 2776Example:: 2777 2778 testpmd> syn_filter 0 add priority high queue 3 2779 2780flex_filter 2781~~~~~~~~~~~ 2782 2783With flex filter, packets can be recognized by any arbitrary pattern within the first 128 bytes of the packet 2784and routed into one of the receive queues:: 2785 2786 flex_filter (port_id) (add|del) len (len_value) bytes (bytes_value) \ 2787 mask (mask_value) priority (prio_value) queue (queue_id) 2788 2789The available information parameters are: 2790 2791* ``port_id``: The port which the Flex filter is assigned on. 2792 2793* ``len_value``: Filter length in bytes, no greater than 128. 2794 2795* ``bytes_value``: A string in hexadecimal, means the value the flex filter needs to match. 2796 2797* ``mask_value``: A string in hexadecimal, bit 1 means corresponding byte participates in the match. 2798 2799* ``prio_value``: The priority of this filter. 2800 2801* ``queue_id``: The receive queue associated with this Flex filter. 2802 2803Example:: 2804 2805 testpmd> flex_filter 0 add len 16 bytes 0x00000000000000000000000008060000 \ 2806 mask 000C priority 3 queue 3 2807 2808 testpmd> flex_filter 0 del len 16 bytes 0x00000000000000000000000008060000 \ 2809 mask 000C priority 3 queue 3 2810 2811 2812.. _testpmd_flow_director: 2813 2814flow_director_filter 2815~~~~~~~~~~~~~~~~~~~~ 2816 2817The Flow Director works in receive mode to identify specific flows or sets of flows and route them to specific queues. 2818 2819Four types of filtering are supported which are referred to as Perfect Match, Signature, Perfect-mac-vlan and 2820Perfect-tunnel filters, the match mode is set by the ``--pkt-filter-mode`` command-line parameter: 2821 2822* Perfect match filters. 2823 The hardware checks a match between the masked fields of the received packets and the programmed filters. 2824 The masked fields are for IP flow. 2825 2826* Signature filters. 2827 The hardware checks a match between a hash-based signature of the masked fields of the received packet. 2828 2829* Perfect-mac-vlan match filters. 2830 The hardware checks a match between the masked fields of the received packets and the programmed filters. 2831 The masked fields are for MAC VLAN flow. 2832 2833* Perfect-tunnel match filters. 2834 The hardware checks a match between the masked fields of the received packets and the programmed filters. 2835 The masked fields are for tunnel flow. 2836 2837* Perfect-raw-flow-type match filters. 2838 The hardware checks a match between the masked fields of the received packets and pre-loaded raw (template) packet. 2839 The masked fields are specified by input sets. 2840 2841The Flow Director filters can match the different fields for different type of packet: flow type, specific input set 2842per flow type and the flexible payload. 2843 2844The Flow Director can also mask out parts of all of these fields so that filters 2845are only applied to certain fields or parts of the fields. 2846 2847Note that for raw flow type mode the source and destination fields in the 2848raw packet buffer need to be presented in a reversed order with respect 2849to the expected received packets. 2850For example: IP source and destination addresses or TCP/UDP/SCTP 2851source and destination ports 2852 2853Different NICs may have different capabilities, command show port fdir (port_id) can be used to acquire the information. 2854 2855# Commands to add flow director filters of different flow types:: 2856 2857 flow_director_filter (port_id) mode IP (add|del|update) \ 2858 flow (ipv4-other|ipv4-frag|ipv6-other|ipv6-frag) \ 2859 src (src_ip_address) dst (dst_ip_address) \ 2860 tos (tos_value) proto (proto_value) ttl (ttl_value) \ 2861 vlan (vlan_value) flexbytes (flexbytes_value) \ 2862 (drop|fwd) pf|vf(vf_id) queue (queue_id) \ 2863 fd_id (fd_id_value) 2864 2865 flow_director_filter (port_id) mode IP (add|del|update) \ 2866 flow (ipv4-tcp|ipv4-udp|ipv6-tcp|ipv6-udp) \ 2867 src (src_ip_address) (src_port) \ 2868 dst (dst_ip_address) (dst_port) \ 2869 tos (tos_value) ttl (ttl_value) \ 2870 vlan (vlan_value) flexbytes (flexbytes_value) \ 2871 (drop|fwd) queue pf|vf(vf_id) (queue_id) \ 2872 fd_id (fd_id_value) 2873 2874 flow_director_filter (port_id) mode IP (add|del|update) \ 2875 flow (ipv4-sctp|ipv6-sctp) \ 2876 src (src_ip_address) (src_port) \ 2877 dst (dst_ip_address) (dst_port) \ 2878 tos (tos_value) ttl (ttl_value) \ 2879 tag (verification_tag) vlan (vlan_value) \ 2880 flexbytes (flexbytes_value) (drop|fwd) \ 2881 pf|vf(vf_id) queue (queue_id) fd_id (fd_id_value) 2882 2883 flow_director_filter (port_id) mode IP (add|del|update) flow l2_payload \ 2884 ether (ethertype) flexbytes (flexbytes_value) \ 2885 (drop|fwd) pf|vf(vf_id) queue (queue_id) 2886 fd_id (fd_id_value) 2887 2888 flow_director_filter (port_id) mode MAC-VLAN (add|del|update) \ 2889 mac (mac_address) vlan (vlan_value) \ 2890 flexbytes (flexbytes_value) (drop|fwd) \ 2891 queue (queue_id) fd_id (fd_id_value) 2892 2893 flow_director_filter (port_id) mode Tunnel (add|del|update) \ 2894 mac (mac_address) vlan (vlan_value) \ 2895 tunnel (NVGRE|VxLAN) tunnel-id (tunnel_id_value) \ 2896 flexbytes (flexbytes_value) (drop|fwd) \ 2897 queue (queue_id) fd_id (fd_id_value) 2898 2899 flow_director_filter (port_id) mode raw (add|del|update) flow (flow_id) \ 2900 (drop|fwd) queue (queue_id) fd_id (fd_id_value) \ 2901 packet (packet file name) 2902 2903For example, to add an ipv4-udp flow type filter:: 2904 2905 testpmd> flow_director_filter 0 mode IP add flow ipv4-udp src 2.2.2.3 32 \ 2906 dst 2.2.2.5 33 tos 2 ttl 40 vlan 0x1 flexbytes (0x88,0x48) \ 2907 fwd pf queue 1 fd_id 1 2908 2909For example, add an ipv4-other flow type filter:: 2910 2911 testpmd> flow_director_filter 0 mode IP add flow ipv4-other src 2.2.2.3 \ 2912 dst 2.2.2.5 tos 2 proto 20 ttl 40 vlan 0x1 \ 2913 flexbytes (0x88,0x48) fwd pf queue 1 fd_id 1 2914 2915flush_flow_director 2916~~~~~~~~~~~~~~~~~~~ 2917 2918Flush all flow director filters on a device:: 2919 2920 testpmd> flush_flow_director (port_id) 2921 2922Example, to flush all flow director filter on port 0:: 2923 2924 testpmd> flush_flow_director 0 2925 2926flow_director_mask 2927~~~~~~~~~~~~~~~~~~ 2928 2929Set flow director's input masks:: 2930 2931 flow_director_mask (port_id) mode IP vlan (vlan_value) \ 2932 src_mask (ipv4_src) (ipv6_src) (src_port) \ 2933 dst_mask (ipv4_dst) (ipv6_dst) (dst_port) 2934 2935 flow_director_mask (port_id) mode MAC-VLAN vlan (vlan_value) 2936 2937 flow_director_mask (port_id) mode Tunnel vlan (vlan_value) \ 2938 mac (mac_value) tunnel-type (tunnel_type_value) \ 2939 tunnel-id (tunnel_id_value) 2940 2941Example, to set flow director mask on port 0:: 2942 2943 testpmd> flow_director_mask 0 mode IP vlan 0xefff \ 2944 src_mask 255.255.255.255 \ 2945 FFFF:FFFF:FFFF:FFFF:FFFF:FFFF:FFFF:FFFF 0xFFFF \ 2946 dst_mask 255.255.255.255 \ 2947 FFFF:FFFF:FFFF:FFFF:FFFF:FFFF:FFFF:FFFF 0xFFFF 2948 2949flow_director_flex_mask 2950~~~~~~~~~~~~~~~~~~~~~~~ 2951 2952set masks of flow director's flexible payload based on certain flow type:: 2953 2954 testpmd> flow_director_flex_mask (port_id) \ 2955 flow (none|ipv4-other|ipv4-frag|ipv4-tcp|ipv4-udp|ipv4-sctp| \ 2956 ipv6-other|ipv6-frag|ipv6-tcp|ipv6-udp|ipv6-sctp| \ 2957 l2_payload|all) (mask) 2958 2959Example, to set flow director's flex mask for all flow type on port 0:: 2960 2961 testpmd> flow_director_flex_mask 0 flow all \ 2962 (0xff,0xff,0,0,0,0,0,0,0,0,0,0,0,0,0,0) 2963 2964 2965flow_director_flex_payload 2966~~~~~~~~~~~~~~~~~~~~~~~~~~ 2967 2968Configure flexible payload selection:: 2969 2970 flow_director_flex_payload (port_id) (raw|l2|l3|l4) (config) 2971 2972For example, to select the first 16 bytes from the offset 4 (bytes) of packet's payload as flexible payload:: 2973 2974 testpmd> flow_director_flex_payload 0 l4 \ 2975 (4,5,6,7,8,9,10,11,12,13,14,15,16,17,18,19) 2976 2977get_sym_hash_ena_per_port 2978~~~~~~~~~~~~~~~~~~~~~~~~~ 2979 2980Get symmetric hash enable configuration per port:: 2981 2982 get_sym_hash_ena_per_port (port_id) 2983 2984For example, to get symmetric hash enable configuration of port 1:: 2985 2986 testpmd> get_sym_hash_ena_per_port 1 2987 2988set_sym_hash_ena_per_port 2989~~~~~~~~~~~~~~~~~~~~~~~~~ 2990 2991Set symmetric hash enable configuration per port to enable or disable:: 2992 2993 set_sym_hash_ena_per_port (port_id) (enable|disable) 2994 2995For example, to set symmetric hash enable configuration of port 1 to enable:: 2996 2997 testpmd> set_sym_hash_ena_per_port 1 enable 2998 2999get_hash_global_config 3000~~~~~~~~~~~~~~~~~~~~~~ 3001 3002Get the global configurations of hash filters:: 3003 3004 get_hash_global_config (port_id) 3005 3006For example, to get the global configurations of hash filters of port 1:: 3007 3008 testpmd> get_hash_global_config 1 3009 3010set_hash_global_config 3011~~~~~~~~~~~~~~~~~~~~~~ 3012 3013Set the global configurations of hash filters:: 3014 3015 set_hash_global_config (port_id) (toeplitz|simple_xor|default) \ 3016 (ipv4|ipv4-frag|ipv4-tcp|ipv4-udp|ipv4-sctp|ipv4-other|ipv6|ipv6-frag| \ 3017 ipv6-tcp|ipv6-udp|ipv6-sctp|ipv6-other|l2_payload|<flow_id>) \ 3018 (enable|disable) 3019 3020For example, to enable simple_xor for flow type of ipv6 on port 2:: 3021 3022 testpmd> set_hash_global_config 2 simple_xor ipv6 enable 3023 3024set_hash_input_set 3025~~~~~~~~~~~~~~~~~~ 3026 3027Set the input set for hash:: 3028 3029 set_hash_input_set (port_id) (ipv4-frag|ipv4-tcp|ipv4-udp|ipv4-sctp| \ 3030 ipv4-other|ipv6-frag|ipv6-tcp|ipv6-udp|ipv6-sctp|ipv6-other| \ 3031 l2_payload|<flow_id>) (ovlan|ivlan|src-ipv4|dst-ipv4|src-ipv6|dst-ipv6| \ 3032 ipv4-tos|ipv4-proto|ipv6-tc|ipv6-next-header|udp-src-port|udp-dst-port| \ 3033 tcp-src-port|tcp-dst-port|sctp-src-port|sctp-dst-port|sctp-veri-tag| \ 3034 udp-key|gre-key|fld-1st|fld-2nd|fld-3rd|fld-4th|fld-5th|fld-6th|fld-7th| \ 3035 fld-8th|none) (select|add) 3036 3037For example, to add source IP to hash input set for flow type of ipv4-udp on port 0:: 3038 3039 testpmd> set_hash_input_set 0 ipv4-udp src-ipv4 add 3040 3041set_fdir_input_set 3042~~~~~~~~~~~~~~~~~~ 3043 3044The Flow Director filters can match the different fields for different type of packet, i.e. specific input set 3045on per flow type and the flexible payload. This command can be used to change input set for each flow type. 3046 3047Set the input set for flow director:: 3048 3049 set_fdir_input_set (port_id) (ipv4-frag|ipv4-tcp|ipv4-udp|ipv4-sctp| \ 3050 ipv4-other|ipv6|ipv6-frag|ipv6-tcp|ipv6-udp|ipv6-sctp|ipv6-other| \ 3051 l2_payload|<flow_id>) (ivlan|ethertype|src-ipv4|dst-ipv4|src-ipv6|dst-ipv6| \ 3052 ipv4-tos|ipv4-proto|ipv4-ttl|ipv6-tc|ipv6-next-header|ipv6-hop-limits| \ 3053 tudp-src-port|udp-dst-port|cp-src-port|tcp-dst-port|sctp-src-port| \ 3054 sctp-dst-port|sctp-veri-tag|none) (select|add) 3055 3056For example to add source IP to FD input set for flow type of ipv4-udp on port 0:: 3057 3058 testpmd> set_fdir_input_set 0 ipv4-udp src-ipv4 add 3059 3060global_config 3061~~~~~~~~~~~~~ 3062 3063Set different GRE key length for input set:: 3064 3065 global_config (port_id) gre-key-len (number in bytes) 3066 3067For example to set GRE key length for input set to 4 bytes on port 0:: 3068 3069 testpmd> global_config 0 gre-key-len 4 3070 3071 3072.. _testpmd_rte_flow: 3073 3074Flow rules management 3075--------------------- 3076 3077Control of the generic flow API (*rte_flow*) is fully exposed through the 3078``flow`` command (validation, creation, destruction, queries and operation 3079modes). 3080 3081Considering *rte_flow* overlaps with all `Filter Functions`_, using both 3082features simultaneously may cause undefined side-effects and is therefore 3083not recommended. 3084 3085``flow`` syntax 3086~~~~~~~~~~~~~~~ 3087 3088Because the ``flow`` command uses dynamic tokens to handle the large number 3089of possible flow rules combinations, its behavior differs slightly from 3090other commands, in particular: 3091 3092- Pressing *?* or the *<tab>* key displays contextual help for the current 3093 token, not that of the entire command. 3094 3095- Optional and repeated parameters are supported (provided they are listed 3096 in the contextual help). 3097 3098The first parameter stands for the operation mode. Possible operations and 3099their general syntax are described below. They are covered in detail in the 3100following sections. 3101 3102- Check whether a flow rule can be created:: 3103 3104 flow validate {port_id} 3105 [group {group_id}] [priority {level}] [ingress] [egress] [transfer] 3106 pattern {item} [/ {item} [...]] / end 3107 actions {action} [/ {action} [...]] / end 3108 3109- Create a flow rule:: 3110 3111 flow create {port_id} 3112 [group {group_id}] [priority {level}] [ingress] [egress] [transfer] 3113 pattern {item} [/ {item} [...]] / end 3114 actions {action} [/ {action} [...]] / end 3115 3116- Destroy specific flow rules:: 3117 3118 flow destroy {port_id} rule {rule_id} [...] 3119 3120- Destroy all flow rules:: 3121 3122 flow flush {port_id} 3123 3124- Query an existing flow rule:: 3125 3126 flow query {port_id} {rule_id} {action} 3127 3128- List existing flow rules sorted by priority, filtered by group 3129 identifiers:: 3130 3131 flow list {port_id} [group {group_id}] [...] 3132 3133- Restrict ingress traffic to the defined flow rules:: 3134 3135 flow isolate {port_id} {boolean} 3136 3137Validating flow rules 3138~~~~~~~~~~~~~~~~~~~~~ 3139 3140``flow validate`` reports whether a flow rule would be accepted by the 3141underlying device in its current state but stops short of creating it. It is 3142bound to ``rte_flow_validate()``:: 3143 3144 flow validate {port_id} 3145 [group {group_id}] [priority {level}] [ingress] [egress] [transfer] 3146 pattern {item} [/ {item} [...]] / end 3147 actions {action} [/ {action} [...]] / end 3148 3149If successful, it will show:: 3150 3151 Flow rule validated 3152 3153Otherwise it will show an error message of the form:: 3154 3155 Caught error type [...] ([...]): [...] 3156 3157This command uses the same parameters as ``flow create``, their format is 3158described in `Creating flow rules`_. 3159 3160Check whether redirecting any Ethernet packet received on port 0 to RX queue 3161index 6 is supported:: 3162 3163 testpmd> flow validate 0 ingress pattern eth / end 3164 actions queue index 6 / end 3165 Flow rule validated 3166 testpmd> 3167 3168Port 0 does not support TCPv6 rules:: 3169 3170 testpmd> flow validate 0 ingress pattern eth / ipv6 / tcp / end 3171 actions drop / end 3172 Caught error type 9 (specific pattern item): Invalid argument 3173 testpmd> 3174 3175Creating flow rules 3176~~~~~~~~~~~~~~~~~~~ 3177 3178``flow create`` validates and creates the specified flow rule. It is bound 3179to ``rte_flow_create()``:: 3180 3181 flow create {port_id} 3182 [group {group_id}] [priority {level}] [ingress] [egress] [transfer] 3183 pattern {item} [/ {item} [...]] / end 3184 actions {action} [/ {action} [...]] / end 3185 3186If successful, it will return a flow rule ID usable with other commands:: 3187 3188 Flow rule #[...] created 3189 3190Otherwise it will show an error message of the form:: 3191 3192 Caught error type [...] ([...]): [...] 3193 3194Parameters describe in the following order: 3195 3196- Attributes (*group*, *priority*, *ingress*, *egress*, *transfer* tokens). 3197- A matching pattern, starting with the *pattern* token and terminated by an 3198 *end* pattern item. 3199- Actions, starting with the *actions* token and terminated by an *end* 3200 action. 3201 3202These translate directly to *rte_flow* objects provided as-is to the 3203underlying functions. 3204 3205The shortest valid definition only comprises mandatory tokens:: 3206 3207 testpmd> flow create 0 pattern end actions end 3208 3209Note that PMDs may refuse rules that essentially do nothing such as this 3210one. 3211 3212**All unspecified object values are automatically initialized to 0.** 3213 3214Attributes 3215^^^^^^^^^^ 3216 3217These tokens affect flow rule attributes (``struct rte_flow_attr``) and are 3218specified before the ``pattern`` token. 3219 3220- ``group {group id}``: priority group. 3221- ``priority {level}``: priority level within group. 3222- ``ingress``: rule applies to ingress traffic. 3223- ``egress``: rule applies to egress traffic. 3224- ``transfer``: apply rule directly to endpoints found in pattern. 3225 3226Each instance of an attribute specified several times overrides the previous 3227value as shown below (group 4 is used):: 3228 3229 testpmd> flow create 0 group 42 group 24 group 4 [...] 3230 3231Note that once enabled, ``ingress`` and ``egress`` cannot be disabled. 3232 3233While not specifying a direction is an error, some rules may allow both 3234simultaneously. 3235 3236Most rules affect RX therefore contain the ``ingress`` token:: 3237 3238 testpmd> flow create 0 ingress pattern [...] 3239 3240Matching pattern 3241^^^^^^^^^^^^^^^^ 3242 3243A matching pattern starts after the ``pattern`` token. It is made of pattern 3244items and is terminated by a mandatory ``end`` item. 3245 3246Items are named after their type (*RTE_FLOW_ITEM_TYPE_* from ``enum 3247rte_flow_item_type``). 3248 3249The ``/`` token is used as a separator between pattern items as shown 3250below:: 3251 3252 testpmd> flow create 0 ingress pattern eth / ipv4 / udp / end [...] 3253 3254Note that protocol items like these must be stacked from lowest to highest 3255layer to make sense. For instance, the following rule is either invalid or 3256unlikely to match any packet:: 3257 3258 testpmd> flow create 0 ingress pattern eth / udp / ipv4 / end [...] 3259 3260More information on these restrictions can be found in the *rte_flow* 3261documentation. 3262 3263Several items support additional specification structures, for example 3264``ipv4`` allows specifying source and destination addresses as follows:: 3265 3266 testpmd> flow create 0 ingress pattern eth / ipv4 src is 10.1.1.1 3267 dst is 10.2.0.0 / end [...] 3268 3269This rule matches all IPv4 traffic with the specified properties. 3270 3271In this example, ``src`` and ``dst`` are field names of the underlying 3272``struct rte_flow_item_ipv4`` object. All item properties can be specified 3273in a similar fashion. 3274 3275The ``is`` token means that the subsequent value must be matched exactly, 3276and assigns ``spec`` and ``mask`` fields in ``struct rte_flow_item`` 3277accordingly. Possible assignment tokens are: 3278 3279- ``is``: match value perfectly (with full bit-mask). 3280- ``spec``: match value according to configured bit-mask. 3281- ``last``: specify upper bound to establish a range. 3282- ``mask``: specify bit-mask with relevant bits set to one. 3283- ``prefix``: generate bit-mask from a prefix length. 3284 3285These yield identical results:: 3286 3287 ipv4 src is 10.1.1.1 3288 3289:: 3290 3291 ipv4 src spec 10.1.1.1 src mask 255.255.255.255 3292 3293:: 3294 3295 ipv4 src spec 10.1.1.1 src prefix 32 3296 3297:: 3298 3299 ipv4 src is 10.1.1.1 src last 10.1.1.1 # range with a single value 3300 3301:: 3302 3303 ipv4 src is 10.1.1.1 src last 0 # 0 disables range 3304 3305Inclusive ranges can be defined with ``last``:: 3306 3307 ipv4 src is 10.1.1.1 src last 10.2.3.4 # 10.1.1.1 to 10.2.3.4 3308 3309Note that ``mask`` affects both ``spec`` and ``last``:: 3310 3311 ipv4 src is 10.1.1.1 src last 10.2.3.4 src mask 255.255.0.0 3312 # matches 10.1.0.0 to 10.2.255.255 3313 3314Properties can be modified multiple times:: 3315 3316 ipv4 src is 10.1.1.1 src is 10.1.2.3 src is 10.2.3.4 # matches 10.2.3.4 3317 3318:: 3319 3320 ipv4 src is 10.1.1.1 src prefix 24 src prefix 16 # matches 10.1.0.0/16 3321 3322Pattern items 3323^^^^^^^^^^^^^ 3324 3325This section lists supported pattern items and their attributes, if any. 3326 3327- ``end``: end list of pattern items. 3328 3329- ``void``: no-op pattern item. 3330 3331- ``invert``: perform actions when pattern does not match. 3332 3333- ``any``: match any protocol for the current layer. 3334 3335 - ``num {unsigned}``: number of layers covered. 3336 3337- ``pf``: match traffic from/to the physical function. 3338 3339- ``vf``: match traffic from/to a virtual function ID. 3340 3341 - ``id {unsigned}``: VF ID. 3342 3343- ``phy_port``: match traffic from/to a specific physical port. 3344 3345 - ``index {unsigned}``: physical port index. 3346 3347- ``port_id``: match traffic from/to a given DPDK port ID. 3348 3349 - ``id {unsigned}``: DPDK port ID. 3350 3351- ``mark``: match value set in previously matched flow rule using the mark action. 3352 3353 - ``id {unsigned}``: arbitrary integer value. 3354 3355- ``raw``: match an arbitrary byte string. 3356 3357 - ``relative {boolean}``: look for pattern after the previous item. 3358 - ``search {boolean}``: search pattern from offset (see also limit). 3359 - ``offset {integer}``: absolute or relative offset for pattern. 3360 - ``limit {unsigned}``: search area limit for start of pattern. 3361 - ``pattern {string}``: byte string to look for. 3362 3363- ``eth``: match Ethernet header. 3364 3365 - ``dst {MAC-48}``: destination MAC. 3366 - ``src {MAC-48}``: source MAC. 3367 - ``type {unsigned}``: EtherType or TPID. 3368 3369- ``vlan``: match 802.1Q/ad VLAN tag. 3370 3371 - ``tci {unsigned}``: tag control information. 3372 - ``pcp {unsigned}``: priority code point. 3373 - ``dei {unsigned}``: drop eligible indicator. 3374 - ``vid {unsigned}``: VLAN identifier. 3375 - ``inner_type {unsigned}``: inner EtherType or TPID. 3376 3377- ``ipv4``: match IPv4 header. 3378 3379 - ``tos {unsigned}``: type of service. 3380 - ``ttl {unsigned}``: time to live. 3381 - ``proto {unsigned}``: next protocol ID. 3382 - ``src {ipv4 address}``: source address. 3383 - ``dst {ipv4 address}``: destination address. 3384 3385- ``ipv6``: match IPv6 header. 3386 3387 - ``tc {unsigned}``: traffic class. 3388 - ``flow {unsigned}``: flow label. 3389 - ``proto {unsigned}``: protocol (next header). 3390 - ``hop {unsigned}``: hop limit. 3391 - ``src {ipv6 address}``: source address. 3392 - ``dst {ipv6 address}``: destination address. 3393 3394- ``icmp``: match ICMP header. 3395 3396 - ``type {unsigned}``: ICMP packet type. 3397 - ``code {unsigned}``: ICMP packet code. 3398 3399- ``udp``: match UDP header. 3400 3401 - ``src {unsigned}``: UDP source port. 3402 - ``dst {unsigned}``: UDP destination port. 3403 3404- ``tcp``: match TCP header. 3405 3406 - ``src {unsigned}``: TCP source port. 3407 - ``dst {unsigned}``: TCP destination port. 3408 3409- ``sctp``: match SCTP header. 3410 3411 - ``src {unsigned}``: SCTP source port. 3412 - ``dst {unsigned}``: SCTP destination port. 3413 - ``tag {unsigned}``: validation tag. 3414 - ``cksum {unsigned}``: checksum. 3415 3416- ``vxlan``: match VXLAN header. 3417 3418 - ``vni {unsigned}``: VXLAN identifier. 3419 3420- ``e_tag``: match IEEE 802.1BR E-Tag header. 3421 3422 - ``grp_ecid_b {unsigned}``: GRP and E-CID base. 3423 3424- ``nvgre``: match NVGRE header. 3425 3426 - ``tni {unsigned}``: virtual subnet ID. 3427 3428- ``mpls``: match MPLS header. 3429 3430 - ``label {unsigned}``: MPLS label. 3431 3432- ``gre``: match GRE header. 3433 3434 - ``protocol {unsigned}``: protocol type. 3435 3436- ``fuzzy``: fuzzy pattern match, expect faster than default. 3437 3438 - ``thresh {unsigned}``: accuracy threshold. 3439 3440- ``gtp``, ``gtpc``, ``gtpu``: match GTPv1 header. 3441 3442 - ``teid {unsigned}``: tunnel endpoint identifier. 3443 3444- ``geneve``: match GENEVE header. 3445 3446 - ``vni {unsigned}``: virtual network identifier. 3447 - ``protocol {unsigned}``: protocol type. 3448 3449- ``vxlan-gpe``: match VXLAN-GPE header. 3450 3451 - ``vni {unsigned}``: VXLAN-GPE identifier. 3452 3453- ``arp_eth_ipv4``: match ARP header for Ethernet/IPv4. 3454 3455 - ``sha {MAC-48}``: sender hardware address. 3456 - ``spa {ipv4 address}``: sender IPv4 address. 3457 - ``tha {MAC-48}``: target hardware address. 3458 - ``tpa {ipv4 address}``: target IPv4 address. 3459 3460- ``ipv6_ext``: match presence of any IPv6 extension header. 3461 3462 - ``next_hdr {unsigned}``: next header. 3463 3464- ``icmp6``: match any ICMPv6 header. 3465 3466 - ``type {unsigned}``: ICMPv6 type. 3467 - ``code {unsigned}``: ICMPv6 code. 3468 3469- ``icmp6_nd_ns``: match ICMPv6 neighbor discovery solicitation. 3470 3471 - ``target_addr {ipv6 address}``: target address. 3472 3473- ``icmp6_nd_na``: match ICMPv6 neighbor discovery advertisement. 3474 3475 - ``target_addr {ipv6 address}``: target address. 3476 3477- ``icmp6_nd_opt``: match presence of any ICMPv6 neighbor discovery option. 3478 3479 - ``type {unsigned}``: ND option type. 3480 3481- ``icmp6_nd_opt_sla_eth``: match ICMPv6 neighbor discovery source Ethernet 3482 link-layer address option. 3483 3484 - ``sla {MAC-48}``: source Ethernet LLA. 3485 3486- ``icmp6_nd_opt_sla_eth``: match ICMPv6 neighbor discovery target Ethernet 3487 link-layer address option. 3488 3489 - ``tla {MAC-48}``: target Ethernet LLA. 3490 3491Actions list 3492^^^^^^^^^^^^ 3493 3494A list of actions starts after the ``actions`` token in the same fashion as 3495`Matching pattern`_; actions are separated by ``/`` tokens and the list is 3496terminated by a mandatory ``end`` action. 3497 3498Actions are named after their type (*RTE_FLOW_ACTION_TYPE_* from ``enum 3499rte_flow_action_type``). 3500 3501Dropping all incoming UDPv4 packets can be expressed as follows:: 3502 3503 testpmd> flow create 0 ingress pattern eth / ipv4 / udp / end 3504 actions drop / end 3505 3506Several actions have configurable properties which must be specified when 3507there is no valid default value. For example, ``queue`` requires a target 3508queue index. 3509 3510This rule redirects incoming UDPv4 traffic to queue index 6:: 3511 3512 testpmd> flow create 0 ingress pattern eth / ipv4 / udp / end 3513 actions queue index 6 / end 3514 3515While this one could be rejected by PMDs (unspecified queue index):: 3516 3517 testpmd> flow create 0 ingress pattern eth / ipv4 / udp / end 3518 actions queue / end 3519 3520As defined by *rte_flow*, the list is not ordered, all actions of a given 3521rule are performed simultaneously. These are equivalent:: 3522 3523 queue index 6 / void / mark id 42 / end 3524 3525:: 3526 3527 void / mark id 42 / queue index 6 / end 3528 3529All actions in a list should have different types, otherwise only the last 3530action of a given type is taken into account:: 3531 3532 queue index 4 / queue index 5 / queue index 6 / end # will use queue 6 3533 3534:: 3535 3536 drop / drop / drop / end # drop is performed only once 3537 3538:: 3539 3540 mark id 42 / queue index 3 / mark id 24 / end # mark will be 24 3541 3542Considering they are performed simultaneously, opposite and overlapping 3543actions can sometimes be combined when the end result is unambiguous:: 3544 3545 drop / queue index 6 / end # drop has no effect 3546 3547:: 3548 3549 queue index 6 / rss queues 6 7 8 / end # queue has no effect 3550 3551:: 3552 3553 drop / passthru / end # drop has no effect 3554 3555Note that PMDs may still refuse such combinations. 3556 3557Actions 3558^^^^^^^ 3559 3560This section lists supported actions and their attributes, if any. 3561 3562- ``end``: end list of actions. 3563 3564- ``void``: no-op action. 3565 3566- ``passthru``: let subsequent rule process matched packets. 3567 3568- ``jump``: redirect traffic to group on device. 3569 3570 - ``group {unsigned}``: group to redirect to. 3571 3572- ``mark``: attach 32 bit value to packets. 3573 3574 - ``id {unsigned}``: 32 bit value to return with packets. 3575 3576- ``flag``: flag packets. 3577 3578- ``queue``: assign packets to a given queue index. 3579 3580 - ``index {unsigned}``: queue index to use. 3581 3582- ``drop``: drop packets (note: passthru has priority). 3583 3584- ``count``: enable counters for this rule. 3585 3586- ``rss``: spread packets among several queues. 3587 3588 - ``func {hash function}``: RSS hash function to apply, allowed tokens are 3589 the same as `set_hash_global_config`_. 3590 3591 - ``level {unsigned}``: encapsulation level for ``types``. 3592 3593 - ``types [{RSS hash type} [...]] end``: specific RSS hash types, allowed 3594 tokens are the same as `set_hash_input_set`_, except that an empty list 3595 does not disable RSS but instead requests unspecified "best-effort" 3596 settings. 3597 3598 - ``key {string}``: RSS hash key, overrides ``key_len``. 3599 3600 - ``key_len {unsigned}``: RSS hash key length in bytes, can be used in 3601 conjunction with ``key`` to pad or truncate it. 3602 3603 - ``queues [{unsigned} [...]] end``: queue indices to use. 3604 3605- ``pf``: direct traffic to physical function. 3606 3607- ``vf``: direct traffic to a virtual function ID. 3608 3609 - ``original {boolean}``: use original VF ID if possible. 3610 - ``id {unsigned}``: VF ID. 3611 3612- ``phy_port``: direct packets to physical port index. 3613 3614 - ``original {boolean}``: use original port index if possible. 3615 - ``index {unsigned}``: physical port index. 3616 3617- ``port_id``: direct matching traffic to a given DPDK port ID. 3618 3619 - ``original {boolean}``: use original DPDK port ID if possible. 3620 - ``id {unsigned}``: DPDK port ID. 3621 3622- ``of_set_mpls_ttl``: OpenFlow's ``OFPAT_SET_MPLS_TTL``. 3623 3624 - ``mpls_ttl``: MPLS TTL. 3625 3626- ``of_dec_mpls_ttl``: OpenFlow's ``OFPAT_DEC_MPLS_TTL``. 3627 3628- ``of_set_nw_ttl``: OpenFlow's ``OFPAT_SET_NW_TTL``. 3629 3630 - ``nw_ttl``: IP TTL. 3631 3632- ``of_dec_nw_ttl``: OpenFlow's ``OFPAT_DEC_NW_TTL``. 3633 3634- ``of_copy_ttl_out``: OpenFlow's ``OFPAT_COPY_TTL_OUT``. 3635 3636- ``of_copy_ttl_in``: OpenFlow's ``OFPAT_COPY_TTL_IN``. 3637 3638- ``of_pop_vlan``: OpenFlow's ``OFPAT_POP_VLAN``. 3639 3640- ``of_push_vlan``: OpenFlow's ``OFPAT_PUSH_VLAN``. 3641 3642 - ``ethertype``: Ethertype. 3643 3644- ``of_set_vlan_vid``: OpenFlow's ``OFPAT_SET_VLAN_VID``. 3645 3646 - ``vlan_vid``: VLAN id. 3647 3648- ``of_set_vlan_pcp``: OpenFlow's ``OFPAT_SET_VLAN_PCP``. 3649 3650 - ``vlan_pcp``: VLAN priority. 3651 3652- ``of_pop_mpls``: OpenFlow's ``OFPAT_POP_MPLS``. 3653 3654 - ``ethertype``: Ethertype. 3655 3656- ``of_push_mpls``: OpenFlow's ``OFPAT_PUSH_MPLS``. 3657 3658 - ``ethertype``: Ethertype. 3659 3660Destroying flow rules 3661~~~~~~~~~~~~~~~~~~~~~ 3662 3663``flow destroy`` destroys one or more rules from their rule ID (as returned 3664by ``flow create``), this command calls ``rte_flow_destroy()`` as many 3665times as necessary:: 3666 3667 flow destroy {port_id} rule {rule_id} [...] 3668 3669If successful, it will show:: 3670 3671 Flow rule #[...] destroyed 3672 3673It does not report anything for rule IDs that do not exist. The usual error 3674message is shown when a rule cannot be destroyed:: 3675 3676 Caught error type [...] ([...]): [...] 3677 3678``flow flush`` destroys all rules on a device and does not take extra 3679arguments. It is bound to ``rte_flow_flush()``:: 3680 3681 flow flush {port_id} 3682 3683Any errors are reported as above. 3684 3685Creating several rules and destroying them:: 3686 3687 testpmd> flow create 0 ingress pattern eth / ipv6 / end 3688 actions queue index 2 / end 3689 Flow rule #0 created 3690 testpmd> flow create 0 ingress pattern eth / ipv4 / end 3691 actions queue index 3 / end 3692 Flow rule #1 created 3693 testpmd> flow destroy 0 rule 0 rule 1 3694 Flow rule #1 destroyed 3695 Flow rule #0 destroyed 3696 testpmd> 3697 3698The same result can be achieved using ``flow flush``:: 3699 3700 testpmd> flow create 0 ingress pattern eth / ipv6 / end 3701 actions queue index 2 / end 3702 Flow rule #0 created 3703 testpmd> flow create 0 ingress pattern eth / ipv4 / end 3704 actions queue index 3 / end 3705 Flow rule #1 created 3706 testpmd> flow flush 0 3707 testpmd> 3708 3709Non-existent rule IDs are ignored:: 3710 3711 testpmd> flow create 0 ingress pattern eth / ipv6 / end 3712 actions queue index 2 / end 3713 Flow rule #0 created 3714 testpmd> flow create 0 ingress pattern eth / ipv4 / end 3715 actions queue index 3 / end 3716 Flow rule #1 created 3717 testpmd> flow destroy 0 rule 42 rule 10 rule 2 3718 testpmd> 3719 testpmd> flow destroy 0 rule 0 3720 Flow rule #0 destroyed 3721 testpmd> 3722 3723Querying flow rules 3724~~~~~~~~~~~~~~~~~~~ 3725 3726``flow query`` queries a specific action of a flow rule having that 3727ability. Such actions collect information that can be reported using this 3728command. It is bound to ``rte_flow_query()``:: 3729 3730 flow query {port_id} {rule_id} {action} 3731 3732If successful, it will display either the retrieved data for known actions 3733or the following message:: 3734 3735 Cannot display result for action type [...] ([...]) 3736 3737Otherwise, it will complain either that the rule does not exist or that some 3738error occurred:: 3739 3740 Flow rule #[...] not found 3741 3742:: 3743 3744 Caught error type [...] ([...]): [...] 3745 3746Currently only the ``count`` action is supported. This action reports the 3747number of packets that hit the flow rule and the total number of bytes. Its 3748output has the following format:: 3749 3750 count: 3751 hits_set: [...] # whether "hits" contains a valid value 3752 bytes_set: [...] # whether "bytes" contains a valid value 3753 hits: [...] # number of packets 3754 bytes: [...] # number of bytes 3755 3756Querying counters for TCPv6 packets redirected to queue 6:: 3757 3758 testpmd> flow create 0 ingress pattern eth / ipv6 / tcp / end 3759 actions queue index 6 / count / end 3760 Flow rule #4 created 3761 testpmd> flow query 0 4 count 3762 count: 3763 hits_set: 1 3764 bytes_set: 0 3765 hits: 386446 3766 bytes: 0 3767 testpmd> 3768 3769Listing flow rules 3770~~~~~~~~~~~~~~~~~~ 3771 3772``flow list`` lists existing flow rules sorted by priority and optionally 3773filtered by group identifiers:: 3774 3775 flow list {port_id} [group {group_id}] [...] 3776 3777This command only fails with the following message if the device does not 3778exist:: 3779 3780 Invalid port [...] 3781 3782Output consists of a header line followed by a short description of each 3783flow rule, one per line. There is no output at all when no flow rules are 3784configured on the device:: 3785 3786 ID Group Prio Attr Rule 3787 [...] [...] [...] [...] [...] 3788 3789``Attr`` column flags: 3790 3791- ``i`` for ``ingress``. 3792- ``e`` for ``egress``. 3793 3794Creating several flow rules and listing them:: 3795 3796 testpmd> flow create 0 ingress pattern eth / ipv4 / end 3797 actions queue index 6 / end 3798 Flow rule #0 created 3799 testpmd> flow create 0 ingress pattern eth / ipv6 / end 3800 actions queue index 2 / end 3801 Flow rule #1 created 3802 testpmd> flow create 0 priority 5 ingress pattern eth / ipv4 / udp / end 3803 actions rss queues 6 7 8 end / end 3804 Flow rule #2 created 3805 testpmd> flow list 0 3806 ID Group Prio Attr Rule 3807 0 0 0 i- ETH IPV4 => QUEUE 3808 1 0 0 i- ETH IPV6 => QUEUE 3809 2 0 5 i- ETH IPV4 UDP => RSS 3810 testpmd> 3811 3812Rules are sorted by priority (i.e. group ID first, then priority level):: 3813 3814 testpmd> flow list 1 3815 ID Group Prio Attr Rule 3816 0 0 0 i- ETH => COUNT 3817 6 0 500 i- ETH IPV6 TCP => DROP COUNT 3818 5 0 1000 i- ETH IPV6 ICMP => QUEUE 3819 1 24 0 i- ETH IPV4 UDP => QUEUE 3820 4 24 10 i- ETH IPV4 TCP => DROP 3821 3 24 20 i- ETH IPV4 => DROP 3822 2 24 42 i- ETH IPV4 UDP => QUEUE 3823 7 63 0 i- ETH IPV6 UDP VXLAN => MARK QUEUE 3824 testpmd> 3825 3826Output can be limited to specific groups:: 3827 3828 testpmd> flow list 1 group 0 group 63 3829 ID Group Prio Attr Rule 3830 0 0 0 i- ETH => COUNT 3831 6 0 500 i- ETH IPV6 TCP => DROP COUNT 3832 5 0 1000 i- ETH IPV6 ICMP => QUEUE 3833 7 63 0 i- ETH IPV6 UDP VXLAN => MARK QUEUE 3834 testpmd> 3835 3836Toggling isolated mode 3837~~~~~~~~~~~~~~~~~~~~~~ 3838 3839``flow isolate`` can be used to tell the underlying PMD that ingress traffic 3840must only be injected from the defined flow rules; that no default traffic 3841is expected outside those rules and the driver is free to assign more 3842resources to handle them. It is bound to ``rte_flow_isolate()``:: 3843 3844 flow isolate {port_id} {boolean} 3845 3846If successful, enabling or disabling isolated mode shows either:: 3847 3848 Ingress traffic on port [...] 3849 is now restricted to the defined flow rules 3850 3851Or:: 3852 3853 Ingress traffic on port [...] 3854 is not restricted anymore to the defined flow rules 3855 3856Otherwise, in case of error:: 3857 3858 Caught error type [...] ([...]): [...] 3859 3860Mainly due to its side effects, PMDs supporting this mode may not have the 3861ability to toggle it more than once without reinitializing affected ports 3862first (e.g. by exiting testpmd). 3863 3864Enabling isolated mode:: 3865 3866 testpmd> flow isolate 0 true 3867 Ingress traffic on port 0 is now restricted to the defined flow rules 3868 testpmd> 3869 3870Disabling isolated mode:: 3871 3872 testpmd> flow isolate 0 false 3873 Ingress traffic on port 0 is not restricted anymore to the defined flow rules 3874 testpmd> 3875 3876Sample QinQ flow rules 3877~~~~~~~~~~~~~~~~~~~~~~ 3878 3879Before creating QinQ rule(s) the following commands should be issued to enable QinQ:: 3880 3881 testpmd> port stop 0 3882 testpmd> vlan set qinq on 0 3883 3884The above command sets the inner and outer TPID's to 0x8100. 3885 3886To change the TPID's the following commands should be used:: 3887 3888 testpmd> vlan set outer tpid 0xa100 0 3889 testpmd> vlan set inner tpid 0x9100 0 3890 testpmd> port start 0 3891 3892Validate and create a QinQ rule on port 0 to steer traffic to a VF queue in a VM. 3893 3894:: 3895 3896 testpmd> flow validate 0 ingress pattern eth / vlan tci is 123 / 3897 vlan tci is 456 / end actions vf id 1 / queue index 0 / end 3898 Flow rule #0 validated 3899 3900 testpmd> flow create 0 ingress pattern eth / vlan tci is 4 / 3901 vlan tci is 456 / end actions vf id 123 / queue index 0 / end 3902 Flow rule #0 created 3903 3904 testpmd> flow list 0 3905 ID Group Prio Attr Rule 3906 0 0 0 i- ETH VLAN VLAN=>VF QUEUE 3907 3908Validate and create a QinQ rule on port 0 to steer traffic to a queue on the host. 3909 3910:: 3911 3912 testpmd> flow validate 0 ingress pattern eth / vlan tci is 321 / 3913 vlan tci is 654 / end actions pf / queue index 0 / end 3914 Flow rule #1 validated 3915 3916 testpmd> flow create 0 ingress pattern eth / vlan tci is 321 / 3917 vlan tci is 654 / end actions pf / queue index 1 / end 3918 Flow rule #1 created 3919 3920 testpmd> flow list 0 3921 ID Group Prio Attr Rule 3922 0 0 0 i- ETH VLAN VLAN=>VF QUEUE 3923 1 0 0 i- ETH VLAN VLAN=>PF QUEUE 3924 3925BPF Functions 3926-------------- 3927 3928The following sections show functions to load/unload eBPF based filters. 3929 3930bpf-load 3931~~~~~~~~ 3932 3933Load an eBPF program as a callback for partciular RX/TX queue:: 3934 3935 testpmd> bpf-load rx|tx (portid) (queueid) (load-flags) (bpf-prog-filename) 3936 3937The available load-flags are: 3938 3939* ``J``: use JIT generated native code, otherwise BPF interpreter will be used. 3940 3941* ``M``: assume input parameter is a pointer to rte_mbuf, otherwise assume it is a pointer to first segment's data. 3942 3943* ``-``: none. 3944 3945.. note:: 3946 3947 You'll need clang v3.7 or above to build bpf program you'd like to load 3948 3949For example: 3950 3951.. code-block:: console 3952 3953 cd test/bpf 3954 clang -O2 -target bpf -c t1.c 3955 3956Then to load (and JIT compile) t1.o at RX queue 0, port 1:: 3957 3958.. code-block:: console 3959 3960 testpmd> bpf-load rx 1 0 J ./dpdk.org/test/bpf/t1.o 3961 3962To load (not JITed) t1.o at TX queue 0, port 0:: 3963 3964.. code-block:: console 3965 3966 testpmd> bpf-load tx 0 0 - ./dpdk.org/test/bpf/t1.o 3967 3968bpf-unload 3969~~~~~~~~~~ 3970 3971Unload previously loaded eBPF program for partciular RX/TX queue:: 3972 3973 testpmd> bpf-unload rx|tx (portid) (queueid) 3974 3975For example to unload BPF filter from TX queue 0, port 0: 3976 3977.. code-block:: console 3978 3979 testpmd> bpf-load tx 0 0 - ./dpdk.org/test/bpf/t1.o 3980