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 395 396 397Configuration Functions 398----------------------- 399 400The testpmd application can be configured from the runtime as well as from the command-line. 401 402This section details the available configuration functions that are available. 403 404.. note:: 405 406 Configuration changes only become active when forwarding is started/restarted. 407 408set default 409~~~~~~~~~~~ 410 411Reset forwarding to the default configuration:: 412 413 testpmd> set default 414 415set verbose 416~~~~~~~~~~~ 417 418Set the debug verbosity level:: 419 420 testpmd> set verbose (level) 421 422Currently the only available levels are 0 (silent except for error) and 1 (fully verbose). 423 424set log 425~~~~~~~ 426 427Set the log level for a log type:: 428 429 testpmd> set log global|(type) (level) 430 431Where: 432 433* ``type`` is the log name. 434 435* ``level`` is the log level. 436 437For example, to change the global log level:: 438 testpmd> set log global (level) 439 440Regexes can also be used for type. To change log level of user1, user2 and user3:: 441 testpmd> set log user[1-3] (level) 442 443set nbport 444~~~~~~~~~~ 445 446Set the number of ports used by the application: 447 448set nbport (num) 449 450This is equivalent to the ``--nb-ports`` command-line option. 451 452set nbcore 453~~~~~~~~~~ 454 455Set the number of cores used by the application:: 456 457 testpmd> set nbcore (num) 458 459This is equivalent to the ``--nb-cores`` command-line option. 460 461.. note:: 462 463 The number of cores used must not be greater than number of ports used multiplied by the number of queues per port. 464 465set coremask 466~~~~~~~~~~~~ 467 468Set the forwarding cores hexadecimal mask:: 469 470 testpmd> set coremask (mask) 471 472This is equivalent to the ``--coremask`` command-line option. 473 474.. note:: 475 476 The master lcore is reserved for command line parsing only and cannot be masked on for packet forwarding. 477 478set portmask 479~~~~~~~~~~~~ 480 481Set the forwarding ports hexadecimal mask:: 482 483 testpmd> set portmask (mask) 484 485This is equivalent to the ``--portmask`` command-line option. 486 487set burst 488~~~~~~~~~ 489 490Set number of packets per burst:: 491 492 testpmd> set burst (num) 493 494This is equivalent to the ``--burst command-line`` option. 495 496When retry is enabled, the transmit delay time and number of retries can also be set:: 497 498 testpmd> set burst tx delay (microseconds) retry (num) 499 500set txpkts 501~~~~~~~~~~ 502 503Set the length of each segment of the TX-ONLY packets or length of packet for FLOWGEN mode:: 504 505 testpmd> set txpkts (x[,y]*) 506 507Where x[,y]* represents a CSV list of values, without white space. 508 509set txsplit 510~~~~~~~~~~~ 511 512Set the split policy for the TX packets, applicable for TX-ONLY and CSUM forwarding modes:: 513 514 testpmd> set txsplit (off|on|rand) 515 516Where: 517 518* ``off`` disable packet copy & split for CSUM mode. 519 520* ``on`` split outgoing packet into multiple segments. Size of each segment 521 and number of segments per packet is determined by ``set txpkts`` command 522 (see above). 523 524* ``rand`` same as 'on', but number of segments per each packet is a random value between 1 and total number of segments. 525 526set corelist 527~~~~~~~~~~~~ 528 529Set the list of forwarding cores:: 530 531 testpmd> set corelist (x[,y]*) 532 533For example, to change the forwarding cores: 534 535.. code-block:: console 536 537 testpmd> set corelist 3,1 538 testpmd> show config fwd 539 540 io packet forwarding - ports=2 - cores=2 - streams=2 - NUMA support disabled 541 Logical Core 3 (socket 0) forwards packets on 1 streams: 542 RX P=0/Q=0 (socket 0) -> TX P=1/Q=0 (socket 0) peer=02:00:00:00:00:01 543 Logical Core 1 (socket 0) forwards packets on 1 streams: 544 RX P=1/Q=0 (socket 0) -> TX P=0/Q=0 (socket 0) peer=02:00:00:00:00:00 545 546.. note:: 547 548 The cores are used in the same order as specified on the command line. 549 550set portlist 551~~~~~~~~~~~~ 552 553Set the list of forwarding ports:: 554 555 testpmd> set portlist (x[,y]*) 556 557For example, to change the port forwarding: 558 559.. code-block:: console 560 561 testpmd> set portlist 0,2,1,3 562 testpmd> show config fwd 563 564 io packet forwarding - ports=4 - cores=1 - streams=4 565 Logical Core 3 (socket 0) forwards packets on 4 streams: 566 RX P=0/Q=0 (socket 0) -> TX P=2/Q=0 (socket 0) peer=02:00:00:00:00:01 567 RX P=2/Q=0 (socket 0) -> TX P=0/Q=0 (socket 0) peer=02:00:00:00:00:00 568 RX P=1/Q=0 (socket 0) -> TX P=3/Q=0 (socket 0) peer=02:00:00:00:00:03 569 RX P=3/Q=0 (socket 0) -> TX P=1/Q=0 (socket 0) peer=02:00:00:00:00:02 570 571set tx loopback 572~~~~~~~~~~~~~~~ 573 574Enable/disable tx loopback:: 575 576 testpmd> set tx loopback (port_id) (on|off) 577 578set drop enable 579~~~~~~~~~~~~~~~ 580 581set drop enable bit for all queues:: 582 583 testpmd> set all queues drop (port_id) (on|off) 584 585set split drop enable (for VF) 586~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ 587 588set split drop enable bit for VF from PF:: 589 590 testpmd> set vf split drop (port_id) (vf_id) (on|off) 591 592set mac antispoof (for VF) 593~~~~~~~~~~~~~~~~~~~~~~~~~~ 594 595Set mac antispoof for a VF from the PF:: 596 597 testpmd> set vf mac antispoof (port_id) (vf_id) (on|off) 598 599set macsec offload 600~~~~~~~~~~~~~~~~~~ 601 602Enable/disable MACsec offload:: 603 604 testpmd> set macsec offload (port_id) on encrypt (on|off) replay-protect (on|off) 605 testpmd> set macsec offload (port_id) off 606 607set macsec sc 608~~~~~~~~~~~~~ 609 610Configure MACsec secure connection (SC):: 611 612 testpmd> set macsec sc (tx|rx) (port_id) (mac) (pi) 613 614.. note:: 615 616 The pi argument is ignored for tx. 617 Check the NIC Datasheet for hardware limits. 618 619set macsec sa 620~~~~~~~~~~~~~ 621 622Configure MACsec secure association (SA):: 623 624 testpmd> set macsec sa (tx|rx) (port_id) (idx) (an) (pn) (key) 625 626.. note:: 627 628 The IDX value must be 0 or 1. 629 Check the NIC Datasheet for hardware limits. 630 631set broadcast mode (for VF) 632~~~~~~~~~~~~~~~~~~~~~~~~~~~ 633 634Set broadcast mode for a VF from the PF:: 635 636 testpmd> set vf broadcast (port_id) (vf_id) (on|off) 637 638vlan set strip 639~~~~~~~~~~~~~~ 640 641Set the VLAN strip on a port:: 642 643 testpmd> vlan set strip (on|off) (port_id) 644 645vlan set stripq 646~~~~~~~~~~~~~~~ 647 648Set the VLAN strip for a queue on a port:: 649 650 testpmd> vlan set stripq (on|off) (port_id,queue_id) 651 652vlan set stripq (for VF) 653~~~~~~~~~~~~~~~~~~~~~~~~ 654 655Set VLAN strip for all queues in a pool for a VF from the PF:: 656 657 testpmd> set vf vlan stripq (port_id) (vf_id) (on|off) 658 659vlan set insert (for VF) 660~~~~~~~~~~~~~~~~~~~~~~~~ 661 662Set VLAN insert for a VF from the PF:: 663 664 testpmd> set vf vlan insert (port_id) (vf_id) (vlan_id) 665 666vlan set tag (for VF) 667~~~~~~~~~~~~~~~~~~~~~ 668 669Set VLAN tag for a VF from the PF:: 670 671 testpmd> set vf vlan tag (port_id) (vf_id) (on|off) 672 673vlan set antispoof (for VF) 674~~~~~~~~~~~~~~~~~~~~~~~~~~~ 675 676Set VLAN antispoof for a VF from the PF:: 677 678 testpmd> set vf vlan antispoof (port_id) (vf_id) (on|off) 679 680vlan set filter 681~~~~~~~~~~~~~~~ 682 683Set the VLAN filter on a port:: 684 685 testpmd> vlan set filter (on|off) (port_id) 686 687vlan set qinq 688~~~~~~~~~~~~~ 689 690Set the VLAN QinQ (extended queue in queue) on for a port:: 691 692 testpmd> vlan set qinq (on|off) (port_id) 693 694vlan set tpid 695~~~~~~~~~~~~~ 696 697Set the inner or outer VLAN TPID for packet filtering on a port:: 698 699 testpmd> vlan set (inner|outer) tpid (value) (port_id) 700 701.. note:: 702 703 TPID value must be a 16-bit number (value <= 65536). 704 705rx_vlan add 706~~~~~~~~~~~ 707 708Add a VLAN ID, or all identifiers, to the set of VLAN identifiers filtered by port ID:: 709 710 testpmd> rx_vlan add (vlan_id|all) (port_id) 711 712.. note:: 713 714 VLAN filter must be set on that port. VLAN ID < 4096. 715 Depending on the NIC used, number of vlan_ids may be limited to the maximum entries 716 in VFTA table. This is important if enabling all vlan_ids. 717 718rx_vlan rm 719~~~~~~~~~~ 720 721Remove a VLAN ID, or all identifiers, from the set of VLAN identifiers filtered by port ID:: 722 723 testpmd> rx_vlan rm (vlan_id|all) (port_id) 724 725rx_vlan add (for VF) 726~~~~~~~~~~~~~~~~~~~~ 727 728Add a VLAN ID, to the set of VLAN identifiers filtered for VF(s) for port ID:: 729 730 testpmd> rx_vlan add (vlan_id) port (port_id) vf (vf_mask) 731 732rx_vlan rm (for VF) 733~~~~~~~~~~~~~~~~~~~ 734 735Remove a VLAN ID, from the set of VLAN identifiers filtered for VF(s) for port ID:: 736 737 testpmd> rx_vlan rm (vlan_id) port (port_id) vf (vf_mask) 738 739tunnel_filter add 740~~~~~~~~~~~~~~~~~ 741 742Add a tunnel filter on a port:: 743 744 testpmd> tunnel_filter add (port_id) (outer_mac) (inner_mac) (ip_addr) \ 745 (inner_vlan) (vxlan|nvgre|ipingre) (imac-ivlan|imac-ivlan-tenid|\ 746 imac-tenid|imac|omac-imac-tenid|oip|iip) (tenant_id) (queue_id) 747 748The available information categories are: 749 750* ``vxlan``: Set tunnel type as VXLAN. 751 752* ``nvgre``: Set tunnel type as NVGRE. 753 754* ``ipingre``: Set tunnel type as IP-in-GRE. 755 756* ``imac-ivlan``: Set filter type as Inner MAC and VLAN. 757 758* ``imac-ivlan-tenid``: Set filter type as Inner MAC, VLAN and tenant ID. 759 760* ``imac-tenid``: Set filter type as Inner MAC and tenant ID. 761 762* ``imac``: Set filter type as Inner MAC. 763 764* ``omac-imac-tenid``: Set filter type as Outer MAC, Inner MAC and tenant ID. 765 766* ``oip``: Set filter type as Outer IP. 767 768* ``iip``: Set filter type as Inner IP. 769 770Example:: 771 772 testpmd> tunnel_filter add 0 68:05:CA:28:09:82 00:00:00:00:00:00 \ 773 192.168.2.2 0 ipingre oip 1 1 774 775 Set an IP-in-GRE tunnel on port 0, and the filter type is Outer IP. 776 777tunnel_filter remove 778~~~~~~~~~~~~~~~~~~~~ 779 780Remove a tunnel filter on a port:: 781 782 testpmd> tunnel_filter rm (port_id) (outer_mac) (inner_mac) (ip_addr) \ 783 (inner_vlan) (vxlan|nvgre|ipingre) (imac-ivlan|imac-ivlan-tenid|\ 784 imac-tenid|imac|omac-imac-tenid|oip|iip) (tenant_id) (queue_id) 785 786rx_vxlan_port add 787~~~~~~~~~~~~~~~~~ 788 789Add an UDP port for VXLAN packet filter on a port:: 790 791 testpmd> rx_vxlan_port add (udp_port) (port_id) 792 793rx_vxlan_port remove 794~~~~~~~~~~~~~~~~~~~~ 795 796Remove an UDP port for VXLAN packet filter on a port:: 797 798 testpmd> rx_vxlan_port rm (udp_port) (port_id) 799 800tx_vlan set 801~~~~~~~~~~~ 802 803Set hardware insertion of VLAN IDs in packets sent on a port:: 804 805 testpmd> tx_vlan set (port_id) vlan_id[, vlan_id_outer] 806 807For example, set a single VLAN ID (5) insertion on port 0:: 808 809 tx_vlan set 0 5 810 811Or, set double VLAN ID (inner: 2, outer: 3) insertion on port 1:: 812 813 tx_vlan set 1 2 3 814 815 816tx_vlan set pvid 817~~~~~~~~~~~~~~~~ 818 819Set port based hardware insertion of VLAN ID in packets sent on a port:: 820 821 testpmd> tx_vlan set pvid (port_id) (vlan_id) (on|off) 822 823tx_vlan reset 824~~~~~~~~~~~~~ 825 826Disable hardware insertion of a VLAN header in packets sent on a port:: 827 828 testpmd> tx_vlan reset (port_id) 829 830csum set 831~~~~~~~~ 832 833Select hardware or software calculation of the checksum when 834transmitting a packet using the ``csum`` forwarding engine:: 835 836 testpmd> csum set (ip|udp|tcp|sctp|outer-ip) (hw|sw) (port_id) 837 838Where: 839 840* ``ip|udp|tcp|sctp`` always relate to the inner layer. 841 842* ``outer-ip`` relates to the outer IP layer (only for IPv4) in the case where the packet is recognized 843 as a tunnel packet by the forwarding engine (vxlan, gre and ipip are 844 supported). See also the ``csum parse-tunnel`` command. 845 846.. note:: 847 848 Check the NIC Datasheet for hardware limits. 849 850RSS queue region 851~~~~~~~~~~~~~~~~ 852 853Set RSS queue region span on a port:: 854 855 testpmd> set port (port_id) queue-region region_id (value) \ 856 queue_start_index (value) queue_num (value) 857 858Set flowtype mapping on a RSS queue region on a port:: 859 860 testpmd> set port (port_id) queue-region region_id (value) flowtype (value) 861 862where: 863 864* For the flowtype(pctype) of packet,the specific index for each type has 865 been defined in file i40e_type.h as enum i40e_filter_pctype. 866 867Set user priority mapping on a RSS queue region on a port:: 868 869 testpmd> set port (port_id) queue-region UP (value) region_id (value) 870 871Flush all queue region related configuration on a port:: 872 873 testpmd> set port (port_id) queue-region flush (on|off) 874 875where: 876 877* "on"is just an enable function which server for other configuration, 878 it is for all configuration about queue region from up layer, 879 at first will only keep in DPDK softwarestored in driver, 880 only after "flush on", it commit all configuration to HW. 881 "off" is just clean all configuration about queue region just now, 882 and restore all to DPDK i40e driver default config when start up. 883 884Show all queue region related configuration info on a port:: 885 886 testpmd> show port (port_id) queue-region 887 888.. note:: 889 890 Queue region only support on PF by now, so these command is 891 only for configuration of queue region on PF port. 892 893csum parse-tunnel 894~~~~~~~~~~~~~~~~~ 895 896Define how tunneled packets should be handled by the csum forward 897engine:: 898 899 testpmd> csum parse-tunnel (on|off) (tx_port_id) 900 901If enabled, the csum forward engine will try to recognize supported 902tunnel headers (vxlan, gre, ipip). 903 904If disabled, treat tunnel packets as non-tunneled packets (a inner 905header is handled as a packet payload). 906 907.. note:: 908 909 The port argument is the TX port like in the ``csum set`` command. 910 911Example: 912 913Consider a packet in packet like the following:: 914 915 eth_out/ipv4_out/udp_out/vxlan/eth_in/ipv4_in/tcp_in 916 917* If parse-tunnel is enabled, the ``ip|udp|tcp|sctp`` parameters of ``csum set`` 918 command relate to the inner headers (here ``ipv4_in`` and ``tcp_in``), and the 919 ``outer-ip parameter`` relates to the outer headers (here ``ipv4_out``). 920 921* If parse-tunnel is disabled, the ``ip|udp|tcp|sctp`` parameters of ``csum set`` 922 command relate to the outer headers, here ``ipv4_out`` and ``udp_out``. 923 924csum show 925~~~~~~~~~ 926 927Display tx checksum offload configuration:: 928 929 testpmd> csum show (port_id) 930 931tso set 932~~~~~~~ 933 934Enable TCP Segmentation Offload (TSO) in the ``csum`` forwarding engine:: 935 936 testpmd> tso set (segsize) (port_id) 937 938.. note:: 939 940 Check the NIC datasheet for hardware limits. 941 942tso show 943~~~~~~~~ 944 945Display the status of TCP Segmentation Offload:: 946 947 testpmd> tso show (port_id) 948 949set port - gro 950~~~~~~~~~~~~~~ 951 952Enable or disable GRO in ``csum`` forwarding engine:: 953 954 testpmd> set port <port_id> gro on|off 955 956If enabled, the csum forwarding engine will perform GRO on the TCP/IPv4 957packets received from the given port. 958 959If disabled, packets received from the given port won't be performed 960GRO. By default, GRO is disabled for all ports. 961 962.. note:: 963 964 When enable GRO for a port, TCP/IPv4 packets received from the port 965 will be performed GRO. After GRO, all merged packets have bad 966 checksums, since the GRO library doesn't re-calculate checksums for 967 the merged packets. Therefore, if users want the merged packets to 968 have correct checksums, please select HW IP checksum calculation and 969 HW TCP checksum calculation for the port which the merged packets are 970 transmitted to. 971 972show port - gro 973~~~~~~~~~~~~~~~ 974 975Display GRO configuration for a given port:: 976 977 testpmd> show port <port_id> gro 978 979set gro flush 980~~~~~~~~~~~~~ 981 982Set the cycle to flush the GROed packets from reassembly tables:: 983 984 testpmd> set gro flush <cycles> 985 986When enable GRO, the csum forwarding engine performs GRO on received 987packets, and the GROed packets are stored in reassembly tables. Users 988can use this command to determine when the GROed packets are flushed 989from the reassembly tables. 990 991The ``cycles`` is measured in GRO operation times. The csum forwarding 992engine flushes the GROed packets from the tables every ``cycles`` GRO 993operations. 994 995By default, the value of ``cycles`` is 1, which means flush GROed packets 996from the reassembly tables as soon as one GRO operation finishes. The value 997of ``cycles`` should be in the range of 1 to ``GRO_MAX_FLUSH_CYCLES``. 998 999Please note that the large value of ``cycles`` may cause the poor TCP/IP 1000stack performance. Because the GROed packets are delayed to arrive the 1001stack, thus causing more duplicated ACKs and TCP retransmissions. 1002 1003set port - gso 1004~~~~~~~~~~~~~~ 1005 1006Toggle per-port GSO support in ``csum`` forwarding engine:: 1007 1008 testpmd> set port <port_id> gso on|off 1009 1010If enabled, the csum forwarding engine will perform GSO on supported IPv4 1011packets, transmitted on the given port. 1012 1013If disabled, packets transmitted on the given port will not undergo GSO. 1014By default, GSO is disabled for all ports. 1015 1016.. note:: 1017 1018 When GSO is enabled on a port, supported IPv4 packets transmitted on that 1019 port undergo GSO. Afterwards, the segmented packets are represented by 1020 multi-segment mbufs; however, the csum forwarding engine doesn't calculation 1021 of checksums for GSO'd segments in SW. As a result, if users want correct 1022 checksums in GSO segments, they should enable HW checksum calculation for 1023 GSO-enabled ports. 1024 1025 For example, HW checksum calculation for VxLAN GSO'd packets may be enabled 1026 by setting the following options in the csum forwarding engine: 1027 1028 testpmd> csum set outer_ip hw <port_id> 1029 1030 testpmd> csum set ip hw <port_id> 1031 1032 testpmd> csum set tcp hw <port_id> 1033 1034set gso segsz 1035~~~~~~~~~~~~~ 1036 1037Set the maximum GSO segment size (measured in bytes), which includes the 1038packet header and the packet payload for GSO-enabled ports (global):: 1039 1040 testpmd> set gso segsz <length> 1041 1042show port - gso 1043~~~~~~~~~~~~~~~ 1044 1045Display the status of Generic Segmentation Offload for a given port:: 1046 1047 testpmd> show port <port_id> gso 1048 1049mac_addr add 1050~~~~~~~~~~~~ 1051 1052Add an alternative MAC address to a port:: 1053 1054 testpmd> mac_addr add (port_id) (XX:XX:XX:XX:XX:XX) 1055 1056mac_addr remove 1057~~~~~~~~~~~~~~~ 1058 1059Remove a MAC address from a port:: 1060 1061 testpmd> mac_addr remove (port_id) (XX:XX:XX:XX:XX:XX) 1062 1063mac_addr add (for VF) 1064~~~~~~~~~~~~~~~~~~~~~ 1065 1066Add an alternative MAC address for a VF to a port:: 1067 1068 testpmd> mac_add add port (port_id) vf (vf_id) (XX:XX:XX:XX:XX:XX) 1069 1070mac_addr set 1071~~~~~~~~~~~~ 1072 1073Set the default MAC address for a port:: 1074 1075 testpmd> mac_addr set (port_id) (XX:XX:XX:XX:XX:XX) 1076 1077mac_addr set (for VF) 1078~~~~~~~~~~~~~~~~~~~~~ 1079 1080Set the MAC address for a VF from the PF:: 1081 1082 testpmd> set vf mac addr (port_id) (vf_id) (XX:XX:XX:XX:XX:XX) 1083 1084set eth-peer 1085~~~~~~~~~~~~ 1086 1087Set the forwarding peer address for certain port:: 1088 1089 testpmd> set eth-peer (port_id) (perr_addr) 1090 1091This is equivalent to the ``--eth-peer`` command-line option. 1092 1093set port-uta 1094~~~~~~~~~~~~ 1095 1096Set the unicast hash filter(s) on/off for a port:: 1097 1098 testpmd> set port (port_id) uta (XX:XX:XX:XX:XX:XX|all) (on|off) 1099 1100set promisc 1101~~~~~~~~~~~ 1102 1103Set the promiscuous mode on for a port or for all ports. 1104In promiscuous mode packets are not dropped if they aren't for the specified MAC address:: 1105 1106 testpmd> set promisc (port_id|all) (on|off) 1107 1108set allmulti 1109~~~~~~~~~~~~ 1110 1111Set the allmulti mode for a port or for all ports:: 1112 1113 testpmd> set allmulti (port_id|all) (on|off) 1114 1115Same as the ifconfig (8) option. Controls how multicast packets are handled. 1116 1117set promisc (for VF) 1118~~~~~~~~~~~~~~~~~~~~ 1119 1120Set the unicast promiscuous mode for a VF from PF. 1121It's supported by Intel i40e NICs now. 1122In promiscuous mode packets are not dropped if they aren't for the specified MAC address:: 1123 1124 testpmd> set vf promisc (port_id) (vf_id) (on|off) 1125 1126set allmulticast (for VF) 1127~~~~~~~~~~~~~~~~~~~~~~~~~ 1128 1129Set the multicast promiscuous mode for a VF from PF. 1130It's supported by Intel i40e NICs now. 1131In promiscuous mode packets are not dropped if they aren't for the specified MAC address:: 1132 1133 testpmd> set vf allmulti (port_id) (vf_id) (on|off) 1134 1135set tx max bandwidth (for VF) 1136~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ 1137 1138Set TX max absolute bandwidth (Mbps) for a VF from PF:: 1139 1140 testpmd> set vf tx max-bandwidth (port_id) (vf_id) (max_bandwidth) 1141 1142set tc tx min bandwidth (for VF) 1143~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ 1144 1145Set all TCs' TX min relative bandwidth (%) for a VF from PF:: 1146 1147 testpmd> set vf tc tx min-bandwidth (port_id) (vf_id) (bw1, bw2, ...) 1148 1149set tc tx max bandwidth (for VF) 1150~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ 1151 1152Set a TC's TX max absolute bandwidth (Mbps) for a VF from PF:: 1153 1154 testpmd> set vf tc tx max-bandwidth (port_id) (vf_id) (tc_no) (max_bandwidth) 1155 1156set tc strict link priority mode 1157~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ 1158 1159Set some TCs' strict link priority mode on a physical port:: 1160 1161 testpmd> set tx strict-link-priority (port_id) (tc_bitmap) 1162 1163set tc tx min bandwidth 1164~~~~~~~~~~~~~~~~~~~~~~~ 1165 1166Set all TCs' TX min relative bandwidth (%) globally for all PF and VFs:: 1167 1168 testpmd> set tc tx min-bandwidth (port_id) (bw1, bw2, ...) 1169 1170set flow_ctrl rx 1171~~~~~~~~~~~~~~~~ 1172 1173Set the link flow control parameter on a port:: 1174 1175 testpmd> set flow_ctrl rx (on|off) tx (on|off) (high_water) (low_water) \ 1176 (pause_time) (send_xon) mac_ctrl_frame_fwd (on|off) \ 1177 autoneg (on|off) (port_id) 1178 1179Where: 1180 1181* ``high_water`` (integer): High threshold value to trigger XOFF. 1182 1183* ``low_water`` (integer): Low threshold value to trigger XON. 1184 1185* ``pause_time`` (integer): Pause quota in the Pause frame. 1186 1187* ``send_xon`` (0/1): Send XON frame. 1188 1189* ``mac_ctrl_frame_fwd``: Enable receiving MAC control frames. 1190 1191* ``autoneg``: Change the auto-negotiation parameter. 1192 1193set pfc_ctrl rx 1194~~~~~~~~~~~~~~~ 1195 1196Set the priority flow control parameter on a port:: 1197 1198 testpmd> set pfc_ctrl rx (on|off) tx (on|off) (high_water) (low_water) \ 1199 (pause_time) (priority) (port_id) 1200 1201Where: 1202 1203* ``high_water`` (integer): High threshold value. 1204 1205* ``low_water`` (integer): Low threshold value. 1206 1207* ``pause_time`` (integer): Pause quota in the Pause frame. 1208 1209* ``priority`` (0-7): VLAN User Priority. 1210 1211set stat_qmap 1212~~~~~~~~~~~~~ 1213 1214Set statistics mapping (qmapping 0..15) for RX/TX queue on port:: 1215 1216 testpmd> set stat_qmap (tx|rx) (port_id) (queue_id) (qmapping) 1217 1218For example, to set rx queue 2 on port 0 to mapping 5:: 1219 1220 testpmd>set stat_qmap rx 0 2 5 1221 1222set xstats-hide-zero 1223~~~~~~~~~~~~~~~~~~~~ 1224 1225Set the option to hide zero values for xstats display:: 1226 1227 testpmd> set xstats-hide-zero on|off 1228 1229.. note:: 1230 1231 By default, the zero values are displayed for xstats. 1232 1233set port - rx/tx (for VF) 1234~~~~~~~~~~~~~~~~~~~~~~~~~ 1235 1236Set VF receive/transmit from a port:: 1237 1238 testpmd> set port (port_id) vf (vf_id) (rx|tx) (on|off) 1239 1240set port - mac address filter (for VF) 1241~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ 1242 1243Add/Remove unicast or multicast MAC addr filter for a VF:: 1244 1245 testpmd> set port (port_id) vf (vf_id) (mac_addr) \ 1246 (exact-mac|exact-mac-vlan|hashmac|hashmac-vlan) (on|off) 1247 1248set port - rx mode(for VF) 1249~~~~~~~~~~~~~~~~~~~~~~~~~~ 1250 1251Set the VF receive mode of a port:: 1252 1253 testpmd> set port (port_id) vf (vf_id) \ 1254 rxmode (AUPE|ROPE|BAM|MPE) (on|off) 1255 1256The available receive modes are: 1257 1258* ``AUPE``: Accepts untagged VLAN. 1259 1260* ``ROPE``: Accepts unicast hash. 1261 1262* ``BAM``: Accepts broadcast packets. 1263 1264* ``MPE``: Accepts all multicast packets. 1265 1266set port - tx_rate (for Queue) 1267~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ 1268 1269Set TX rate limitation for a queue on a port:: 1270 1271 testpmd> set port (port_id) queue (queue_id) rate (rate_value) 1272 1273set port - tx_rate (for VF) 1274~~~~~~~~~~~~~~~~~~~~~~~~~~~ 1275 1276Set TX rate limitation for queues in VF on a port:: 1277 1278 testpmd> set port (port_id) vf (vf_id) rate (rate_value) queue_mask (queue_mask) 1279 1280set port - mirror rule 1281~~~~~~~~~~~~~~~~~~~~~~ 1282 1283Set pool or vlan type mirror rule for a port:: 1284 1285 testpmd> set port (port_id) mirror-rule (rule_id) \ 1286 (pool-mirror-up|pool-mirror-down|vlan-mirror) \ 1287 (poolmask|vlanid[,vlanid]*) dst-pool (pool_id) (on|off) 1288 1289Set link mirror rule for a port:: 1290 1291 testpmd> set port (port_id) mirror-rule (rule_id) \ 1292 (uplink-mirror|downlink-mirror) dst-pool (pool_id) (on|off) 1293 1294For example to enable mirror traffic with vlan 0,1 to pool 0:: 1295 1296 set port 0 mirror-rule 0 vlan-mirror 0,1 dst-pool 0 on 1297 1298reset port - mirror rule 1299~~~~~~~~~~~~~~~~~~~~~~~~ 1300 1301Reset a mirror rule for a port:: 1302 1303 testpmd> reset port (port_id) mirror-rule (rule_id) 1304 1305set flush_rx 1306~~~~~~~~~~~~ 1307 1308Set the flush on RX streams before forwarding. 1309The default is flush ``on``. 1310Mainly used with PCAP drivers to turn off the default behavior of flushing the first 512 packets on RX streams:: 1311 1312 testpmd> set flush_rx off 1313 1314set bypass mode 1315~~~~~~~~~~~~~~~ 1316 1317Set the bypass mode for the lowest port on bypass enabled NIC:: 1318 1319 testpmd> set bypass mode (normal|bypass|isolate) (port_id) 1320 1321set bypass event 1322~~~~~~~~~~~~~~~~ 1323 1324Set the event required to initiate specified bypass mode for the lowest port on a bypass enabled:: 1325 1326 testpmd> set bypass event (timeout|os_on|os_off|power_on|power_off) \ 1327 mode (normal|bypass|isolate) (port_id) 1328 1329Where: 1330 1331* ``timeout``: Enable bypass after watchdog timeout. 1332 1333* ``os_on``: Enable bypass when OS/board is powered on. 1334 1335* ``os_off``: Enable bypass when OS/board is powered off. 1336 1337* ``power_on``: Enable bypass when power supply is turned on. 1338 1339* ``power_off``: Enable bypass when power supply is turned off. 1340 1341 1342set bypass timeout 1343~~~~~~~~~~~~~~~~~~ 1344 1345Set the bypass watchdog timeout to ``n`` seconds where 0 = instant:: 1346 1347 testpmd> set bypass timeout (0|1.5|2|3|4|8|16|32) 1348 1349show bypass config 1350~~~~~~~~~~~~~~~~~~ 1351 1352Show the bypass configuration for a bypass enabled NIC using the lowest port on the NIC:: 1353 1354 testpmd> show bypass config (port_id) 1355 1356set link up 1357~~~~~~~~~~~ 1358 1359Set link up for a port:: 1360 1361 testpmd> set link-up port (port id) 1362 1363set link down 1364~~~~~~~~~~~~~ 1365 1366Set link down for a port:: 1367 1368 testpmd> set link-down port (port id) 1369 1370E-tag set 1371~~~~~~~~~ 1372 1373Enable E-tag insertion for a VF on a port:: 1374 1375 testpmd> E-tag set insertion on port-tag-id (value) port (port_id) vf (vf_id) 1376 1377Disable E-tag insertion for a VF on a port:: 1378 1379 testpmd> E-tag set insertion off port (port_id) vf (vf_id) 1380 1381Enable/disable E-tag stripping on a port:: 1382 1383 testpmd> E-tag set stripping (on|off) port (port_id) 1384 1385Enable/disable E-tag based forwarding on a port:: 1386 1387 testpmd> E-tag set forwarding (on|off) port (port_id) 1388 1389Add an E-tag forwarding filter on a port:: 1390 1391 testpmd> E-tag set filter add e-tag-id (value) dst-pool (pool_id) port (port_id) 1392 1393Delete an E-tag forwarding filter on a port:: 1394 testpmd> E-tag set filter del e-tag-id (value) port (port_id) 1395 1396ddp add 1397~~~~~~~ 1398 1399Load a dynamic device personalization (DDP) profile and store backup profile:: 1400 1401 testpmd> ddp add (port_id) (profile_path[,backup_profile_path]) 1402 1403ddp del 1404~~~~~~~ 1405 1406Delete a dynamic device personalization profile and restore backup profile:: 1407 1408 testpmd> ddp del (port_id) (backup_profile_path) 1409 1410ptype mapping 1411~~~~~~~~~~~~~ 1412 1413List all items from the ptype mapping table:: 1414 1415 testpmd> ptype mapping get (port_id) (valid_only) 1416 1417Where: 1418 1419* ``valid_only``: A flag indicates if only list valid items(=1) or all itemss(=0). 1420 1421Replace a specific or a group of software defined ptype with a new one:: 1422 1423 testpmd> ptype mapping replace (port_id) (target) (mask) (pkt_type) 1424 1425where: 1426 1427* ``target``: A specific software ptype or a mask to represent a group of software ptypes. 1428 1429* ``mask``: A flag indicate if "target" is a specific software ptype(=0) or a ptype mask(=1). 1430 1431* ``pkt_type``: The new software ptype to replace the old ones. 1432 1433Update hardware defined ptype to software defined packet type mapping table:: 1434 1435 testpmd> ptype mapping update (port_id) (hw_ptype) (sw_ptype) 1436 1437where: 1438 1439* ``hw_ptype``: hardware ptype as the index of the ptype mapping table. 1440 1441* ``sw_ptype``: software ptype as the value of the ptype mapping table. 1442 1443Reset ptype mapping table:: 1444 1445 testpmd> ptype mapping reset (port_id) 1446 1447Port Functions 1448-------------- 1449 1450The following sections show functions for configuring ports. 1451 1452.. note:: 1453 1454 Port configuration changes only become active when forwarding is started/restarted. 1455 1456port attach 1457~~~~~~~~~~~ 1458 1459Attach a port specified by pci address or virtual device args:: 1460 1461 testpmd> port attach (identifier) 1462 1463To attach a new pci device, the device should be recognized by kernel first. 1464Then it should be moved under DPDK management. 1465Finally the port can be attached to testpmd. 1466 1467For example, to move a pci device using ixgbe under DPDK management: 1468 1469.. code-block:: console 1470 1471 # Check the status of the available devices. 1472 ./usertools/dpdk-devbind.py --status 1473 1474 Network devices using DPDK-compatible driver 1475 ============================================ 1476 <none> 1477 1478 Network devices using kernel driver 1479 =================================== 1480 0000:0a:00.0 '82599ES 10-Gigabit' if=eth2 drv=ixgbe unused= 1481 1482 1483 # Bind the device to igb_uio. 1484 sudo ./usertools/dpdk-devbind.py -b igb_uio 0000:0a:00.0 1485 1486 1487 # Recheck the status of the devices. 1488 ./usertools/dpdk-devbind.py --status 1489 Network devices using DPDK-compatible driver 1490 ============================================ 1491 0000:0a:00.0 '82599ES 10-Gigabit' drv=igb_uio unused= 1492 1493To attach a port created by virtual device, above steps are not needed. 1494 1495For example, to attach a port whose pci address is 0000:0a:00.0. 1496 1497.. code-block:: console 1498 1499 testpmd> port attach 0000:0a:00.0 1500 Attaching a new port... 1501 EAL: PCI device 0000:0a:00.0 on NUMA socket -1 1502 EAL: probe driver: 8086:10fb rte_ixgbe_pmd 1503 EAL: PCI memory mapped at 0x7f83bfa00000 1504 EAL: PCI memory mapped at 0x7f83bfa80000 1505 PMD: eth_ixgbe_dev_init(): MAC: 2, PHY: 18, SFP+: 5 1506 PMD: eth_ixgbe_dev_init(): port 0 vendorID=0x8086 deviceID=0x10fb 1507 Port 0 is attached. Now total ports is 1 1508 Done 1509 1510For example, to attach a port created by pcap PMD. 1511 1512.. code-block:: console 1513 1514 testpmd> port attach net_pcap0 1515 Attaching a new port... 1516 PMD: Initializing pmd_pcap for net_pcap0 1517 PMD: Creating pcap-backed ethdev on numa socket 0 1518 Port 0 is attached. Now total ports is 1 1519 Done 1520 1521In this case, identifier is ``net_pcap0``. 1522This identifier format is the same as ``--vdev`` format of DPDK applications. 1523 1524For example, to re-attach a bonded port which has been previously detached, 1525the mode and slave parameters must be given. 1526 1527.. code-block:: console 1528 1529 testpmd> port attach net_bond_0,mode=0,slave=1 1530 Attaching a new port... 1531 EAL: Initializing pmd_bond for net_bond_0 1532 EAL: Create bonded device net_bond_0 on port 0 in mode 0 on socket 0. 1533 Port 0 is attached. Now total ports is 1 1534 Done 1535 1536 1537port detach 1538~~~~~~~~~~~ 1539 1540Detach a specific port:: 1541 1542 testpmd> port detach (port_id) 1543 1544Before detaching a port, the port should be stopped and closed. 1545 1546For example, to detach a pci device port 0. 1547 1548.. code-block:: console 1549 1550 testpmd> port stop 0 1551 Stopping ports... 1552 Done 1553 testpmd> port close 0 1554 Closing ports... 1555 Done 1556 1557 testpmd> port detach 0 1558 Detaching a port... 1559 EAL: PCI device 0000:0a:00.0 on NUMA socket -1 1560 EAL: remove driver: 8086:10fb rte_ixgbe_pmd 1561 EAL: PCI memory unmapped at 0x7f83bfa00000 1562 EAL: PCI memory unmapped at 0x7f83bfa80000 1563 Done 1564 1565 1566For example, to detach a virtual device port 0. 1567 1568.. code-block:: console 1569 1570 testpmd> port stop 0 1571 Stopping ports... 1572 Done 1573 testpmd> port close 0 1574 Closing ports... 1575 Done 1576 1577 testpmd> port detach 0 1578 Detaching a port... 1579 PMD: Closing pcap ethdev on numa socket 0 1580 Port 'net_pcap0' is detached. Now total ports is 0 1581 Done 1582 1583To remove a pci device completely from the system, first detach the port from testpmd. 1584Then the device should be moved under kernel management. 1585Finally the device can be removed using kernel pci hotplug functionality. 1586 1587For example, to move a pci device under kernel management: 1588 1589.. code-block:: console 1590 1591 sudo ./usertools/dpdk-devbind.py -b ixgbe 0000:0a:00.0 1592 1593 ./usertools/dpdk-devbind.py --status 1594 1595 Network devices using DPDK-compatible driver 1596 ============================================ 1597 <none> 1598 1599 Network devices using kernel driver 1600 =================================== 1601 0000:0a:00.0 '82599ES 10-Gigabit' if=eth2 drv=ixgbe unused=igb_uio 1602 1603To remove a port created by a virtual device, above steps are not needed. 1604 1605port start 1606~~~~~~~~~~ 1607 1608Start all ports or a specific port:: 1609 1610 testpmd> port start (port_id|all) 1611 1612port stop 1613~~~~~~~~~ 1614 1615Stop all ports or a specific port:: 1616 1617 testpmd> port stop (port_id|all) 1618 1619port close 1620~~~~~~~~~~ 1621 1622Close all ports or a specific port:: 1623 1624 testpmd> port close (port_id|all) 1625 1626port config - queue ring size 1627~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ 1628 1629Configure a rx/tx queue ring size:: 1630 1631 testpmd> port (port_id) (rxq|txq) (queue_id) ring_size (value) 1632 1633Only take effect after command that (re-)start the port or command that setup specific queue. 1634 1635port start/stop queue 1636~~~~~~~~~~~~~~~~~~~~~ 1637 1638Start/stop a rx/tx queue on a specific port:: 1639 1640 testpmd> port (port_id) (rxq|txq) (queue_id) (start|stop) 1641 1642port setup queue 1643~~~~~~~~~~~~~~~~~~~~~ 1644 1645Setup a rx/tx queue on a specific port:: 1646 1647 testpmd> port (port_id) (rxq|txq) (queue_id) setup 1648 1649Only take effect when port is started. 1650 1651port config - speed 1652~~~~~~~~~~~~~~~~~~~ 1653 1654Set the speed and duplex mode for all ports or a specific port:: 1655 1656 testpmd> port config (port_id|all) speed (10|100|1000|10000|25000|40000|50000|100000|auto) \ 1657 duplex (half|full|auto) 1658 1659port config - queues/descriptors 1660~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ 1661 1662Set number of queues/descriptors for rxq, txq, rxd and txd:: 1663 1664 testpmd> port config all (rxq|txq|rxd|txd) (value) 1665 1666This is equivalent to the ``--rxq``, ``--txq``, ``--rxd`` and ``--txd`` command-line options. 1667 1668port config - max-pkt-len 1669~~~~~~~~~~~~~~~~~~~~~~~~~ 1670 1671Set the maximum packet length:: 1672 1673 testpmd> port config all max-pkt-len (value) 1674 1675This is equivalent to the ``--max-pkt-len`` command-line option. 1676 1677port config - CRC Strip 1678~~~~~~~~~~~~~~~~~~~~~~~ 1679 1680Set hardware CRC stripping on or off for all ports:: 1681 1682 testpmd> port config all crc-strip (on|off) 1683 1684CRC stripping is on by default. 1685 1686The ``off`` option is equivalent to the ``--disable-crc-strip`` command-line option. 1687 1688port config - scatter 1689~~~~~~~~~~~~~~~~~~~~~~~ 1690 1691Set RX scatter mode on or off for all ports:: 1692 1693 testpmd> port config all scatter (on|off) 1694 1695RX scatter mode is off by default. 1696 1697The ``on`` option is equivalent to the ``--enable-scatter`` command-line option. 1698 1699port config - RX Checksum 1700~~~~~~~~~~~~~~~~~~~~~~~~~ 1701 1702Set hardware RX checksum offload to on or off for all ports:: 1703 1704 testpmd> port config all rx-cksum (on|off) 1705 1706Checksum offload is off by default. 1707 1708The ``on`` option is equivalent to the ``--enable-rx-cksum`` command-line option. 1709 1710port config - VLAN 1711~~~~~~~~~~~~~~~~~~ 1712 1713Set hardware VLAN on or off for all ports:: 1714 1715 testpmd> port config all hw-vlan (on|off) 1716 1717Hardware VLAN is off by default. 1718 1719The ``on`` option is equivalent to the ``--enable-hw-vlan`` command-line option. 1720 1721port config - VLAN filter 1722~~~~~~~~~~~~~~~~~~~~~~~~~ 1723 1724Set hardware VLAN filter on or off for all ports:: 1725 1726 testpmd> port config all hw-vlan-filter (on|off) 1727 1728Hardware VLAN filter is off by default. 1729 1730The ``on`` option is equivalent to the ``--enable-hw-vlan-filter`` command-line option. 1731 1732port config - VLAN strip 1733~~~~~~~~~~~~~~~~~~~~~~~~ 1734 1735Set hardware VLAN strip on or off for all ports:: 1736 1737 testpmd> port config all hw-vlan-strip (on|off) 1738 1739Hardware VLAN strip is off by default. 1740 1741The ``on`` option is equivalent to the ``--enable-hw-vlan-strip`` command-line option. 1742 1743port config - VLAN extend 1744~~~~~~~~~~~~~~~~~~~~~~~~~ 1745 1746Set hardware VLAN extend on or off for all ports:: 1747 1748 testpmd> port config all hw-vlan-extend (on|off) 1749 1750Hardware VLAN extend is off by default. 1751 1752The ``on`` option is equivalent to the ``--enable-hw-vlan-extend`` command-line option. 1753 1754port config - Drop Packets 1755~~~~~~~~~~~~~~~~~~~~~~~~~~ 1756 1757Set packet drop for packets with no descriptors on or off for all ports:: 1758 1759 testpmd> port config all drop-en (on|off) 1760 1761Packet dropping for packets with no descriptors is off by default. 1762 1763The ``on`` option is equivalent to the ``--enable-drop-en`` command-line option. 1764 1765port config - RSS 1766~~~~~~~~~~~~~~~~~ 1767 1768Set the RSS (Receive Side Scaling) mode on or off:: 1769 1770 testpmd> port config all rss (all|default|ip|tcp|udp|sctp|ether|port|vxlan|geneve|nvgre|none) 1771 1772RSS is on by default. 1773 1774The ``all`` option is equivalent to ip|tcp|udp|sctp|ether. 1775The ``default`` option enables all supported RSS types reported by device info. 1776The ``none`` option is equivalent to the ``--disable-rss`` command-line option. 1777 1778port config - RSS Reta 1779~~~~~~~~~~~~~~~~~~~~~~ 1780 1781Set the RSS (Receive Side Scaling) redirection table:: 1782 1783 testpmd> port config all rss reta (hash,queue)[,(hash,queue)] 1784 1785port config - DCB 1786~~~~~~~~~~~~~~~~~ 1787 1788Set the DCB mode for an individual port:: 1789 1790 testpmd> port config (port_id) dcb vt (on|off) (traffic_class) pfc (on|off) 1791 1792The traffic class should be 4 or 8. 1793 1794port config - Burst 1795~~~~~~~~~~~~~~~~~~~ 1796 1797Set the number of packets per burst:: 1798 1799 testpmd> port config all burst (value) 1800 1801This is equivalent to the ``--burst`` command-line option. 1802 1803port config - Threshold 1804~~~~~~~~~~~~~~~~~~~~~~~ 1805 1806Set thresholds for TX/RX queues:: 1807 1808 testpmd> port config all (threshold) (value) 1809 1810Where the threshold type can be: 1811 1812* ``txpt:`` Set the prefetch threshold register of the TX rings, 0 <= value <= 255. 1813 1814* ``txht:`` Set the host threshold register of the TX rings, 0 <= value <= 255. 1815 1816* ``txwt:`` Set the write-back threshold register of the TX rings, 0 <= value <= 255. 1817 1818* ``rxpt:`` Set the prefetch threshold register of the RX rings, 0 <= value <= 255. 1819 1820* ``rxht:`` Set the host threshold register of the RX rings, 0 <= value <= 255. 1821 1822* ``rxwt:`` Set the write-back threshold register of the RX rings, 0 <= value <= 255. 1823 1824* ``txfreet:`` Set the transmit free threshold of the TX rings, 0 <= value <= txd. 1825 1826* ``rxfreet:`` Set the transmit free threshold of the RX rings, 0 <= value <= rxd. 1827 1828* ``txrst:`` Set the transmit RS bit threshold of TX rings, 0 <= value <= txd. 1829 1830These threshold options are also available from the command-line. 1831 1832port config - E-tag 1833~~~~~~~~~~~~~~~~~~~ 1834 1835Set the value of ether-type for E-tag:: 1836 1837 testpmd> port config (port_id|all) l2-tunnel E-tag ether-type (value) 1838 1839Enable/disable the E-tag support:: 1840 1841 testpmd> port config (port_id|all) l2-tunnel E-tag (enable|disable) 1842 1843port config pctype mapping 1844~~~~~~~~~~~~~~~~~~~~~~~~~~ 1845 1846Reset pctype mapping table:: 1847 1848 testpmd> port config (port_id) pctype mapping reset 1849 1850Update hardware defined pctype to software defined flow type mapping table:: 1851 1852 testpmd> port config (port_id) pctype mapping update (pctype_id_0[,pctype_id_1]*) (flow_type_id) 1853 1854where: 1855 1856* ``pctype_id_x``: hardware pctype id as index of bit in bitmask value of the pctype mapping table. 1857 1858* ``flow_type_id``: software flow type id as the index of the pctype mapping table. 1859 1860port config input set 1861~~~~~~~~~~~~~~~~~~~~~ 1862 1863Config RSS/FDIR/FDIR flexible payload input set for some pctype:: 1864 testpmd> port config (port_id) pctype (pctype_id) \ 1865 (hash_inset|fdir_inset|fdir_flx_inset) \ 1866 (get|set|clear) field (field_idx) 1867 1868Clear RSS/FDIR/FDIR flexible payload input set for some pctype:: 1869 testpmd> port config (port_id) pctype (pctype_id) \ 1870 (hash_inset|fdir_inset|fdir_flx_inset) clear all 1871 1872where: 1873 1874* ``pctype_id``: hardware packet classification types. 1875* ``field_idx``: hardware field index. 1876 1877port config udp_tunnel_port 1878~~~~~~~~~~~~~~~~~~~~~~~~~~~ 1879 1880Add/remove UDP tunnel port for VXLAN/GENEVE tunneling protocols:: 1881 testpmd> port config (port_id) udp_tunnel_port add|rm vxlan|geneve (udp_port) 1882 1883Link Bonding Functions 1884---------------------- 1885 1886The Link Bonding functions make it possible to dynamically create and 1887manage link bonding devices from within testpmd interactive prompt. 1888 1889create bonded device 1890~~~~~~~~~~~~~~~~~~~~ 1891 1892Create a new bonding device:: 1893 1894 testpmd> create bonded device (mode) (socket) 1895 1896For example, to create a bonded device in mode 1 on socket 0:: 1897 1898 testpmd> create bonded 1 0 1899 created new bonded device (port X) 1900 1901add bonding slave 1902~~~~~~~~~~~~~~~~~ 1903 1904Adds Ethernet device to a Link Bonding device:: 1905 1906 testpmd> add bonding slave (slave id) (port id) 1907 1908For example, to add Ethernet device (port 6) to a Link Bonding device (port 10):: 1909 1910 testpmd> add bonding slave 6 10 1911 1912 1913remove bonding slave 1914~~~~~~~~~~~~~~~~~~~~ 1915 1916Removes an Ethernet slave device from a Link Bonding device:: 1917 1918 testpmd> remove bonding slave (slave id) (port id) 1919 1920For example, to remove Ethernet slave device (port 6) to a Link Bonding device (port 10):: 1921 1922 testpmd> remove bonding slave 6 10 1923 1924set bonding mode 1925~~~~~~~~~~~~~~~~ 1926 1927Set the Link Bonding mode of a Link Bonding device:: 1928 1929 testpmd> set bonding mode (value) (port id) 1930 1931For example, to set the bonding mode of a Link Bonding device (port 10) to broadcast (mode 3):: 1932 1933 testpmd> set bonding mode 3 10 1934 1935set bonding primary 1936~~~~~~~~~~~~~~~~~~~ 1937 1938Set an Ethernet slave device as the primary device on a Link Bonding device:: 1939 1940 testpmd> set bonding primary (slave id) (port id) 1941 1942For example, to set the Ethernet slave device (port 6) as the primary port of a Link Bonding device (port 10):: 1943 1944 testpmd> set bonding primary 6 10 1945 1946set bonding mac 1947~~~~~~~~~~~~~~~ 1948 1949Set the MAC address of a Link Bonding device:: 1950 1951 testpmd> set bonding mac (port id) (mac) 1952 1953For example, to set the MAC address of a Link Bonding device (port 10) to 00:00:00:00:00:01:: 1954 1955 testpmd> set bonding mac 10 00:00:00:00:00:01 1956 1957set bonding xmit_balance_policy 1958~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ 1959 1960Set the transmission policy for a Link Bonding device when it is in Balance XOR mode:: 1961 1962 testpmd> set bonding xmit_balance_policy (port_id) (l2|l23|l34) 1963 1964For example, set a Link Bonding device (port 10) to use a balance policy of layer 3+4 (IP addresses & UDP ports):: 1965 1966 testpmd> set bonding xmit_balance_policy 10 l34 1967 1968 1969set bonding mon_period 1970~~~~~~~~~~~~~~~~~~~~~~ 1971 1972Set the link status monitoring polling period in milliseconds for a bonding device. 1973 1974This adds support for PMD slave devices which do not support link status interrupts. 1975When the mon_period is set to a value greater than 0 then all PMD's which do not support 1976link status ISR will be queried every polling interval to check if their link status has changed:: 1977 1978 testpmd> set bonding mon_period (port_id) (value) 1979 1980For example, to set the link status monitoring polling period of bonded device (port 5) to 150ms:: 1981 1982 testpmd> set bonding mon_period 5 150 1983 1984 1985set bonding lacp dedicated_queue 1986~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ 1987 1988Enable dedicated tx/rx queues on bonding devices slaves to handle LACP control plane traffic 1989when in mode 4 (link-aggregration-802.3ad):: 1990 1991 testpmd> set bonding lacp dedicated_queues (port_id) (enable|disable) 1992 1993 1994set bonding agg_mode 1995~~~~~~~~~~~~~~~~~~~~ 1996 1997Enable one of the specific aggregators mode when in mode 4 (link-aggregration-802.3ad):: 1998 1999 testpmd> set bonding agg_mode (port_id) (bandwidth|count|stable) 2000 2001 2002show bonding config 2003~~~~~~~~~~~~~~~~~~~ 2004 2005Show the current configuration of a Link Bonding device:: 2006 2007 testpmd> show bonding config (port id) 2008 2009For example, 2010to show the configuration a Link Bonding device (port 9) with 3 slave devices (1, 3, 4) 2011in balance mode with a transmission policy of layer 2+3:: 2012 2013 testpmd> show bonding config 9 2014 Bonding mode: 2 2015 Balance Xmit Policy: BALANCE_XMIT_POLICY_LAYER23 2016 Slaves (3): [1 3 4] 2017 Active Slaves (3): [1 3 4] 2018 Primary: [3] 2019 2020 2021Register Functions 2022------------------ 2023 2024The Register Functions can be used to read from and write to registers on the network card referenced by a port number. 2025This is mainly useful for debugging purposes. 2026Reference should be made to the appropriate datasheet for the network card for details on the register addresses 2027and fields that can be accessed. 2028 2029read reg 2030~~~~~~~~ 2031 2032Display the value of a port register:: 2033 2034 testpmd> read reg (port_id) (address) 2035 2036For example, to examine the Flow Director control register (FDIRCTL, 0x0000EE000) on an Intel 82599 10 GbE Controller:: 2037 2038 testpmd> read reg 0 0xEE00 2039 port 0 PCI register at offset 0xEE00: 0x4A060029 (1241907241) 2040 2041read regfield 2042~~~~~~~~~~~~~ 2043 2044Display a port register bit field:: 2045 2046 testpmd> read regfield (port_id) (address) (bit_x) (bit_y) 2047 2048For example, reading the lowest two bits from the register in the example above:: 2049 2050 testpmd> read regfield 0 0xEE00 0 1 2051 port 0 PCI register at offset 0xEE00: bits[0, 1]=0x1 (1) 2052 2053read regbit 2054~~~~~~~~~~~ 2055 2056Display a single port register bit:: 2057 2058 testpmd> read regbit (port_id) (address) (bit_x) 2059 2060For example, reading the lowest bit from the register in the example above:: 2061 2062 testpmd> read regbit 0 0xEE00 0 2063 port 0 PCI register at offset 0xEE00: bit 0=1 2064 2065write reg 2066~~~~~~~~~ 2067 2068Set the value of a port register:: 2069 2070 testpmd> write reg (port_id) (address) (value) 2071 2072For example, to clear a register:: 2073 2074 testpmd> write reg 0 0xEE00 0x0 2075 port 0 PCI register at offset 0xEE00: 0x00000000 (0) 2076 2077write regfield 2078~~~~~~~~~~~~~~ 2079 2080Set bit field of a port register:: 2081 2082 testpmd> write regfield (port_id) (address) (bit_x) (bit_y) (value) 2083 2084For example, writing to the register cleared in the example above:: 2085 2086 testpmd> write regfield 0 0xEE00 0 1 2 2087 port 0 PCI register at offset 0xEE00: 0x00000002 (2) 2088 2089write regbit 2090~~~~~~~~~~~~ 2091 2092Set single bit value of a port register:: 2093 2094 testpmd> write regbit (port_id) (address) (bit_x) (value) 2095 2096For example, to set the high bit in the register from the example above:: 2097 2098 testpmd> write regbit 0 0xEE00 31 1 2099 port 0 PCI register at offset 0xEE00: 0x8000000A (2147483658) 2100 2101Traffic Metering and Policing 2102----------------------------- 2103 2104The following section shows functions for configuring traffic metering and 2105policing on the ethernet device through the use of generic ethdev API. 2106 2107show port traffic management capability 2108~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ 2109 2110Show traffic metering and policing capability of the port:: 2111 2112 testpmd> show port meter cap (port_id) 2113 2114add port meter profile (srTCM rfc2967) 2115~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ 2116 2117Add meter profile (srTCM rfc2697) to the ethernet device:: 2118 2119 testpmd> add port meter profile srtcm_rfc2697 (port_id) (profile_id) \ 2120 (cir) (cbs) (ebs) 2121 2122where: 2123 2124* ``profile_id``: ID for the meter profile. 2125* ``cir``: Committed Information Rate (CIR) (bytes/second). 2126* ``cbs``: Committed Burst Size (CBS) (bytes). 2127* ``ebs``: Excess Burst Size (EBS) (bytes). 2128 2129add port meter profile (trTCM rfc2968) 2130~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ 2131 2132Add meter profile (srTCM rfc2698) to the ethernet device:: 2133 2134 testpmd> add port meter profile trtcm_rfc2698 (port_id) (profile_id) \ 2135 (cir) (pir) (cbs) (pbs) 2136 2137where: 2138 2139* ``profile_id``: ID for the meter profile. 2140* ``cir``: Committed information rate (bytes/second). 2141* ``pir``: Peak information rate (bytes/second). 2142* ``cbs``: Committed burst size (bytes). 2143* ``pbs``: Peak burst size (bytes). 2144 2145add port meter profile (trTCM rfc4115) 2146~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ 2147 2148Add meter profile (trTCM rfc4115) to the ethernet device:: 2149 2150 testpmd> add port meter profile trtcm_rfc4115 (port_id) (profile_id) \ 2151 (cir) (eir) (cbs) (ebs) 2152 2153where: 2154 2155* ``profile_id``: ID for the meter profile. 2156* ``cir``: Committed information rate (bytes/second). 2157* ``eir``: Excess information rate (bytes/second). 2158* ``cbs``: Committed burst size (bytes). 2159* ``ebs``: Excess burst size (bytes). 2160 2161delete port meter profile 2162~~~~~~~~~~~~~~~~~~~~~~~~~ 2163 2164Delete meter profile from the ethernet device:: 2165 2166 testpmd> del port meter profile (port_id) (profile_id) 2167 2168create port meter 2169~~~~~~~~~~~~~~~~~ 2170 2171Create new meter object for the ethernet device:: 2172 2173 testpmd> create port meter (port_id) (mtr_id) (profile_id) \ 2174 (meter_enable) (g_action) (y_action) (r_action) (stats_mask) (shared) \ 2175 (use_pre_meter_color) [(dscp_tbl_entry0) (dscp_tbl_entry1)...\ 2176 (dscp_tbl_entry63)] 2177 2178where: 2179 2180* ``mtr_id``: meter object ID. 2181* ``profile_id``: ID for the meter profile. 2182* ``meter_enable``: When this parameter has a non-zero value, the meter object 2183 gets enabled at the time of creation, otherwise remains disabled. 2184* ``g_action``: Policer action for the packet with green color. 2185* ``y_action``: Policer action for the packet with yellow color. 2186* ``r_action``: Policer action for the packet with red color. 2187* ``stats_mask``: Mask of statistics counter types to be enabled for the 2188 meter object. 2189* ``shared``: When this parameter has a non-zero value, the meter object is 2190 shared by multiple flows. Otherwise, meter object is used by single flow. 2191* ``use_pre_meter_color``: When this parameter has a non-zero value, the 2192 input color for the current meter object is determined by the latest meter 2193 object in the same flow. Otherwise, the current meter object uses the 2194 *dscp_table* to determine the input color. 2195* ``dscp_tbl_entryx``: DSCP table entry x providing meter providing input 2196 color, 0 <= x <= 63. 2197 2198enable port meter 2199~~~~~~~~~~~~~~~~~ 2200 2201Enable meter for the ethernet device:: 2202 2203 testpmd> enable port meter (port_id) (mtr_id) 2204 2205disable port meter 2206~~~~~~~~~~~~~~~~~~ 2207 2208Disable meter for the ethernet device:: 2209 2210 testpmd> disable port meter (port_id) (mtr_id) 2211 2212delete port meter 2213~~~~~~~~~~~~~~~~~ 2214 2215Delete meter for the ethernet device:: 2216 2217 testpmd> del port meter (port_id) (mtr_id) 2218 2219Set port meter profile 2220~~~~~~~~~~~~~~~~~~~~~~ 2221 2222Set meter profile for the ethernet device:: 2223 2224 testpmd> set port meter profile (port_id) (mtr_id) (profile_id) 2225 2226set port meter dscp table 2227~~~~~~~~~~~~~~~~~~~~~~~~~ 2228 2229Set meter dscp table for the ethernet device:: 2230 2231 testpmd> set port meter dscp table (port_id) (mtr_id) [(dscp_tbl_entry0) \ 2232 (dscp_tbl_entry1)...(dscp_tbl_entry63)] 2233 2234set port meter policer action 2235~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ 2236 2237Set meter policer action for the ethernet device:: 2238 2239 testpmd> set port meter policer action (port_id) (mtr_id) (action_mask) \ 2240 (action0) [(action1) (action1)] 2241 2242where: 2243 2244* ``action_mask``: Bit mask indicating which policer actions need to be 2245 updated. One or more policer actions can be updated in a single function 2246 invocation. To update the policer action associated with color C, bit 2247 (1 << C) needs to be set in *action_mask* and element at position C 2248 in the *actions* array needs to be valid. 2249* ``actionx``: Policer action for the color x, 2250 RTE_MTR_GREEN <= x < RTE_MTR_COLORS 2251 2252set port meter stats mask 2253~~~~~~~~~~~~~~~~~~~~~~~~~ 2254 2255Set meter stats mask for the ethernet device:: 2256 2257 testpmd> set port meter stats mask (port_id) (mtr_id) (stats_mask) 2258 2259where: 2260 2261* ``stats_mask``: Bit mask indicating statistics counter types to be enabled. 2262 2263show port meter stats 2264~~~~~~~~~~~~~~~~~~~~~ 2265 2266Show meter stats of the ethernet device:: 2267 2268 testpmd> show port meter stats (port_id) (mtr_id) (clear) 2269 2270where: 2271 2272* ``clear``: Flag that indicates whether the statistics counters should 2273 be cleared (i.e. set to zero) immediately after they have been read or not. 2274 2275Traffic Management 2276------------------ 2277 2278The following section shows functions for configuring traffic management on 2279on the ethernet device through the use of generic TM API. 2280 2281show port traffic management capability 2282~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ 2283 2284Show traffic management capability of the port:: 2285 2286 testpmd> show port tm cap (port_id) 2287 2288show port traffic management capability (hierarchy level) 2289~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ 2290 2291Show traffic management hierarchy level capability of the port:: 2292 2293 testpmd> show port tm level cap (port_id) (level_id) 2294 2295show port traffic management capability (hierarchy node level) 2296~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ 2297 2298Show the traffic management hierarchy node capability of the port:: 2299 2300 testpmd> show port tm node cap (port_id) (node_id) 2301 2302show port traffic management hierarchy node type 2303~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ 2304 2305Show the port traffic management hierarchy node type:: 2306 2307 testpmd> show port tm node type (port_id) (node_id) 2308 2309show port traffic management hierarchy node stats 2310~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ 2311 2312Show the port traffic management hierarchy node statistics:: 2313 2314 testpmd> show port tm node stats (port_id) (node_id) (clear) 2315 2316where: 2317 2318* ``clear``: When this parameter has a non-zero value, the statistics counters 2319 are cleared (i.e. set to zero) immediately after they have been read, 2320 otherwise the statistics counters are left untouched. 2321 2322Add port traffic management private shaper profile 2323~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ 2324 2325Add the port traffic management private shaper profile:: 2326 2327 testpmd> add port tm node shaper profile (port_id) (shaper_profile_id) \ 2328 (tb_rate) (tb_size) (packet_length_adjust) 2329 2330where: 2331 2332* ``shaper_profile id``: Shaper profile ID for the new profile. 2333* ``tb_rate``: Token bucket rate (bytes per second). 2334* ``tb_size``: Token bucket size (bytes). 2335* ``packet_length_adjust``: The value (bytes) to be added to the length of 2336 each packet for the purpose of shaping. This parameter value can be used to 2337 correct the packet length with the framing overhead bytes that are consumed 2338 on the wire. 2339 2340Delete port traffic management private shaper profile 2341~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ 2342 2343Delete the port traffic management private shaper:: 2344 2345 testpmd> del port tm node shaper profile (port_id) (shaper_profile_id) 2346 2347where: 2348 2349* ``shaper_profile id``: Shaper profile ID that needs to be deleted. 2350 2351Add port traffic management shared shaper 2352~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ 2353 2354Create the port traffic management shared shaper:: 2355 2356 testpmd> add port tm node shared shaper (port_id) (shared_shaper_id) \ 2357 (shaper_profile_id) 2358 2359where: 2360 2361* ``shared_shaper_id``: Shared shaper ID to be created. 2362* ``shaper_profile id``: Shaper profile ID for shared shaper. 2363 2364Set port traffic management shared shaper 2365~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ 2366 2367Update the port traffic management shared shaper:: 2368 2369 testpmd> set port tm node shared shaper (port_id) (shared_shaper_id) \ 2370 (shaper_profile_id) 2371 2372where: 2373 2374* ``shared_shaper_id``: Shared shaper ID to be update. 2375* ``shaper_profile id``: Shaper profile ID for shared shaper. 2376 2377Delete port traffic management shared shaper 2378~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ 2379 2380Delete the port traffic management shared shaper:: 2381 2382 testpmd> del port tm node shared shaper (port_id) (shared_shaper_id) 2383 2384where: 2385 2386* ``shared_shaper_id``: Shared shaper ID to be deleted. 2387 2388Set port traffic management hiearchy node private shaper 2389~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ 2390 2391set the port traffic management hierarchy node private shaper:: 2392 2393 testpmd> set port tm node shaper profile (port_id) (node_id) \ 2394 (shaper_profile_id) 2395 2396where: 2397 2398* ``shaper_profile id``: Private shaper profile ID to be enabled on the 2399 hierarchy node. 2400 2401Add port traffic management WRED profile 2402~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ 2403 2404Create a new WRED profile:: 2405 2406 testpmd> add port tm node wred profile (port_id) (wred_profile_id) \ 2407 (color_g) (min_th_g) (max_th_g) (maxp_inv_g) (wq_log2_g) \ 2408 (color_y) (min_th_y) (max_th_y) (maxp_inv_y) (wq_log2_y) \ 2409 (color_r) (min_th_r) (max_th_r) (maxp_inv_r) (wq_log2_r) 2410 2411where: 2412 2413* ``wred_profile id``: Identifier for the newly create WRED profile 2414* ``color_g``: Packet color (green) 2415* ``min_th_g``: Minimum queue threshold for packet with green color 2416* ``max_th_g``: Minimum queue threshold for packet with green color 2417* ``maxp_inv_g``: Inverse of packet marking probability maximum value (maxp) 2418* ``wq_log2_g``: Negated log2 of queue weight (wq) 2419* ``color_y``: Packet color (yellow) 2420* ``min_th_y``: Minimum queue threshold for packet with yellow color 2421* ``max_th_y``: Minimum queue threshold for packet with yellow color 2422* ``maxp_inv_y``: Inverse of packet marking probability maximum value (maxp) 2423* ``wq_log2_y``: Negated log2 of queue weight (wq) 2424* ``color_r``: Packet color (red) 2425* ``min_th_r``: Minimum queue threshold for packet with yellow color 2426* ``max_th_r``: Minimum queue threshold for packet with yellow color 2427* ``maxp_inv_r``: Inverse of packet marking probability maximum value (maxp) 2428* ``wq_log2_r``: Negated log2 of queue weight (wq) 2429 2430Delete port traffic management WRED profile 2431~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ 2432 2433Delete the WRED profile:: 2434 2435 testpmd> del port tm node wred profile (port_id) (wred_profile_id) 2436 2437Add port traffic management hierarchy nonleaf node 2438~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ 2439 2440Add nonleaf node to port traffic management hiearchy:: 2441 2442 testpmd> add port tm nonleaf node (port_id) (node_id) (parent_node_id) \ 2443 (priority) (weight) (level_id) (shaper_profile_id) \ 2444 (n_sp_priorities) (stats_mask) (n_shared_shapers) \ 2445 [(shared_shaper_0) (shared_shaper_1) ...] \ 2446 2447where: 2448 2449* ``parent_node_id``: Node ID of the parent. 2450* ``priority``: Node priority (highest node priority is zero). This is used by 2451 the SP algorithm running on the parent node for scheduling this node. 2452* ``weight``: Node weight (lowest weight is one). The node weight is relative 2453 to the weight sum of all siblings that have the same priority. It is used by 2454 the WFQ algorithm running on the parent node for scheduling this node. 2455* ``level_id``: Hiearchy level of the node. 2456* ``shaper_profile_id``: Shaper profile ID of the private shaper to be used by 2457 the node. 2458* ``n_sp_priorities``: Number of strict priorities. 2459* ``stats_mask``: Mask of statistics counter types to be enabled for this node. 2460* ``n_shared_shapers``: Number of shared shapers. 2461* ``shared_shaper_id``: Shared shaper id. 2462 2463Add port traffic management hierarchy leaf node 2464~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ 2465 2466Add leaf node to port traffic management hiearchy:: 2467 2468 testpmd> add port tm leaf node (port_id) (node_id) (parent_node_id) \ 2469 (priority) (weight) (level_id) (shaper_profile_id) \ 2470 (cman_mode) (wred_profile_id) (stats_mask) (n_shared_shapers) \ 2471 [(shared_shaper_id) (shared_shaper_id) ...] \ 2472 2473where: 2474 2475* ``parent_node_id``: Node ID of the parent. 2476* ``priority``: Node priority (highest node priority is zero). This is used by 2477 the SP algorithm running on the parent node for scheduling this node. 2478* ``weight``: Node weight (lowest weight is one). The node weight is relative 2479 to the weight sum of all siblings that have the same priority. It is used by 2480 the WFQ algorithm running on the parent node for scheduling this node. 2481* ``level_id``: Hiearchy level of the node. 2482* ``shaper_profile_id``: Shaper profile ID of the private shaper to be used by 2483 the node. 2484* ``cman_mode``: Congestion management mode to be enabled for this node. 2485* ``wred_profile_id``: WRED profile id to be enabled for this node. 2486* ``stats_mask``: Mask of statistics counter types to be enabled for this node. 2487* ``n_shared_shapers``: Number of shared shapers. 2488* ``shared_shaper_id``: Shared shaper id. 2489 2490Delete port traffic management hierarchy node 2491~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ 2492 2493Delete node from port traffic management hiearchy:: 2494 2495 testpmd> del port tm node (port_id) (node_id) 2496 2497Update port traffic management hierarchy parent node 2498~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ 2499 2500Update port traffic management hierarchy parent node:: 2501 2502 testpmd> set port tm node parent (port_id) (node_id) (parent_node_id) \ 2503 (priority) (weight) 2504 2505This function can only be called after the hierarchy commit invocation. Its 2506success depends on the port support for this operation, as advertised through 2507the port capability set. This function is valid for all nodes of the traffic 2508management hierarchy except root node. 2509 2510Suspend port traffic management hierarchy node 2511~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ 2512 2513 testpmd> suspend port tm node (port_id) (node_id) 2514 2515Commit port traffic management hierarchy 2516~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ 2517 2518Commit the traffic management hierarchy on the port:: 2519 2520 testpmd> port tm hierarchy commit (port_id) (clean_on_fail) 2521 2522where: 2523 2524* ``clean_on_fail``: When set to non-zero, hierarchy is cleared on function 2525 call failure. On the other hand, hierarchy is preserved when this parameter 2526 is equal to zero. 2527 2528Set port traffic management default hierarchy (tm forwarding mode) 2529~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ 2530 2531set the traffic management default hierarchy on the port:: 2532 2533 testpmd> set port tm hierarchy default (port_id) 2534 2535Filter Functions 2536---------------- 2537 2538This section details the available filter functions that are available. 2539 2540Note these functions interface the deprecated legacy filtering framework, 2541superseded by *rte_flow*. See `Flow rules management`_. 2542 2543ethertype_filter 2544~~~~~~~~~~~~~~~~~~~~ 2545 2546Add or delete a L2 Ethertype filter, which identify packets by their L2 Ethertype mainly assign them to a receive queue:: 2547 2548 ethertype_filter (port_id) (add|del) (mac_addr|mac_ignr) (mac_address) \ 2549 ethertype (ether_type) (drop|fwd) queue (queue_id) 2550 2551The available information parameters are: 2552 2553* ``port_id``: The port which the Ethertype filter assigned on. 2554 2555* ``mac_addr``: Compare destination mac address. 2556 2557* ``mac_ignr``: Ignore destination mac address match. 2558 2559* ``mac_address``: Destination mac address to match. 2560 2561* ``ether_type``: The EtherType value want to match, 2562 for example 0x0806 for ARP packet. 0x0800 (IPv4) and 0x86DD (IPv6) are invalid. 2563 2564* ``queue_id``: The receive queue associated with this EtherType filter. 2565 It is meaningless when deleting or dropping. 2566 2567Example, to add/remove an ethertype filter rule:: 2568 2569 testpmd> ethertype_filter 0 add mac_ignr 00:11:22:33:44:55 \ 2570 ethertype 0x0806 fwd queue 3 2571 2572 testpmd> ethertype_filter 0 del mac_ignr 00:11:22:33:44:55 \ 2573 ethertype 0x0806 fwd queue 3 2574 25752tuple_filter 2576~~~~~~~~~~~~~~~~~ 2577 2578Add or delete a 2-tuple filter, 2579which identifies packets by specific protocol and destination TCP/UDP port 2580and forwards packets into one of the receive queues:: 2581 2582 2tuple_filter (port_id) (add|del) dst_port (dst_port_value) \ 2583 protocol (protocol_value) mask (mask_value) \ 2584 tcp_flags (tcp_flags_value) priority (prio_value) \ 2585 queue (queue_id) 2586 2587The available information parameters are: 2588 2589* ``port_id``: The port which the 2-tuple filter assigned on. 2590 2591* ``dst_port_value``: Destination port in L4. 2592 2593* ``protocol_value``: IP L4 protocol. 2594 2595* ``mask_value``: Participates in the match or not by bit for field above, 1b means participate. 2596 2597* ``tcp_flags_value``: TCP control bits. The non-zero value is invalid, when the pro_value is not set to 0x06 (TCP). 2598 2599* ``prio_value``: Priority of this filter. 2600 2601* ``queue_id``: The receive queue associated with this 2-tuple filter. 2602 2603Example, to add/remove an 2tuple filter rule:: 2604 2605 testpmd> 2tuple_filter 0 add dst_port 32 protocol 0x06 mask 0x03 \ 2606 tcp_flags 0x02 priority 3 queue 3 2607 2608 testpmd> 2tuple_filter 0 del dst_port 32 protocol 0x06 mask 0x03 \ 2609 tcp_flags 0x02 priority 3 queue 3 2610 26115tuple_filter 2612~~~~~~~~~~~~~~~~~ 2613 2614Add or delete a 5-tuple filter, 2615which consists of a 5-tuple (protocol, source and destination IP addresses, source and destination TCP/UDP/SCTP port) 2616and routes packets into one of the receive queues:: 2617 2618 5tuple_filter (port_id) (add|del) dst_ip (dst_address) src_ip \ 2619 (src_address) dst_port (dst_port_value) \ 2620 src_port (src_port_value) protocol (protocol_value) \ 2621 mask (mask_value) tcp_flags (tcp_flags_value) \ 2622 priority (prio_value) queue (queue_id) 2623 2624The available information parameters are: 2625 2626* ``port_id``: The port which the 5-tuple filter assigned on. 2627 2628* ``dst_address``: Destination IP address. 2629 2630* ``src_address``: Source IP address. 2631 2632* ``dst_port_value``: TCP/UDP destination port. 2633 2634* ``src_port_value``: TCP/UDP source port. 2635 2636* ``protocol_value``: L4 protocol. 2637 2638* ``mask_value``: Participates in the match or not by bit for field above, 1b means participate 2639 2640* ``tcp_flags_value``: TCP control bits. The non-zero value is invalid, when the protocol_value is not set to 0x06 (TCP). 2641 2642* ``prio_value``: The priority of this filter. 2643 2644* ``queue_id``: The receive queue associated with this 5-tuple filter. 2645 2646Example, to add/remove an 5tuple filter rule:: 2647 2648 testpmd> 5tuple_filter 0 add dst_ip 2.2.2.5 src_ip 2.2.2.4 \ 2649 dst_port 64 src_port 32 protocol 0x06 mask 0x1F \ 2650 flags 0x0 priority 3 queue 3 2651 2652 testpmd> 5tuple_filter 0 del dst_ip 2.2.2.5 src_ip 2.2.2.4 \ 2653 dst_port 64 src_port 32 protocol 0x06 mask 0x1F \ 2654 flags 0x0 priority 3 queue 3 2655 2656syn_filter 2657~~~~~~~~~~ 2658 2659Using the SYN filter, TCP packets whose *SYN* flag is set can be forwarded to a separate queue:: 2660 2661 syn_filter (port_id) (add|del) priority (high|low) queue (queue_id) 2662 2663The available information parameters are: 2664 2665* ``port_id``: The port which the SYN filter assigned on. 2666 2667* ``high``: This SYN filter has higher priority than other filters. 2668 2669* ``low``: This SYN filter has lower priority than other filters. 2670 2671* ``queue_id``: The receive queue associated with this SYN filter 2672 2673Example:: 2674 2675 testpmd> syn_filter 0 add priority high queue 3 2676 2677flex_filter 2678~~~~~~~~~~~ 2679 2680With flex filter, packets can be recognized by any arbitrary pattern within the first 128 bytes of the packet 2681and routed into one of the receive queues:: 2682 2683 flex_filter (port_id) (add|del) len (len_value) bytes (bytes_value) \ 2684 mask (mask_value) priority (prio_value) queue (queue_id) 2685 2686The available information parameters are: 2687 2688* ``port_id``: The port which the Flex filter is assigned on. 2689 2690* ``len_value``: Filter length in bytes, no greater than 128. 2691 2692* ``bytes_value``: A string in hexadecimal, means the value the flex filter needs to match. 2693 2694* ``mask_value``: A string in hexadecimal, bit 1 means corresponding byte participates in the match. 2695 2696* ``prio_value``: The priority of this filter. 2697 2698* ``queue_id``: The receive queue associated with this Flex filter. 2699 2700Example:: 2701 2702 testpmd> flex_filter 0 add len 16 bytes 0x00000000000000000000000008060000 \ 2703 mask 000C priority 3 queue 3 2704 2705 testpmd> flex_filter 0 del len 16 bytes 0x00000000000000000000000008060000 \ 2706 mask 000C priority 3 queue 3 2707 2708 2709.. _testpmd_flow_director: 2710 2711flow_director_filter 2712~~~~~~~~~~~~~~~~~~~~ 2713 2714The Flow Director works in receive mode to identify specific flows or sets of flows and route them to specific queues. 2715 2716Four types of filtering are supported which are referred to as Perfect Match, Signature, Perfect-mac-vlan and 2717Perfect-tunnel filters, the match mode is set by the ``--pkt-filter-mode`` command-line parameter: 2718 2719* Perfect match filters. 2720 The hardware checks a match between the masked fields of the received packets and the programmed filters. 2721 The masked fields are for IP flow. 2722 2723* Signature filters. 2724 The hardware checks a match between a hash-based signature of the masked fields of the received packet. 2725 2726* Perfect-mac-vlan match filters. 2727 The hardware checks a match between the masked fields of the received packets and the programmed filters. 2728 The masked fields are for MAC VLAN flow. 2729 2730* Perfect-tunnel match filters. 2731 The hardware checks a match between the masked fields of the received packets and the programmed filters. 2732 The masked fields are for tunnel flow. 2733 2734* Perfect-raw-flow-type match filters. 2735 The hardware checks a match between the masked fields of the received packets and pre-loaded raw (template) packet. 2736 The masked fields are specified by input sets. 2737 2738The Flow Director filters can match the different fields for different type of packet: flow type, specific input set 2739per flow type and the flexible payload. 2740 2741The Flow Director can also mask out parts of all of these fields so that filters 2742are only applied to certain fields or parts of the fields. 2743 2744Note that for raw flow type mode the source and destination fields in the 2745raw packet buffer need to be presented in a reversed order with respect 2746to the expected received packets. 2747For example: IP source and destination addresses or TCP/UDP/SCTP 2748source and destination ports 2749 2750Different NICs may have different capabilities, command show port fdir (port_id) can be used to acquire the information. 2751 2752# Commands to add flow director filters of different flow types:: 2753 2754 flow_director_filter (port_id) mode IP (add|del|update) \ 2755 flow (ipv4-other|ipv4-frag|ipv6-other|ipv6-frag) \ 2756 src (src_ip_address) dst (dst_ip_address) \ 2757 tos (tos_value) proto (proto_value) ttl (ttl_value) \ 2758 vlan (vlan_value) flexbytes (flexbytes_value) \ 2759 (drop|fwd) pf|vf(vf_id) queue (queue_id) \ 2760 fd_id (fd_id_value) 2761 2762 flow_director_filter (port_id) mode IP (add|del|update) \ 2763 flow (ipv4-tcp|ipv4-udp|ipv6-tcp|ipv6-udp) \ 2764 src (src_ip_address) (src_port) \ 2765 dst (dst_ip_address) (dst_port) \ 2766 tos (tos_value) ttl (ttl_value) \ 2767 vlan (vlan_value) flexbytes (flexbytes_value) \ 2768 (drop|fwd) queue pf|vf(vf_id) (queue_id) \ 2769 fd_id (fd_id_value) 2770 2771 flow_director_filter (port_id) mode IP (add|del|update) \ 2772 flow (ipv4-sctp|ipv6-sctp) \ 2773 src (src_ip_address) (src_port) \ 2774 dst (dst_ip_address) (dst_port) \ 2775 tos (tos_value) ttl (ttl_value) \ 2776 tag (verification_tag) vlan (vlan_value) \ 2777 flexbytes (flexbytes_value) (drop|fwd) \ 2778 pf|vf(vf_id) queue (queue_id) fd_id (fd_id_value) 2779 2780 flow_director_filter (port_id) mode IP (add|del|update) flow l2_payload \ 2781 ether (ethertype) flexbytes (flexbytes_value) \ 2782 (drop|fwd) pf|vf(vf_id) queue (queue_id) 2783 fd_id (fd_id_value) 2784 2785 flow_director_filter (port_id) mode MAC-VLAN (add|del|update) \ 2786 mac (mac_address) vlan (vlan_value) \ 2787 flexbytes (flexbytes_value) (drop|fwd) \ 2788 queue (queue_id) fd_id (fd_id_value) 2789 2790 flow_director_filter (port_id) mode Tunnel (add|del|update) \ 2791 mac (mac_address) vlan (vlan_value) \ 2792 tunnel (NVGRE|VxLAN) tunnel-id (tunnel_id_value) \ 2793 flexbytes (flexbytes_value) (drop|fwd) \ 2794 queue (queue_id) fd_id (fd_id_value) 2795 2796 flow_director_filter (port_id) mode raw (add|del|update) flow (flow_id) \ 2797 (drop|fwd) queue (queue_id) fd_id (fd_id_value) \ 2798 packet (packet file name) 2799 2800For example, to add an ipv4-udp flow type filter:: 2801 2802 testpmd> flow_director_filter 0 mode IP add flow ipv4-udp src 2.2.2.3 32 \ 2803 dst 2.2.2.5 33 tos 2 ttl 40 vlan 0x1 flexbytes (0x88,0x48) \ 2804 fwd pf queue 1 fd_id 1 2805 2806For example, add an ipv4-other flow type filter:: 2807 2808 testpmd> flow_director_filter 0 mode IP add flow ipv4-other src 2.2.2.3 \ 2809 dst 2.2.2.5 tos 2 proto 20 ttl 40 vlan 0x1 \ 2810 flexbytes (0x88,0x48) fwd pf queue 1 fd_id 1 2811 2812flush_flow_director 2813~~~~~~~~~~~~~~~~~~~ 2814 2815Flush all flow director filters on a device:: 2816 2817 testpmd> flush_flow_director (port_id) 2818 2819Example, to flush all flow director filter on port 0:: 2820 2821 testpmd> flush_flow_director 0 2822 2823flow_director_mask 2824~~~~~~~~~~~~~~~~~~ 2825 2826Set flow director's input masks:: 2827 2828 flow_director_mask (port_id) mode IP vlan (vlan_value) \ 2829 src_mask (ipv4_src) (ipv6_src) (src_port) \ 2830 dst_mask (ipv4_dst) (ipv6_dst) (dst_port) 2831 2832 flow_director_mask (port_id) mode MAC-VLAN vlan (vlan_value) 2833 2834 flow_director_mask (port_id) mode Tunnel vlan (vlan_value) \ 2835 mac (mac_value) tunnel-type (tunnel_type_value) \ 2836 tunnel-id (tunnel_id_value) 2837 2838Example, to set flow director mask on port 0:: 2839 2840 testpmd> flow_director_mask 0 mode IP vlan 0xefff \ 2841 src_mask 255.255.255.255 \ 2842 FFFF:FFFF:FFFF:FFFF:FFFF:FFFF:FFFF:FFFF 0xFFFF \ 2843 dst_mask 255.255.255.255 \ 2844 FFFF:FFFF:FFFF:FFFF:FFFF:FFFF:FFFF:FFFF 0xFFFF 2845 2846flow_director_flex_mask 2847~~~~~~~~~~~~~~~~~~~~~~~ 2848 2849set masks of flow director's flexible payload based on certain flow type:: 2850 2851 testpmd> flow_director_flex_mask (port_id) \ 2852 flow (none|ipv4-other|ipv4-frag|ipv4-tcp|ipv4-udp|ipv4-sctp| \ 2853 ipv6-other|ipv6-frag|ipv6-tcp|ipv6-udp|ipv6-sctp| \ 2854 l2_payload|all) (mask) 2855 2856Example, to set flow director's flex mask for all flow type on port 0:: 2857 2858 testpmd> flow_director_flex_mask 0 flow all \ 2859 (0xff,0xff,0,0,0,0,0,0,0,0,0,0,0,0,0,0) 2860 2861 2862flow_director_flex_payload 2863~~~~~~~~~~~~~~~~~~~~~~~~~~ 2864 2865Configure flexible payload selection:: 2866 2867 flow_director_flex_payload (port_id) (raw|l2|l3|l4) (config) 2868 2869For example, to select the first 16 bytes from the offset 4 (bytes) of packet's payload as flexible payload:: 2870 2871 testpmd> flow_director_flex_payload 0 l4 \ 2872 (4,5,6,7,8,9,10,11,12,13,14,15,16,17,18,19) 2873 2874get_sym_hash_ena_per_port 2875~~~~~~~~~~~~~~~~~~~~~~~~~ 2876 2877Get symmetric hash enable configuration per port:: 2878 2879 get_sym_hash_ena_per_port (port_id) 2880 2881For example, to get symmetric hash enable configuration of port 1:: 2882 2883 testpmd> get_sym_hash_ena_per_port 1 2884 2885set_sym_hash_ena_per_port 2886~~~~~~~~~~~~~~~~~~~~~~~~~ 2887 2888Set symmetric hash enable configuration per port to enable or disable:: 2889 2890 set_sym_hash_ena_per_port (port_id) (enable|disable) 2891 2892For example, to set symmetric hash enable configuration of port 1 to enable:: 2893 2894 testpmd> set_sym_hash_ena_per_port 1 enable 2895 2896get_hash_global_config 2897~~~~~~~~~~~~~~~~~~~~~~ 2898 2899Get the global configurations of hash filters:: 2900 2901 get_hash_global_config (port_id) 2902 2903For example, to get the global configurations of hash filters of port 1:: 2904 2905 testpmd> get_hash_global_config 1 2906 2907set_hash_global_config 2908~~~~~~~~~~~~~~~~~~~~~~ 2909 2910Set the global configurations of hash filters:: 2911 2912 set_hash_global_config (port_id) (toeplitz|simple_xor|default) \ 2913 (ipv4|ipv4-frag|ipv4-tcp|ipv4-udp|ipv4-sctp|ipv4-other|ipv6|ipv6-frag| \ 2914 ipv6-tcp|ipv6-udp|ipv6-sctp|ipv6-other|l2_payload|<flow_id>) \ 2915 (enable|disable) 2916 2917For example, to enable simple_xor for flow type of ipv6 on port 2:: 2918 2919 testpmd> set_hash_global_config 2 simple_xor ipv6 enable 2920 2921set_hash_input_set 2922~~~~~~~~~~~~~~~~~~ 2923 2924Set the input set for hash:: 2925 2926 set_hash_input_set (port_id) (ipv4-frag|ipv4-tcp|ipv4-udp|ipv4-sctp| \ 2927 ipv4-other|ipv6-frag|ipv6-tcp|ipv6-udp|ipv6-sctp|ipv6-other| \ 2928 l2_payload|<flow_id>) (ovlan|ivlan|src-ipv4|dst-ipv4|src-ipv6|dst-ipv6| \ 2929 ipv4-tos|ipv4-proto|ipv6-tc|ipv6-next-header|udp-src-port|udp-dst-port| \ 2930 tcp-src-port|tcp-dst-port|sctp-src-port|sctp-dst-port|sctp-veri-tag| \ 2931 udp-key|gre-key|fld-1st|fld-2nd|fld-3rd|fld-4th|fld-5th|fld-6th|fld-7th| \ 2932 fld-8th|none) (select|add) 2933 2934For example, to add source IP to hash input set for flow type of ipv4-udp on port 0:: 2935 2936 testpmd> set_hash_input_set 0 ipv4-udp src-ipv4 add 2937 2938set_fdir_input_set 2939~~~~~~~~~~~~~~~~~~ 2940 2941The Flow Director filters can match the different fields for different type of packet, i.e. specific input set 2942on per flow type and the flexible payload. This command can be used to change input set for each flow type. 2943 2944Set the input set for flow director:: 2945 2946 set_fdir_input_set (port_id) (ipv4-frag|ipv4-tcp|ipv4-udp|ipv4-sctp| \ 2947 ipv4-other|ipv6|ipv6-frag|ipv6-tcp|ipv6-udp|ipv6-sctp|ipv6-other| \ 2948 l2_payload|<flow_id>) (ivlan|ethertype|src-ipv4|dst-ipv4|src-ipv6|dst-ipv6| \ 2949 ipv4-tos|ipv4-proto|ipv4-ttl|ipv6-tc|ipv6-next-header|ipv6-hop-limits| \ 2950 tudp-src-port|udp-dst-port|cp-src-port|tcp-dst-port|sctp-src-port| \ 2951 sctp-dst-port|sctp-veri-tag|none) (select|add) 2952 2953For example to add source IP to FD input set for flow type of ipv4-udp on port 0:: 2954 2955 testpmd> set_fdir_input_set 0 ipv4-udp src-ipv4 add 2956 2957global_config 2958~~~~~~~~~~~~~ 2959 2960Set different GRE key length for input set:: 2961 2962 global_config (port_id) gre-key-len (number in bytes) 2963 2964For example to set GRE key length for input set to 4 bytes on port 0:: 2965 2966 testpmd> global_config 0 gre-key-len 4 2967 2968 2969.. _testpmd_rte_flow: 2970 2971Flow rules management 2972--------------------- 2973 2974Control of the generic flow API (*rte_flow*) is fully exposed through the 2975``flow`` command (validation, creation, destruction, queries and operation 2976modes). 2977 2978Considering *rte_flow* overlaps with all `Filter Functions`_, using both 2979features simultaneously may cause undefined side-effects and is therefore 2980not recommended. 2981 2982``flow`` syntax 2983~~~~~~~~~~~~~~~ 2984 2985Because the ``flow`` command uses dynamic tokens to handle the large number 2986of possible flow rules combinations, its behavior differs slightly from 2987other commands, in particular: 2988 2989- Pressing *?* or the *<tab>* key displays contextual help for the current 2990 token, not that of the entire command. 2991 2992- Optional and repeated parameters are supported (provided they are listed 2993 in the contextual help). 2994 2995The first parameter stands for the operation mode. Possible operations and 2996their general syntax are described below. They are covered in detail in the 2997following sections. 2998 2999- Check whether a flow rule can be created:: 3000 3001 flow validate {port_id} 3002 [group {group_id}] [priority {level}] [ingress] [egress] [transfer] 3003 pattern {item} [/ {item} [...]] / end 3004 actions {action} [/ {action} [...]] / end 3005 3006- Create a flow rule:: 3007 3008 flow create {port_id} 3009 [group {group_id}] [priority {level}] [ingress] [egress] [transfer] 3010 pattern {item} [/ {item} [...]] / end 3011 actions {action} [/ {action} [...]] / end 3012 3013- Destroy specific flow rules:: 3014 3015 flow destroy {port_id} rule {rule_id} [...] 3016 3017- Destroy all flow rules:: 3018 3019 flow flush {port_id} 3020 3021- Query an existing flow rule:: 3022 3023 flow query {port_id} {rule_id} {action} 3024 3025- List existing flow rules sorted by priority, filtered by group 3026 identifiers:: 3027 3028 flow list {port_id} [group {group_id}] [...] 3029 3030- Restrict ingress traffic to the defined flow rules:: 3031 3032 flow isolate {port_id} {boolean} 3033 3034Validating flow rules 3035~~~~~~~~~~~~~~~~~~~~~ 3036 3037``flow validate`` reports whether a flow rule would be accepted by the 3038underlying device in its current state but stops short of creating it. It is 3039bound to ``rte_flow_validate()``:: 3040 3041 flow validate {port_id} 3042 [group {group_id}] [priority {level}] [ingress] [egress] [transfer] 3043 pattern {item} [/ {item} [...]] / end 3044 actions {action} [/ {action} [...]] / end 3045 3046If successful, it will show:: 3047 3048 Flow rule validated 3049 3050Otherwise it will show an error message of the form:: 3051 3052 Caught error type [...] ([...]): [...] 3053 3054This command uses the same parameters as ``flow create``, their format is 3055described in `Creating flow rules`_. 3056 3057Check whether redirecting any Ethernet packet received on port 0 to RX queue 3058index 6 is supported:: 3059 3060 testpmd> flow validate 0 ingress pattern eth / end 3061 actions queue index 6 / end 3062 Flow rule validated 3063 testpmd> 3064 3065Port 0 does not support TCPv6 rules:: 3066 3067 testpmd> flow validate 0 ingress pattern eth / ipv6 / tcp / end 3068 actions drop / end 3069 Caught error type 9 (specific pattern item): Invalid argument 3070 testpmd> 3071 3072Creating flow rules 3073~~~~~~~~~~~~~~~~~~~ 3074 3075``flow create`` validates and creates the specified flow rule. It is bound 3076to ``rte_flow_create()``:: 3077 3078 flow create {port_id} 3079 [group {group_id}] [priority {level}] [ingress] [egress] [transfer] 3080 pattern {item} [/ {item} [...]] / end 3081 actions {action} [/ {action} [...]] / end 3082 3083If successful, it will return a flow rule ID usable with other commands:: 3084 3085 Flow rule #[...] created 3086 3087Otherwise it will show an error message of the form:: 3088 3089 Caught error type [...] ([...]): [...] 3090 3091Parameters describe in the following order: 3092 3093- Attributes (*group*, *priority*, *ingress*, *egress*, *transfer* tokens). 3094- A matching pattern, starting with the *pattern* token and terminated by an 3095 *end* pattern item. 3096- Actions, starting with the *actions* token and terminated by an *end* 3097 action. 3098 3099These translate directly to *rte_flow* objects provided as-is to the 3100underlying functions. 3101 3102The shortest valid definition only comprises mandatory tokens:: 3103 3104 testpmd> flow create 0 pattern end actions end 3105 3106Note that PMDs may refuse rules that essentially do nothing such as this 3107one. 3108 3109**All unspecified object values are automatically initialized to 0.** 3110 3111Attributes 3112^^^^^^^^^^ 3113 3114These tokens affect flow rule attributes (``struct rte_flow_attr``) and are 3115specified before the ``pattern`` token. 3116 3117- ``group {group id}``: priority group. 3118- ``priority {level}``: priority level within group. 3119- ``ingress``: rule applies to ingress traffic. 3120- ``egress``: rule applies to egress traffic. 3121- ``transfer``: apply rule directly to endpoints found in pattern. 3122 3123Each instance of an attribute specified several times overrides the previous 3124value as shown below (group 4 is used):: 3125 3126 testpmd> flow create 0 group 42 group 24 group 4 [...] 3127 3128Note that once enabled, ``ingress`` and ``egress`` cannot be disabled. 3129 3130While not specifying a direction is an error, some rules may allow both 3131simultaneously. 3132 3133Most rules affect RX therefore contain the ``ingress`` token:: 3134 3135 testpmd> flow create 0 ingress pattern [...] 3136 3137Matching pattern 3138^^^^^^^^^^^^^^^^ 3139 3140A matching pattern starts after the ``pattern`` token. It is made of pattern 3141items and is terminated by a mandatory ``end`` item. 3142 3143Items are named after their type (*RTE_FLOW_ITEM_TYPE_* from ``enum 3144rte_flow_item_type``). 3145 3146The ``/`` token is used as a separator between pattern items as shown 3147below:: 3148 3149 testpmd> flow create 0 ingress pattern eth / ipv4 / udp / end [...] 3150 3151Note that protocol items like these must be stacked from lowest to highest 3152layer to make sense. For instance, the following rule is either invalid or 3153unlikely to match any packet:: 3154 3155 testpmd> flow create 0 ingress pattern eth / udp / ipv4 / end [...] 3156 3157More information on these restrictions can be found in the *rte_flow* 3158documentation. 3159 3160Several items support additional specification structures, for example 3161``ipv4`` allows specifying source and destination addresses as follows:: 3162 3163 testpmd> flow create 0 ingress pattern eth / ipv4 src is 10.1.1.1 3164 dst is 10.2.0.0 / end [...] 3165 3166This rule matches all IPv4 traffic with the specified properties. 3167 3168In this example, ``src`` and ``dst`` are field names of the underlying 3169``struct rte_flow_item_ipv4`` object. All item properties can be specified 3170in a similar fashion. 3171 3172The ``is`` token means that the subsequent value must be matched exactly, 3173and assigns ``spec`` and ``mask`` fields in ``struct rte_flow_item`` 3174accordingly. Possible assignment tokens are: 3175 3176- ``is``: match value perfectly (with full bit-mask). 3177- ``spec``: match value according to configured bit-mask. 3178- ``last``: specify upper bound to establish a range. 3179- ``mask``: specify bit-mask with relevant bits set to one. 3180- ``prefix``: generate bit-mask from a prefix length. 3181 3182These yield identical results:: 3183 3184 ipv4 src is 10.1.1.1 3185 3186:: 3187 3188 ipv4 src spec 10.1.1.1 src mask 255.255.255.255 3189 3190:: 3191 3192 ipv4 src spec 10.1.1.1 src prefix 32 3193 3194:: 3195 3196 ipv4 src is 10.1.1.1 src last 10.1.1.1 # range with a single value 3197 3198:: 3199 3200 ipv4 src is 10.1.1.1 src last 0 # 0 disables range 3201 3202Inclusive ranges can be defined with ``last``:: 3203 3204 ipv4 src is 10.1.1.1 src last 10.2.3.4 # 10.1.1.1 to 10.2.3.4 3205 3206Note that ``mask`` affects both ``spec`` and ``last``:: 3207 3208 ipv4 src is 10.1.1.1 src last 10.2.3.4 src mask 255.255.0.0 3209 # matches 10.1.0.0 to 10.2.255.255 3210 3211Properties can be modified multiple times:: 3212 3213 ipv4 src is 10.1.1.1 src is 10.1.2.3 src is 10.2.3.4 # matches 10.2.3.4 3214 3215:: 3216 3217 ipv4 src is 10.1.1.1 src prefix 24 src prefix 16 # matches 10.1.0.0/16 3218 3219Pattern items 3220^^^^^^^^^^^^^ 3221 3222This section lists supported pattern items and their attributes, if any. 3223 3224- ``end``: end list of pattern items. 3225 3226- ``void``: no-op pattern item. 3227 3228- ``invert``: perform actions when pattern does not match. 3229 3230- ``any``: match any protocol for the current layer. 3231 3232 - ``num {unsigned}``: number of layers covered. 3233 3234- ``pf``: match traffic from/to the physical function. 3235 3236- ``vf``: match traffic from/to a virtual function ID. 3237 3238 - ``id {unsigned}``: VF ID. 3239 3240- ``phy_port``: match traffic from/to a specific physical port. 3241 3242 - ``index {unsigned}``: physical port index. 3243 3244- ``port_id``: match traffic from/to a given DPDK port ID. 3245 3246 - ``id {unsigned}``: DPDK port ID. 3247 3248- ``mark``: match value set in previously matched flow rule using the mark action. 3249 3250 - ``id {unsigned}``: arbitrary integer value. 3251 3252- ``raw``: match an arbitrary byte string. 3253 3254 - ``relative {boolean}``: look for pattern after the previous item. 3255 - ``search {boolean}``: search pattern from offset (see also limit). 3256 - ``offset {integer}``: absolute or relative offset for pattern. 3257 - ``limit {unsigned}``: search area limit for start of pattern. 3258 - ``pattern {string}``: byte string to look for. 3259 3260- ``eth``: match Ethernet header. 3261 3262 - ``dst {MAC-48}``: destination MAC. 3263 - ``src {MAC-48}``: source MAC. 3264 - ``type {unsigned}``: EtherType or TPID. 3265 3266- ``vlan``: match 802.1Q/ad VLAN tag. 3267 3268 - ``tci {unsigned}``: tag control information. 3269 - ``pcp {unsigned}``: priority code point. 3270 - ``dei {unsigned}``: drop eligible indicator. 3271 - ``vid {unsigned}``: VLAN identifier. 3272 - ``inner_type {unsigned}``: inner EtherType or TPID. 3273 3274- ``ipv4``: match IPv4 header. 3275 3276 - ``tos {unsigned}``: type of service. 3277 - ``ttl {unsigned}``: time to live. 3278 - ``proto {unsigned}``: next protocol ID. 3279 - ``src {ipv4 address}``: source address. 3280 - ``dst {ipv4 address}``: destination address. 3281 3282- ``ipv6``: match IPv6 header. 3283 3284 - ``tc {unsigned}``: traffic class. 3285 - ``flow {unsigned}``: flow label. 3286 - ``proto {unsigned}``: protocol (next header). 3287 - ``hop {unsigned}``: hop limit. 3288 - ``src {ipv6 address}``: source address. 3289 - ``dst {ipv6 address}``: destination address. 3290 3291- ``icmp``: match ICMP header. 3292 3293 - ``type {unsigned}``: ICMP packet type. 3294 - ``code {unsigned}``: ICMP packet code. 3295 3296- ``udp``: match UDP header. 3297 3298 - ``src {unsigned}``: UDP source port. 3299 - ``dst {unsigned}``: UDP destination port. 3300 3301- ``tcp``: match TCP header. 3302 3303 - ``src {unsigned}``: TCP source port. 3304 - ``dst {unsigned}``: TCP destination port. 3305 3306- ``sctp``: match SCTP header. 3307 3308 - ``src {unsigned}``: SCTP source port. 3309 - ``dst {unsigned}``: SCTP destination port. 3310 - ``tag {unsigned}``: validation tag. 3311 - ``cksum {unsigned}``: checksum. 3312 3313- ``vxlan``: match VXLAN header. 3314 3315 - ``vni {unsigned}``: VXLAN identifier. 3316 3317- ``e_tag``: match IEEE 802.1BR E-Tag header. 3318 3319 - ``grp_ecid_b {unsigned}``: GRP and E-CID base. 3320 3321- ``nvgre``: match NVGRE header. 3322 3323 - ``tni {unsigned}``: virtual subnet ID. 3324 3325- ``mpls``: match MPLS header. 3326 3327 - ``label {unsigned}``: MPLS label. 3328 3329- ``gre``: match GRE header. 3330 3331 - ``protocol {unsigned}``: protocol type. 3332 3333- ``fuzzy``: fuzzy pattern match, expect faster than default. 3334 3335 - ``thresh {unsigned}``: accuracy threshold. 3336 3337- ``gtp``, ``gtpc``, ``gtpu``: match GTPv1 header. 3338 3339 - ``teid {unsigned}``: tunnel endpoint identifier. 3340 3341- ``geneve``: match GENEVE header. 3342 3343 - ``vni {unsigned}``: virtual network identifier. 3344 - ``protocol {unsigned}``: protocol type. 3345 3346- ``vxlan-gpe``: match VXLAN-GPE header. 3347 3348 - ``vni {unsigned}``: VXLAN-GPE identifier. 3349 3350- ``arp_eth_ipv4``: match ARP header for Ethernet/IPv4. 3351 3352 - ``sha {MAC-48}``: sender hardware address. 3353 - ``spa {ipv4 address}``: sender IPv4 address. 3354 - ``tha {MAC-48}``: target hardware address. 3355 - ``tpa {ipv4 address}``: target IPv4 address. 3356 3357- ``ipv6_ext``: match presence of any IPv6 extension header. 3358 3359 - ``next_hdr {unsigned}``: next header. 3360 3361- ``icmp6``: match any ICMPv6 header. 3362 3363 - ``type {unsigned}``: ICMPv6 type. 3364 - ``code {unsigned}``: ICMPv6 code. 3365 3366- ``icmp6_nd_ns``: match ICMPv6 neighbor discovery solicitation. 3367 3368 - ``target_addr {ipv6 address}``: target address. 3369 3370- ``icmp6_nd_na``: match ICMPv6 neighbor discovery advertisement. 3371 3372 - ``target_addr {ipv6 address}``: target address. 3373 3374- ``icmp6_nd_opt``: match presence of any ICMPv6 neighbor discovery option. 3375 3376 - ``type {unsigned}``: ND option type. 3377 3378- ``icmp6_nd_opt_sla_eth``: match ICMPv6 neighbor discovery source Ethernet 3379 link-layer address option. 3380 3381 - ``sla {MAC-48}``: source Ethernet LLA. 3382 3383- ``icmp6_nd_opt_sla_eth``: match ICMPv6 neighbor discovery target Ethernet 3384 link-layer address option. 3385 3386 - ``tla {MAC-48}``: target Ethernet LLA. 3387 3388Actions list 3389^^^^^^^^^^^^ 3390 3391A list of actions starts after the ``actions`` token in the same fashion as 3392`Matching pattern`_; actions are separated by ``/`` tokens and the list is 3393terminated by a mandatory ``end`` action. 3394 3395Actions are named after their type (*RTE_FLOW_ACTION_TYPE_* from ``enum 3396rte_flow_action_type``). 3397 3398Dropping all incoming UDPv4 packets can be expressed as follows:: 3399 3400 testpmd> flow create 0 ingress pattern eth / ipv4 / udp / end 3401 actions drop / end 3402 3403Several actions have configurable properties which must be specified when 3404there is no valid default value. For example, ``queue`` requires a target 3405queue index. 3406 3407This rule redirects incoming UDPv4 traffic to queue index 6:: 3408 3409 testpmd> flow create 0 ingress pattern eth / ipv4 / udp / end 3410 actions queue index 6 / end 3411 3412While this one could be rejected by PMDs (unspecified queue index):: 3413 3414 testpmd> flow create 0 ingress pattern eth / ipv4 / udp / end 3415 actions queue / end 3416 3417As defined by *rte_flow*, the list is not ordered, all actions of a given 3418rule are performed simultaneously. These are equivalent:: 3419 3420 queue index 6 / void / mark id 42 / end 3421 3422:: 3423 3424 void / mark id 42 / queue index 6 / end 3425 3426All actions in a list should have different types, otherwise only the last 3427action of a given type is taken into account:: 3428 3429 queue index 4 / queue index 5 / queue index 6 / end # will use queue 6 3430 3431:: 3432 3433 drop / drop / drop / end # drop is performed only once 3434 3435:: 3436 3437 mark id 42 / queue index 3 / mark id 24 / end # mark will be 24 3438 3439Considering they are performed simultaneously, opposite and overlapping 3440actions can sometimes be combined when the end result is unambiguous:: 3441 3442 drop / queue index 6 / end # drop has no effect 3443 3444:: 3445 3446 queue index 6 / rss queues 6 7 8 / end # queue has no effect 3447 3448:: 3449 3450 drop / passthru / end # drop has no effect 3451 3452Note that PMDs may still refuse such combinations. 3453 3454Actions 3455^^^^^^^ 3456 3457This section lists supported actions and their attributes, if any. 3458 3459- ``end``: end list of actions. 3460 3461- ``void``: no-op action. 3462 3463- ``passthru``: let subsequent rule process matched packets. 3464 3465- ``jump``: redirect traffic to group on device. 3466 3467 - ``group {unsigned}``: group to redirect to. 3468 3469- ``mark``: attach 32 bit value to packets. 3470 3471 - ``id {unsigned}``: 32 bit value to return with packets. 3472 3473- ``flag``: flag packets. 3474 3475- ``queue``: assign packets to a given queue index. 3476 3477 - ``index {unsigned}``: queue index to use. 3478 3479- ``drop``: drop packets (note: passthru has priority). 3480 3481- ``count``: enable counters for this rule. 3482 3483- ``rss``: spread packets among several queues. 3484 3485 - ``func {hash function}``: RSS hash function to apply, allowed tokens are 3486 the same as `set_hash_global_config`_. 3487 3488 - ``level {unsigned}``: encapsulation level for ``types``. 3489 3490 - ``types [{RSS hash type} [...]] end``: specific RSS hash types, allowed 3491 tokens are the same as `set_hash_input_set`_, except that an empty list 3492 does not disable RSS but instead requests unspecified "best-effort" 3493 settings. 3494 3495 - ``key {string}``: RSS hash key, overrides ``key_len``. 3496 3497 - ``key_len {unsigned}``: RSS hash key length in bytes, can be used in 3498 conjunction with ``key`` to pad or truncate it. 3499 3500 - ``queues [{unsigned} [...]] end``: queue indices to use. 3501 3502- ``pf``: direct traffic to physical function. 3503 3504- ``vf``: direct traffic to a virtual function ID. 3505 3506 - ``original {boolean}``: use original VF ID if possible. 3507 - ``id {unsigned}``: VF ID. 3508 3509- ``phy_port``: direct packets to physical port index. 3510 3511 - ``original {boolean}``: use original port index if possible. 3512 - ``index {unsigned}``: physical port index. 3513 3514- ``port_id``: direct matching traffic to a given DPDK port ID. 3515 3516 - ``original {boolean}``: use original DPDK port ID if possible. 3517 - ``id {unsigned}``: DPDK port ID. 3518 3519- ``of_set_mpls_ttl``: OpenFlow's ``OFPAT_SET_MPLS_TTL``. 3520 3521 - ``mpls_ttl``: MPLS TTL. 3522 3523- ``of_dec_mpls_ttl``: OpenFlow's ``OFPAT_DEC_MPLS_TTL``. 3524 3525- ``of_set_nw_ttl``: OpenFlow's ``OFPAT_SET_NW_TTL``. 3526 3527 - ``nw_ttl``: IP TTL. 3528 3529- ``of_dec_nw_ttl``: OpenFlow's ``OFPAT_DEC_NW_TTL``. 3530 3531- ``of_copy_ttl_out``: OpenFlow's ``OFPAT_COPY_TTL_OUT``. 3532 3533- ``of_copy_ttl_in``: OpenFlow's ``OFPAT_COPY_TTL_IN``. 3534 3535- ``of_pop_vlan``: OpenFlow's ``OFPAT_POP_VLAN``. 3536 3537- ``of_push_vlan``: OpenFlow's ``OFPAT_PUSH_VLAN``. 3538 3539 - ``ethertype``: Ethertype. 3540 3541- ``of_set_vlan_vid``: OpenFlow's ``OFPAT_SET_VLAN_VID``. 3542 3543 - ``vlan_vid``: VLAN id. 3544 3545- ``of_set_vlan_pcp``: OpenFlow's ``OFPAT_SET_VLAN_PCP``. 3546 3547 - ``vlan_pcp``: VLAN priority. 3548 3549- ``of_pop_mpls``: OpenFlow's ``OFPAT_POP_MPLS``. 3550 3551 - ``ethertype``: Ethertype. 3552 3553- ``of_push_mpls``: OpenFlow's ``OFPAT_PUSH_MPLS``. 3554 3555 - ``ethertype``: Ethertype. 3556 3557Destroying flow rules 3558~~~~~~~~~~~~~~~~~~~~~ 3559 3560``flow destroy`` destroys one or more rules from their rule ID (as returned 3561by ``flow create``), this command calls ``rte_flow_destroy()`` as many 3562times as necessary:: 3563 3564 flow destroy {port_id} rule {rule_id} [...] 3565 3566If successful, it will show:: 3567 3568 Flow rule #[...] destroyed 3569 3570It does not report anything for rule IDs that do not exist. The usual error 3571message is shown when a rule cannot be destroyed:: 3572 3573 Caught error type [...] ([...]): [...] 3574 3575``flow flush`` destroys all rules on a device and does not take extra 3576arguments. It is bound to ``rte_flow_flush()``:: 3577 3578 flow flush {port_id} 3579 3580Any errors are reported as above. 3581 3582Creating several rules and destroying them:: 3583 3584 testpmd> flow create 0 ingress pattern eth / ipv6 / end 3585 actions queue index 2 / end 3586 Flow rule #0 created 3587 testpmd> flow create 0 ingress pattern eth / ipv4 / end 3588 actions queue index 3 / end 3589 Flow rule #1 created 3590 testpmd> flow destroy 0 rule 0 rule 1 3591 Flow rule #1 destroyed 3592 Flow rule #0 destroyed 3593 testpmd> 3594 3595The same result can be achieved using ``flow flush``:: 3596 3597 testpmd> flow create 0 ingress pattern eth / ipv6 / end 3598 actions queue index 2 / end 3599 Flow rule #0 created 3600 testpmd> flow create 0 ingress pattern eth / ipv4 / end 3601 actions queue index 3 / end 3602 Flow rule #1 created 3603 testpmd> flow flush 0 3604 testpmd> 3605 3606Non-existent rule IDs are ignored:: 3607 3608 testpmd> flow create 0 ingress pattern eth / ipv6 / end 3609 actions queue index 2 / end 3610 Flow rule #0 created 3611 testpmd> flow create 0 ingress pattern eth / ipv4 / end 3612 actions queue index 3 / end 3613 Flow rule #1 created 3614 testpmd> flow destroy 0 rule 42 rule 10 rule 2 3615 testpmd> 3616 testpmd> flow destroy 0 rule 0 3617 Flow rule #0 destroyed 3618 testpmd> 3619 3620Querying flow rules 3621~~~~~~~~~~~~~~~~~~~ 3622 3623``flow query`` queries a specific action of a flow rule having that 3624ability. Such actions collect information that can be reported using this 3625command. It is bound to ``rte_flow_query()``:: 3626 3627 flow query {port_id} {rule_id} {action} 3628 3629If successful, it will display either the retrieved data for known actions 3630or the following message:: 3631 3632 Cannot display result for action type [...] ([...]) 3633 3634Otherwise, it will complain either that the rule does not exist or that some 3635error occurred:: 3636 3637 Flow rule #[...] not found 3638 3639:: 3640 3641 Caught error type [...] ([...]): [...] 3642 3643Currently only the ``count`` action is supported. This action reports the 3644number of packets that hit the flow rule and the total number of bytes. Its 3645output has the following format:: 3646 3647 count: 3648 hits_set: [...] # whether "hits" contains a valid value 3649 bytes_set: [...] # whether "bytes" contains a valid value 3650 hits: [...] # number of packets 3651 bytes: [...] # number of bytes 3652 3653Querying counters for TCPv6 packets redirected to queue 6:: 3654 3655 testpmd> flow create 0 ingress pattern eth / ipv6 / tcp / end 3656 actions queue index 6 / count / end 3657 Flow rule #4 created 3658 testpmd> flow query 0 4 count 3659 count: 3660 hits_set: 1 3661 bytes_set: 0 3662 hits: 386446 3663 bytes: 0 3664 testpmd> 3665 3666Listing flow rules 3667~~~~~~~~~~~~~~~~~~ 3668 3669``flow list`` lists existing flow rules sorted by priority and optionally 3670filtered by group identifiers:: 3671 3672 flow list {port_id} [group {group_id}] [...] 3673 3674This command only fails with the following message if the device does not 3675exist:: 3676 3677 Invalid port [...] 3678 3679Output consists of a header line followed by a short description of each 3680flow rule, one per line. There is no output at all when no flow rules are 3681configured on the device:: 3682 3683 ID Group Prio Attr Rule 3684 [...] [...] [...] [...] [...] 3685 3686``Attr`` column flags: 3687 3688- ``i`` for ``ingress``. 3689- ``e`` for ``egress``. 3690 3691Creating several flow rules and listing them:: 3692 3693 testpmd> flow create 0 ingress pattern eth / ipv4 / end 3694 actions queue index 6 / end 3695 Flow rule #0 created 3696 testpmd> flow create 0 ingress pattern eth / ipv6 / end 3697 actions queue index 2 / end 3698 Flow rule #1 created 3699 testpmd> flow create 0 priority 5 ingress pattern eth / ipv4 / udp / end 3700 actions rss queues 6 7 8 end / end 3701 Flow rule #2 created 3702 testpmd> flow list 0 3703 ID Group Prio Attr Rule 3704 0 0 0 i- ETH IPV4 => QUEUE 3705 1 0 0 i- ETH IPV6 => QUEUE 3706 2 0 5 i- ETH IPV4 UDP => RSS 3707 testpmd> 3708 3709Rules are sorted by priority (i.e. group ID first, then priority level):: 3710 3711 testpmd> flow list 1 3712 ID Group Prio Attr Rule 3713 0 0 0 i- ETH => COUNT 3714 6 0 500 i- ETH IPV6 TCP => DROP COUNT 3715 5 0 1000 i- ETH IPV6 ICMP => QUEUE 3716 1 24 0 i- ETH IPV4 UDP => QUEUE 3717 4 24 10 i- ETH IPV4 TCP => DROP 3718 3 24 20 i- ETH IPV4 => DROP 3719 2 24 42 i- ETH IPV4 UDP => QUEUE 3720 7 63 0 i- ETH IPV6 UDP VXLAN => MARK QUEUE 3721 testpmd> 3722 3723Output can be limited to specific groups:: 3724 3725 testpmd> flow list 1 group 0 group 63 3726 ID Group Prio Attr Rule 3727 0 0 0 i- ETH => COUNT 3728 6 0 500 i- ETH IPV6 TCP => DROP COUNT 3729 5 0 1000 i- ETH IPV6 ICMP => QUEUE 3730 7 63 0 i- ETH IPV6 UDP VXLAN => MARK QUEUE 3731 testpmd> 3732 3733Toggling isolated mode 3734~~~~~~~~~~~~~~~~~~~~~~ 3735 3736``flow isolate`` can be used to tell the underlying PMD that ingress traffic 3737must only be injected from the defined flow rules; that no default traffic 3738is expected outside those rules and the driver is free to assign more 3739resources to handle them. It is bound to ``rte_flow_isolate()``:: 3740 3741 flow isolate {port_id} {boolean} 3742 3743If successful, enabling or disabling isolated mode shows either:: 3744 3745 Ingress traffic on port [...] 3746 is now restricted to the defined flow rules 3747 3748Or:: 3749 3750 Ingress traffic on port [...] 3751 is not restricted anymore to the defined flow rules 3752 3753Otherwise, in case of error:: 3754 3755 Caught error type [...] ([...]): [...] 3756 3757Mainly due to its side effects, PMDs supporting this mode may not have the 3758ability to toggle it more than once without reinitializing affected ports 3759first (e.g. by exiting testpmd). 3760 3761Enabling isolated mode:: 3762 3763 testpmd> flow isolate 0 true 3764 Ingress traffic on port 0 is now restricted to the defined flow rules 3765 testpmd> 3766 3767Disabling isolated mode:: 3768 3769 testpmd> flow isolate 0 false 3770 Ingress traffic on port 0 is not restricted anymore to the defined flow rules 3771 testpmd> 3772 3773Sample QinQ flow rules 3774~~~~~~~~~~~~~~~~~~~~~~ 3775 3776Before creating QinQ rule(s) the following commands should be issued to enable QinQ:: 3777 3778 testpmd> port stop 0 3779 testpmd> vlan set qinq on 0 3780 3781The above command sets the inner and outer TPID's to 0x8100. 3782 3783To change the TPID's the following commands should be used:: 3784 3785 testpmd> vlan set outer tpid 0xa100 0 3786 testpmd> vlan set inner tpid 0x9100 0 3787 testpmd> port start 0 3788 3789Validate and create a QinQ rule on port 0 to steer traffic to a VF queue in a VM. 3790 3791:: 3792 3793 testpmd> flow validate 0 ingress pattern eth / vlan tci is 123 / 3794 vlan tci is 456 / end actions vf id 1 / queue index 0 / end 3795 Flow rule #0 validated 3796 3797 testpmd> flow create 0 ingress pattern eth / vlan tci is 4 / 3798 vlan tci is 456 / end actions vf id 123 / queue index 0 / end 3799 Flow rule #0 created 3800 3801 testpmd> flow list 0 3802 ID Group Prio Attr Rule 3803 0 0 0 i- ETH VLAN VLAN=>VF QUEUE 3804 3805Validate and create a QinQ rule on port 0 to steer traffic to a queue on the host. 3806 3807:: 3808 3809 testpmd> flow validate 0 ingress pattern eth / vlan tci is 321 / 3810 vlan tci is 654 / end actions pf / queue index 0 / end 3811 Flow rule #1 validated 3812 3813 testpmd> flow create 0 ingress pattern eth / vlan tci is 321 / 3814 vlan tci is 654 / end actions pf / queue index 1 / end 3815 Flow rule #1 created 3816 3817 testpmd> flow list 0 3818 ID Group Prio Attr Rule 3819 0 0 0 i- ETH VLAN VLAN=>VF QUEUE 3820 1 0 0 i- ETH VLAN VLAN=>PF QUEUE 3821