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 start/stop queue 1627~~~~~~~~~~~~~~~~~~~~~ 1628 1629Start/stop a rx/tx queue on a specific port:: 1630 1631 testpmd> port (port_id) (rxq|txq) (queue_id) (start|stop) 1632 1633port setup queue 1634~~~~~~~~~~~~~~~~~~~~~ 1635 1636Setup a rx/tx queue on a specific port:: 1637 1638 testpmd> port (port_id) (rxq|txq) (queue_id) setup 1639 1640Only take effect when port is started. 1641 1642port config - speed 1643~~~~~~~~~~~~~~~~~~~ 1644 1645Set the speed and duplex mode for all ports or a specific port:: 1646 1647 testpmd> port config (port_id|all) speed (10|100|1000|10000|25000|40000|50000|100000|auto) \ 1648 duplex (half|full|auto) 1649 1650port config - queues/descriptors 1651~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ 1652 1653Set number of queues/descriptors for rxq, txq, rxd and txd:: 1654 1655 testpmd> port config all (rxq|txq|rxd|txd) (value) 1656 1657This is equivalent to the ``--rxq``, ``--txq``, ``--rxd`` and ``--txd`` command-line options. 1658 1659port config - max-pkt-len 1660~~~~~~~~~~~~~~~~~~~~~~~~~ 1661 1662Set the maximum packet length:: 1663 1664 testpmd> port config all max-pkt-len (value) 1665 1666This is equivalent to the ``--max-pkt-len`` command-line option. 1667 1668port config - CRC Strip 1669~~~~~~~~~~~~~~~~~~~~~~~ 1670 1671Set hardware CRC stripping on or off for all ports:: 1672 1673 testpmd> port config all crc-strip (on|off) 1674 1675CRC stripping is on by default. 1676 1677The ``off`` option is equivalent to the ``--disable-crc-strip`` command-line option. 1678 1679port config - scatter 1680~~~~~~~~~~~~~~~~~~~~~~~ 1681 1682Set RX scatter mode on or off for all ports:: 1683 1684 testpmd> port config all scatter (on|off) 1685 1686RX scatter mode is off by default. 1687 1688The ``on`` option is equivalent to the ``--enable-scatter`` command-line option. 1689 1690port config - RX Checksum 1691~~~~~~~~~~~~~~~~~~~~~~~~~ 1692 1693Set hardware RX checksum offload to on or off for all ports:: 1694 1695 testpmd> port config all rx-cksum (on|off) 1696 1697Checksum offload is off by default. 1698 1699The ``on`` option is equivalent to the ``--enable-rx-cksum`` command-line option. 1700 1701port config - VLAN 1702~~~~~~~~~~~~~~~~~~ 1703 1704Set hardware VLAN on or off for all ports:: 1705 1706 testpmd> port config all hw-vlan (on|off) 1707 1708Hardware VLAN is off by default. 1709 1710The ``on`` option is equivalent to the ``--enable-hw-vlan`` command-line option. 1711 1712port config - VLAN filter 1713~~~~~~~~~~~~~~~~~~~~~~~~~ 1714 1715Set hardware VLAN filter on or off for all ports:: 1716 1717 testpmd> port config all hw-vlan-filter (on|off) 1718 1719Hardware VLAN filter is off by default. 1720 1721The ``on`` option is equivalent to the ``--enable-hw-vlan-filter`` command-line option. 1722 1723port config - VLAN strip 1724~~~~~~~~~~~~~~~~~~~~~~~~ 1725 1726Set hardware VLAN strip on or off for all ports:: 1727 1728 testpmd> port config all hw-vlan-strip (on|off) 1729 1730Hardware VLAN strip is off by default. 1731 1732The ``on`` option is equivalent to the ``--enable-hw-vlan-strip`` command-line option. 1733 1734port config - VLAN extend 1735~~~~~~~~~~~~~~~~~~~~~~~~~ 1736 1737Set hardware VLAN extend on or off for all ports:: 1738 1739 testpmd> port config all hw-vlan-extend (on|off) 1740 1741Hardware VLAN extend is off by default. 1742 1743The ``on`` option is equivalent to the ``--enable-hw-vlan-extend`` command-line option. 1744 1745port config - Drop Packets 1746~~~~~~~~~~~~~~~~~~~~~~~~~~ 1747 1748Set packet drop for packets with no descriptors on or off for all ports:: 1749 1750 testpmd> port config all drop-en (on|off) 1751 1752Packet dropping for packets with no descriptors is off by default. 1753 1754The ``on`` option is equivalent to the ``--enable-drop-en`` command-line option. 1755 1756port config - RSS 1757~~~~~~~~~~~~~~~~~ 1758 1759Set the RSS (Receive Side Scaling) mode on or off:: 1760 1761 testpmd> port config all rss (all|default|ip|tcp|udp|sctp|ether|port|vxlan|geneve|nvgre|none) 1762 1763RSS is on by default. 1764 1765The ``all`` option is equivalent to ip|tcp|udp|sctp|ether. 1766The ``default`` option enables all supported RSS types reported by device info. 1767The ``none`` option is equivalent to the ``--disable-rss`` command-line option. 1768 1769port config - RSS Reta 1770~~~~~~~~~~~~~~~~~~~~~~ 1771 1772Set the RSS (Receive Side Scaling) redirection table:: 1773 1774 testpmd> port config all rss reta (hash,queue)[,(hash,queue)] 1775 1776port config - DCB 1777~~~~~~~~~~~~~~~~~ 1778 1779Set the DCB mode for an individual port:: 1780 1781 testpmd> port config (port_id) dcb vt (on|off) (traffic_class) pfc (on|off) 1782 1783The traffic class should be 4 or 8. 1784 1785port config - Burst 1786~~~~~~~~~~~~~~~~~~~ 1787 1788Set the number of packets per burst:: 1789 1790 testpmd> port config all burst (value) 1791 1792This is equivalent to the ``--burst`` command-line option. 1793 1794port config - Threshold 1795~~~~~~~~~~~~~~~~~~~~~~~ 1796 1797Set thresholds for TX/RX queues:: 1798 1799 testpmd> port config all (threshold) (value) 1800 1801Where the threshold type can be: 1802 1803* ``txpt:`` Set the prefetch threshold register of the TX rings, 0 <= value <= 255. 1804 1805* ``txht:`` Set the host threshold register of the TX rings, 0 <= value <= 255. 1806 1807* ``txwt:`` Set the write-back threshold register of the TX rings, 0 <= value <= 255. 1808 1809* ``rxpt:`` Set the prefetch threshold register of the RX rings, 0 <= value <= 255. 1810 1811* ``rxht:`` Set the host threshold register of the RX rings, 0 <= value <= 255. 1812 1813* ``rxwt:`` Set the write-back threshold register of the RX rings, 0 <= value <= 255. 1814 1815* ``txfreet:`` Set the transmit free threshold of the TX rings, 0 <= value <= txd. 1816 1817* ``rxfreet:`` Set the transmit free threshold of the RX rings, 0 <= value <= rxd. 1818 1819* ``txrst:`` Set the transmit RS bit threshold of TX rings, 0 <= value <= txd. 1820 1821These threshold options are also available from the command-line. 1822 1823port config - E-tag 1824~~~~~~~~~~~~~~~~~~~ 1825 1826Set the value of ether-type for E-tag:: 1827 1828 testpmd> port config (port_id|all) l2-tunnel E-tag ether-type (value) 1829 1830Enable/disable the E-tag support:: 1831 1832 testpmd> port config (port_id|all) l2-tunnel E-tag (enable|disable) 1833 1834port config pctype mapping 1835~~~~~~~~~~~~~~~~~~~~~~~~~~ 1836 1837Reset pctype mapping table:: 1838 1839 testpmd> port config (port_id) pctype mapping reset 1840 1841Update hardware defined pctype to software defined flow type mapping table:: 1842 1843 testpmd> port config (port_id) pctype mapping update (pctype_id_0[,pctype_id_1]*) (flow_type_id) 1844 1845where: 1846 1847* ``pctype_id_x``: hardware pctype id as index of bit in bitmask value of the pctype mapping table. 1848 1849* ``flow_type_id``: software flow type id as the index of the pctype mapping table. 1850 1851port config input set 1852~~~~~~~~~~~~~~~~~~~~~ 1853 1854Config RSS/FDIR/FDIR flexible payload input set for some pctype:: 1855 testpmd> port config (port_id) pctype (pctype_id) \ 1856 (hash_inset|fdir_inset|fdir_flx_inset) \ 1857 (get|set|clear) field (field_idx) 1858 1859Clear RSS/FDIR/FDIR flexible payload input set for some pctype:: 1860 testpmd> port config (port_id) pctype (pctype_id) \ 1861 (hash_inset|fdir_inset|fdir_flx_inset) clear all 1862 1863where: 1864 1865* ``pctype_id``: hardware packet classification types. 1866* ``field_idx``: hardware field index. 1867 1868Link Bonding Functions 1869---------------------- 1870 1871The Link Bonding functions make it possible to dynamically create and 1872manage link bonding devices from within testpmd interactive prompt. 1873 1874create bonded device 1875~~~~~~~~~~~~~~~~~~~~ 1876 1877Create a new bonding device:: 1878 1879 testpmd> create bonded device (mode) (socket) 1880 1881For example, to create a bonded device in mode 1 on socket 0:: 1882 1883 testpmd> create bonded 1 0 1884 created new bonded device (port X) 1885 1886add bonding slave 1887~~~~~~~~~~~~~~~~~ 1888 1889Adds Ethernet device to a Link Bonding device:: 1890 1891 testpmd> add bonding slave (slave id) (port id) 1892 1893For example, to add Ethernet device (port 6) to a Link Bonding device (port 10):: 1894 1895 testpmd> add bonding slave 6 10 1896 1897 1898remove bonding slave 1899~~~~~~~~~~~~~~~~~~~~ 1900 1901Removes an Ethernet slave device from a Link Bonding device:: 1902 1903 testpmd> remove bonding slave (slave id) (port id) 1904 1905For example, to remove Ethernet slave device (port 6) to a Link Bonding device (port 10):: 1906 1907 testpmd> remove bonding slave 6 10 1908 1909set bonding mode 1910~~~~~~~~~~~~~~~~ 1911 1912Set the Link Bonding mode of a Link Bonding device:: 1913 1914 testpmd> set bonding mode (value) (port id) 1915 1916For example, to set the bonding mode of a Link Bonding device (port 10) to broadcast (mode 3):: 1917 1918 testpmd> set bonding mode 3 10 1919 1920set bonding primary 1921~~~~~~~~~~~~~~~~~~~ 1922 1923Set an Ethernet slave device as the primary device on a Link Bonding device:: 1924 1925 testpmd> set bonding primary (slave id) (port id) 1926 1927For example, to set the Ethernet slave device (port 6) as the primary port of a Link Bonding device (port 10):: 1928 1929 testpmd> set bonding primary 6 10 1930 1931set bonding mac 1932~~~~~~~~~~~~~~~ 1933 1934Set the MAC address of a Link Bonding device:: 1935 1936 testpmd> set bonding mac (port id) (mac) 1937 1938For example, to set the MAC address of a Link Bonding device (port 10) to 00:00:00:00:00:01:: 1939 1940 testpmd> set bonding mac 10 00:00:00:00:00:01 1941 1942set bonding xmit_balance_policy 1943~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ 1944 1945Set the transmission policy for a Link Bonding device when it is in Balance XOR mode:: 1946 1947 testpmd> set bonding xmit_balance_policy (port_id) (l2|l23|l34) 1948 1949For example, set a Link Bonding device (port 10) to use a balance policy of layer 3+4 (IP addresses & UDP ports):: 1950 1951 testpmd> set bonding xmit_balance_policy 10 l34 1952 1953 1954set bonding mon_period 1955~~~~~~~~~~~~~~~~~~~~~~ 1956 1957Set the link status monitoring polling period in milliseconds for a bonding device. 1958 1959This adds support for PMD slave devices which do not support link status interrupts. 1960When the mon_period is set to a value greater than 0 then all PMD's which do not support 1961link status ISR will be queried every polling interval to check if their link status has changed:: 1962 1963 testpmd> set bonding mon_period (port_id) (value) 1964 1965For example, to set the link status monitoring polling period of bonded device (port 5) to 150ms:: 1966 1967 testpmd> set bonding mon_period 5 150 1968 1969 1970set bonding lacp dedicated_queue 1971~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ 1972 1973Enable dedicated tx/rx queues on bonding devices slaves to handle LACP control plane traffic 1974when in mode 4 (link-aggregration-802.3ad):: 1975 1976 testpmd> set bonding lacp dedicated_queues (port_id) (enable|disable) 1977 1978 1979set bonding agg_mode 1980~~~~~~~~~~~~~~~~~~~~ 1981 1982Enable one of the specific aggregators mode when in mode 4 (link-aggregration-802.3ad):: 1983 1984 testpmd> set bonding agg_mode (port_id) (bandwidth|count|stable) 1985 1986 1987show bonding config 1988~~~~~~~~~~~~~~~~~~~ 1989 1990Show the current configuration of a Link Bonding device:: 1991 1992 testpmd> show bonding config (port id) 1993 1994For example, 1995to show the configuration a Link Bonding device (port 9) with 3 slave devices (1, 3, 4) 1996in balance mode with a transmission policy of layer 2+3:: 1997 1998 testpmd> show bonding config 9 1999 Bonding mode: 2 2000 Balance Xmit Policy: BALANCE_XMIT_POLICY_LAYER23 2001 Slaves (3): [1 3 4] 2002 Active Slaves (3): [1 3 4] 2003 Primary: [3] 2004 2005 2006Register Functions 2007------------------ 2008 2009The Register Functions can be used to read from and write to registers on the network card referenced by a port number. 2010This is mainly useful for debugging purposes. 2011Reference should be made to the appropriate datasheet for the network card for details on the register addresses 2012and fields that can be accessed. 2013 2014read reg 2015~~~~~~~~ 2016 2017Display the value of a port register:: 2018 2019 testpmd> read reg (port_id) (address) 2020 2021For example, to examine the Flow Director control register (FDIRCTL, 0x0000EE000) on an Intel 82599 10 GbE Controller:: 2022 2023 testpmd> read reg 0 0xEE00 2024 port 0 PCI register at offset 0xEE00: 0x4A060029 (1241907241) 2025 2026read regfield 2027~~~~~~~~~~~~~ 2028 2029Display a port register bit field:: 2030 2031 testpmd> read regfield (port_id) (address) (bit_x) (bit_y) 2032 2033For example, reading the lowest two bits from the register in the example above:: 2034 2035 testpmd> read regfield 0 0xEE00 0 1 2036 port 0 PCI register at offset 0xEE00: bits[0, 1]=0x1 (1) 2037 2038read regbit 2039~~~~~~~~~~~ 2040 2041Display a single port register bit:: 2042 2043 testpmd> read regbit (port_id) (address) (bit_x) 2044 2045For example, reading the lowest bit from the register in the example above:: 2046 2047 testpmd> read regbit 0 0xEE00 0 2048 port 0 PCI register at offset 0xEE00: bit 0=1 2049 2050write reg 2051~~~~~~~~~ 2052 2053Set the value of a port register:: 2054 2055 testpmd> write reg (port_id) (address) (value) 2056 2057For example, to clear a register:: 2058 2059 testpmd> write reg 0 0xEE00 0x0 2060 port 0 PCI register at offset 0xEE00: 0x00000000 (0) 2061 2062write regfield 2063~~~~~~~~~~~~~~ 2064 2065Set bit field of a port register:: 2066 2067 testpmd> write regfield (port_id) (address) (bit_x) (bit_y) (value) 2068 2069For example, writing to the register cleared in the example above:: 2070 2071 testpmd> write regfield 0 0xEE00 0 1 2 2072 port 0 PCI register at offset 0xEE00: 0x00000002 (2) 2073 2074write regbit 2075~~~~~~~~~~~~ 2076 2077Set single bit value of a port register:: 2078 2079 testpmd> write regbit (port_id) (address) (bit_x) (value) 2080 2081For example, to set the high bit in the register from the example above:: 2082 2083 testpmd> write regbit 0 0xEE00 31 1 2084 port 0 PCI register at offset 0xEE00: 0x8000000A (2147483658) 2085 2086Traffic Metering and Policing 2087----------------------------- 2088 2089The following section shows functions for configuring traffic metering and 2090policing on the ethernet device through the use of generic ethdev API. 2091 2092show port traffic management capability 2093~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ 2094 2095Show traffic metering and policing capability of the port:: 2096 2097 testpmd> show port meter cap (port_id) 2098 2099add port meter profile (srTCM rfc2967) 2100~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ 2101 2102Add meter profile (srTCM rfc2697) to the ethernet device:: 2103 2104 testpmd> add port meter profile srtcm_rfc2697 (port_id) (profile_id) \ 2105 (cir) (cbs) (ebs) 2106 2107where: 2108 2109* ``profile_id``: ID for the meter profile. 2110* ``cir``: Committed Information Rate (CIR) (bytes/second). 2111* ``cbs``: Committed Burst Size (CBS) (bytes). 2112* ``ebs``: Excess Burst Size (EBS) (bytes). 2113 2114add port meter profile (trTCM rfc2968) 2115~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ 2116 2117Add meter profile (srTCM rfc2698) to the ethernet device:: 2118 2119 testpmd> add port meter profile trtcm_rfc2698 (port_id) (profile_id) \ 2120 (cir) (pir) (cbs) (pbs) 2121 2122where: 2123 2124* ``profile_id``: ID for the meter profile. 2125* ``cir``: Committed information rate (bytes/second). 2126* ``pir``: Peak information rate (bytes/second). 2127* ``cbs``: Committed burst size (bytes). 2128* ``pbs``: Peak burst size (bytes). 2129 2130add port meter profile (trTCM rfc4115) 2131~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ 2132 2133Add meter profile (trTCM rfc4115) to the ethernet device:: 2134 2135 testpmd> add port meter profile trtcm_rfc4115 (port_id) (profile_id) \ 2136 (cir) (eir) (cbs) (ebs) 2137 2138where: 2139 2140* ``profile_id``: ID for the meter profile. 2141* ``cir``: Committed information rate (bytes/second). 2142* ``eir``: Excess information rate (bytes/second). 2143* ``cbs``: Committed burst size (bytes). 2144* ``ebs``: Excess burst size (bytes). 2145 2146delete port meter profile 2147~~~~~~~~~~~~~~~~~~~~~~~~~ 2148 2149Delete meter profile from the ethernet device:: 2150 2151 testpmd> del port meter profile (port_id) (profile_id) 2152 2153create port meter 2154~~~~~~~~~~~~~~~~~ 2155 2156Create new meter object for the ethernet device:: 2157 2158 testpmd> create port meter (port_id) (mtr_id) (profile_id) \ 2159 (meter_enable) (g_action) (y_action) (r_action) (stats_mask) (shared) \ 2160 (use_pre_meter_color) [(dscp_tbl_entry0) (dscp_tbl_entry1)...\ 2161 (dscp_tbl_entry63)] 2162 2163where: 2164 2165* ``mtr_id``: meter object ID. 2166* ``profile_id``: ID for the meter profile. 2167* ``meter_enable``: When this parameter has a non-zero value, the meter object 2168 gets enabled at the time of creation, otherwise remains disabled. 2169* ``g_action``: Policer action for the packet with green color. 2170* ``y_action``: Policer action for the packet with yellow color. 2171* ``r_action``: Policer action for the packet with red color. 2172* ``stats_mask``: Mask of statistics counter types to be enabled for the 2173 meter object. 2174* ``shared``: When this parameter has a non-zero value, the meter object is 2175 shared by multiple flows. Otherwise, meter object is used by single flow. 2176* ``use_pre_meter_color``: When this parameter has a non-zero value, the 2177 input color for the current meter object is determined by the latest meter 2178 object in the same flow. Otherwise, the current meter object uses the 2179 *dscp_table* to determine the input color. 2180* ``dscp_tbl_entryx``: DSCP table entry x providing meter providing input 2181 color, 0 <= x <= 63. 2182 2183enable port meter 2184~~~~~~~~~~~~~~~~~ 2185 2186Enable meter for the ethernet device:: 2187 2188 testpmd> enable port meter (port_id) (mtr_id) 2189 2190disable port meter 2191~~~~~~~~~~~~~~~~~~ 2192 2193Disable meter for the ethernet device:: 2194 2195 testpmd> disable port meter (port_id) (mtr_id) 2196 2197delete port meter 2198~~~~~~~~~~~~~~~~~ 2199 2200Delete meter for the ethernet device:: 2201 2202 testpmd> del port meter (port_id) (mtr_id) 2203 2204Set port meter profile 2205~~~~~~~~~~~~~~~~~~~~~~ 2206 2207Set meter profile for the ethernet device:: 2208 2209 testpmd> set port meter profile (port_id) (mtr_id) (profile_id) 2210 2211set port meter dscp table 2212~~~~~~~~~~~~~~~~~~~~~~~~~ 2213 2214Set meter dscp table for the ethernet device:: 2215 2216 testpmd> set port meter dscp table (port_id) (mtr_id) [(dscp_tbl_entry0) \ 2217 (dscp_tbl_entry1)...(dscp_tbl_entry63)] 2218 2219set port meter policer action 2220~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ 2221 2222Set meter policer action for the ethernet device:: 2223 2224 testpmd> set port meter policer action (port_id) (mtr_id) (action_mask) \ 2225 (action0) [(action1) (action1)] 2226 2227where: 2228 2229* ``action_mask``: Bit mask indicating which policer actions need to be 2230 updated. One or more policer actions can be updated in a single function 2231 invocation. To update the policer action associated with color C, bit 2232 (1 << C) needs to be set in *action_mask* and element at position C 2233 in the *actions* array needs to be valid. 2234* ``actionx``: Policer action for the color x, 2235 RTE_MTR_GREEN <= x < RTE_MTR_COLORS 2236 2237set port meter stats mask 2238~~~~~~~~~~~~~~~~~~~~~~~~~ 2239 2240Set meter stats mask for the ethernet device:: 2241 2242 testpmd> set port meter stats mask (port_id) (mtr_id) (stats_mask) 2243 2244where: 2245 2246* ``stats_mask``: Bit mask indicating statistics counter types to be enabled. 2247 2248show port meter stats 2249~~~~~~~~~~~~~~~~~~~~~ 2250 2251Show meter stats of the ethernet device:: 2252 2253 testpmd> show port meter stats (port_id) (mtr_id) (clear) 2254 2255where: 2256 2257* ``clear``: Flag that indicates whether the statistics counters should 2258 be cleared (i.e. set to zero) immediately after they have been read or not. 2259 2260Traffic Management 2261------------------ 2262 2263The following section shows functions for configuring traffic management on 2264on the ethernet device through the use of generic TM API. 2265 2266show port traffic management capability 2267~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ 2268 2269Show traffic management capability of the port:: 2270 2271 testpmd> show port tm cap (port_id) 2272 2273show port traffic management capability (hierarchy level) 2274~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ 2275 2276Show traffic management hierarchy level capability of the port:: 2277 2278 testpmd> show port tm level cap (port_id) (level_id) 2279 2280show port traffic management capability (hierarchy node level) 2281~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ 2282 2283Show the traffic management hierarchy node capability of the port:: 2284 2285 testpmd> show port tm node cap (port_id) (node_id) 2286 2287show port traffic management hierarchy node type 2288~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ 2289 2290Show the port traffic management hierarchy node type:: 2291 2292 testpmd> show port tm node type (port_id) (node_id) 2293 2294show port traffic management hierarchy node stats 2295~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ 2296 2297Show the port traffic management hierarchy node statistics:: 2298 2299 testpmd> show port tm node stats (port_id) (node_id) (clear) 2300 2301where: 2302 2303* ``clear``: When this parameter has a non-zero value, the statistics counters 2304 are cleared (i.e. set to zero) immediately after they have been read, 2305 otherwise the statistics counters are left untouched. 2306 2307Add port traffic management private shaper profile 2308~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ 2309 2310Add the port traffic management private shaper profile:: 2311 2312 testpmd> add port tm node shaper profile (port_id) (shaper_profile_id) \ 2313 (tb_rate) (tb_size) (packet_length_adjust) 2314 2315where: 2316 2317* ``shaper_profile id``: Shaper profile ID for the new profile. 2318* ``tb_rate``: Token bucket rate (bytes per second). 2319* ``tb_size``: Token bucket size (bytes). 2320* ``packet_length_adjust``: The value (bytes) to be added to the length of 2321 each packet for the purpose of shaping. This parameter value can be used to 2322 correct the packet length with the framing overhead bytes that are consumed 2323 on the wire. 2324 2325Delete port traffic management private shaper profile 2326~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ 2327 2328Delete the port traffic management private shaper:: 2329 2330 testpmd> del port tm node shaper profile (port_id) (shaper_profile_id) 2331 2332where: 2333 2334* ``shaper_profile id``: Shaper profile ID that needs to be deleted. 2335 2336Add port traffic management shared shaper 2337~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ 2338 2339Create the port traffic management shared shaper:: 2340 2341 testpmd> add port tm node shared shaper (port_id) (shared_shaper_id) \ 2342 (shaper_profile_id) 2343 2344where: 2345 2346* ``shared_shaper_id``: Shared shaper ID to be created. 2347* ``shaper_profile id``: Shaper profile ID for shared shaper. 2348 2349Set port traffic management shared shaper 2350~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ 2351 2352Update the port traffic management shared shaper:: 2353 2354 testpmd> set port tm node shared shaper (port_id) (shared_shaper_id) \ 2355 (shaper_profile_id) 2356 2357where: 2358 2359* ``shared_shaper_id``: Shared shaper ID to be update. 2360* ``shaper_profile id``: Shaper profile ID for shared shaper. 2361 2362Delete port traffic management shared shaper 2363~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ 2364 2365Delete the port traffic management shared shaper:: 2366 2367 testpmd> del port tm node shared shaper (port_id) (shared_shaper_id) 2368 2369where: 2370 2371* ``shared_shaper_id``: Shared shaper ID to be deleted. 2372 2373Set port traffic management hiearchy node private shaper 2374~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ 2375 2376set the port traffic management hierarchy node private shaper:: 2377 2378 testpmd> set port tm node shaper profile (port_id) (node_id) \ 2379 (shaper_profile_id) 2380 2381where: 2382 2383* ``shaper_profile id``: Private shaper profile ID to be enabled on the 2384 hierarchy node. 2385 2386Add port traffic management WRED profile 2387~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ 2388 2389Create a new WRED profile:: 2390 2391 testpmd> add port tm node wred profile (port_id) (wred_profile_id) \ 2392 (color_g) (min_th_g) (max_th_g) (maxp_inv_g) (wq_log2_g) \ 2393 (color_y) (min_th_y) (max_th_y) (maxp_inv_y) (wq_log2_y) \ 2394 (color_r) (min_th_r) (max_th_r) (maxp_inv_r) (wq_log2_r) 2395 2396where: 2397 2398* ``wred_profile id``: Identifier for the newly create WRED profile 2399* ``color_g``: Packet color (green) 2400* ``min_th_g``: Minimum queue threshold for packet with green color 2401* ``max_th_g``: Minimum queue threshold for packet with green color 2402* ``maxp_inv_g``: Inverse of packet marking probability maximum value (maxp) 2403* ``wq_log2_g``: Negated log2 of queue weight (wq) 2404* ``color_y``: Packet color (yellow) 2405* ``min_th_y``: Minimum queue threshold for packet with yellow color 2406* ``max_th_y``: Minimum queue threshold for packet with yellow color 2407* ``maxp_inv_y``: Inverse of packet marking probability maximum value (maxp) 2408* ``wq_log2_y``: Negated log2 of queue weight (wq) 2409* ``color_r``: Packet color (red) 2410* ``min_th_r``: Minimum queue threshold for packet with yellow color 2411* ``max_th_r``: Minimum queue threshold for packet with yellow color 2412* ``maxp_inv_r``: Inverse of packet marking probability maximum value (maxp) 2413* ``wq_log2_r``: Negated log2 of queue weight (wq) 2414 2415Delete port traffic management WRED profile 2416~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ 2417 2418Delete the WRED profile:: 2419 2420 testpmd> del port tm node wred profile (port_id) (wred_profile_id) 2421 2422Add port traffic management hierarchy nonleaf node 2423~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ 2424 2425Add nonleaf node to port traffic management hiearchy:: 2426 2427 testpmd> add port tm nonleaf node (port_id) (node_id) (parent_node_id) \ 2428 (priority) (weight) (level_id) (shaper_profile_id) \ 2429 (n_sp_priorities) (stats_mask) (n_shared_shapers) \ 2430 [(shared_shaper_0) (shared_shaper_1) ...] \ 2431 2432where: 2433 2434* ``parent_node_id``: Node ID of the parent. 2435* ``priority``: Node priority (highest node priority is zero). This is used by 2436 the SP algorithm running on the parent node for scheduling this node. 2437* ``weight``: Node weight (lowest weight is one). The node weight is relative 2438 to the weight sum of all siblings that have the same priority. It is used by 2439 the WFQ algorithm running on the parent node for scheduling this node. 2440* ``level_id``: Hiearchy level of the node. 2441* ``shaper_profile_id``: Shaper profile ID of the private shaper to be used by 2442 the node. 2443* ``n_sp_priorities``: Number of strict priorities. 2444* ``stats_mask``: Mask of statistics counter types to be enabled for this node. 2445* ``n_shared_shapers``: Number of shared shapers. 2446* ``shared_shaper_id``: Shared shaper id. 2447 2448Add port traffic management hierarchy leaf node 2449~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ 2450 2451Add leaf node to port traffic management hiearchy:: 2452 2453 testpmd> add port tm leaf node (port_id) (node_id) (parent_node_id) \ 2454 (priority) (weight) (level_id) (shaper_profile_id) \ 2455 (cman_mode) (wred_profile_id) (stats_mask) (n_shared_shapers) \ 2456 [(shared_shaper_id) (shared_shaper_id) ...] \ 2457 2458where: 2459 2460* ``parent_node_id``: Node ID of the parent. 2461* ``priority``: Node priority (highest node priority is zero). This is used by 2462 the SP algorithm running on the parent node for scheduling this node. 2463* ``weight``: Node weight (lowest weight is one). The node weight is relative 2464 to the weight sum of all siblings that have the same priority. It is used by 2465 the WFQ algorithm running on the parent node for scheduling this node. 2466* ``level_id``: Hiearchy level of the node. 2467* ``shaper_profile_id``: Shaper profile ID of the private shaper to be used by 2468 the node. 2469* ``cman_mode``: Congestion management mode to be enabled for this node. 2470* ``wred_profile_id``: WRED profile id to be enabled for this node. 2471* ``stats_mask``: Mask of statistics counter types to be enabled for this node. 2472* ``n_shared_shapers``: Number of shared shapers. 2473* ``shared_shaper_id``: Shared shaper id. 2474 2475Delete port traffic management hierarchy node 2476~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ 2477 2478Delete node from port traffic management hiearchy:: 2479 2480 testpmd> del port tm node (port_id) (node_id) 2481 2482Update port traffic management hierarchy parent node 2483~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ 2484 2485Update port traffic management hierarchy parent node:: 2486 2487 testpmd> set port tm node parent (port_id) (node_id) (parent_node_id) \ 2488 (priority) (weight) 2489 2490This function can only be called after the hierarchy commit invocation. Its 2491success depends on the port support for this operation, as advertised through 2492the port capability set. This function is valid for all nodes of the traffic 2493management hierarchy except root node. 2494 2495Commit port traffic management hierarchy 2496~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ 2497 2498Commit the traffic management hierarchy on the port:: 2499 2500 testpmd> port tm hierarchy commit (port_id) (clean_on_fail) 2501 2502where: 2503 2504* ``clean_on_fail``: When set to non-zero, hierarchy is cleared on function 2505 call failure. On the other hand, hierarchy is preserved when this parameter 2506 is equal to zero. 2507 2508Set port traffic management default hierarchy (tm forwarding mode) 2509~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ 2510 2511set the traffic management default hierarchy on the port:: 2512 2513 testpmd> set port tm hierarchy default (port_id) 2514 2515Filter Functions 2516---------------- 2517 2518This section details the available filter functions that are available. 2519 2520Note these functions interface the deprecated legacy filtering framework, 2521superseded by *rte_flow*. See `Flow rules management`_. 2522 2523ethertype_filter 2524~~~~~~~~~~~~~~~~~~~~ 2525 2526Add or delete a L2 Ethertype filter, which identify packets by their L2 Ethertype mainly assign them to a receive queue:: 2527 2528 ethertype_filter (port_id) (add|del) (mac_addr|mac_ignr) (mac_address) \ 2529 ethertype (ether_type) (drop|fwd) queue (queue_id) 2530 2531The available information parameters are: 2532 2533* ``port_id``: The port which the Ethertype filter assigned on. 2534 2535* ``mac_addr``: Compare destination mac address. 2536 2537* ``mac_ignr``: Ignore destination mac address match. 2538 2539* ``mac_address``: Destination mac address to match. 2540 2541* ``ether_type``: The EtherType value want to match, 2542 for example 0x0806 for ARP packet. 0x0800 (IPv4) and 0x86DD (IPv6) are invalid. 2543 2544* ``queue_id``: The receive queue associated with this EtherType filter. 2545 It is meaningless when deleting or dropping. 2546 2547Example, to add/remove an ethertype filter rule:: 2548 2549 testpmd> ethertype_filter 0 add mac_ignr 00:11:22:33:44:55 \ 2550 ethertype 0x0806 fwd queue 3 2551 2552 testpmd> ethertype_filter 0 del mac_ignr 00:11:22:33:44:55 \ 2553 ethertype 0x0806 fwd queue 3 2554 25552tuple_filter 2556~~~~~~~~~~~~~~~~~ 2557 2558Add or delete a 2-tuple filter, 2559which identifies packets by specific protocol and destination TCP/UDP port 2560and forwards packets into one of the receive queues:: 2561 2562 2tuple_filter (port_id) (add|del) dst_port (dst_port_value) \ 2563 protocol (protocol_value) mask (mask_value) \ 2564 tcp_flags (tcp_flags_value) priority (prio_value) \ 2565 queue (queue_id) 2566 2567The available information parameters are: 2568 2569* ``port_id``: The port which the 2-tuple filter assigned on. 2570 2571* ``dst_port_value``: Destination port in L4. 2572 2573* ``protocol_value``: IP L4 protocol. 2574 2575* ``mask_value``: Participates in the match or not by bit for field above, 1b means participate. 2576 2577* ``tcp_flags_value``: TCP control bits. The non-zero value is invalid, when the pro_value is not set to 0x06 (TCP). 2578 2579* ``prio_value``: Priority of this filter. 2580 2581* ``queue_id``: The receive queue associated with this 2-tuple filter. 2582 2583Example, to add/remove an 2tuple filter rule:: 2584 2585 testpmd> 2tuple_filter 0 add dst_port 32 protocol 0x06 mask 0x03 \ 2586 tcp_flags 0x02 priority 3 queue 3 2587 2588 testpmd> 2tuple_filter 0 del dst_port 32 protocol 0x06 mask 0x03 \ 2589 tcp_flags 0x02 priority 3 queue 3 2590 25915tuple_filter 2592~~~~~~~~~~~~~~~~~ 2593 2594Add or delete a 5-tuple filter, 2595which consists of a 5-tuple (protocol, source and destination IP addresses, source and destination TCP/UDP/SCTP port) 2596and routes packets into one of the receive queues:: 2597 2598 5tuple_filter (port_id) (add|del) dst_ip (dst_address) src_ip \ 2599 (src_address) dst_port (dst_port_value) \ 2600 src_port (src_port_value) protocol (protocol_value) \ 2601 mask (mask_value) tcp_flags (tcp_flags_value) \ 2602 priority (prio_value) queue (queue_id) 2603 2604The available information parameters are: 2605 2606* ``port_id``: The port which the 5-tuple filter assigned on. 2607 2608* ``dst_address``: Destination IP address. 2609 2610* ``src_address``: Source IP address. 2611 2612* ``dst_port_value``: TCP/UDP destination port. 2613 2614* ``src_port_value``: TCP/UDP source port. 2615 2616* ``protocol_value``: L4 protocol. 2617 2618* ``mask_value``: Participates in the match or not by bit for field above, 1b means participate 2619 2620* ``tcp_flags_value``: TCP control bits. The non-zero value is invalid, when the protocol_value is not set to 0x06 (TCP). 2621 2622* ``prio_value``: The priority of this filter. 2623 2624* ``queue_id``: The receive queue associated with this 5-tuple filter. 2625 2626Example, to add/remove an 5tuple filter rule:: 2627 2628 testpmd> 5tuple_filter 0 add dst_ip 2.2.2.5 src_ip 2.2.2.4 \ 2629 dst_port 64 src_port 32 protocol 0x06 mask 0x1F \ 2630 flags 0x0 priority 3 queue 3 2631 2632 testpmd> 5tuple_filter 0 del dst_ip 2.2.2.5 src_ip 2.2.2.4 \ 2633 dst_port 64 src_port 32 protocol 0x06 mask 0x1F \ 2634 flags 0x0 priority 3 queue 3 2635 2636syn_filter 2637~~~~~~~~~~ 2638 2639Using the SYN filter, TCP packets whose *SYN* flag is set can be forwarded to a separate queue:: 2640 2641 syn_filter (port_id) (add|del) priority (high|low) queue (queue_id) 2642 2643The available information parameters are: 2644 2645* ``port_id``: The port which the SYN filter assigned on. 2646 2647* ``high``: This SYN filter has higher priority than other filters. 2648 2649* ``low``: This SYN filter has lower priority than other filters. 2650 2651* ``queue_id``: The receive queue associated with this SYN filter 2652 2653Example:: 2654 2655 testpmd> syn_filter 0 add priority high queue 3 2656 2657flex_filter 2658~~~~~~~~~~~ 2659 2660With flex filter, packets can be recognized by any arbitrary pattern within the first 128 bytes of the packet 2661and routed into one of the receive queues:: 2662 2663 flex_filter (port_id) (add|del) len (len_value) bytes (bytes_value) \ 2664 mask (mask_value) priority (prio_value) queue (queue_id) 2665 2666The available information parameters are: 2667 2668* ``port_id``: The port which the Flex filter is assigned on. 2669 2670* ``len_value``: Filter length in bytes, no greater than 128. 2671 2672* ``bytes_value``: A string in hexadecimal, means the value the flex filter needs to match. 2673 2674* ``mask_value``: A string in hexadecimal, bit 1 means corresponding byte participates in the match. 2675 2676* ``prio_value``: The priority of this filter. 2677 2678* ``queue_id``: The receive queue associated with this Flex filter. 2679 2680Example:: 2681 2682 testpmd> flex_filter 0 add len 16 bytes 0x00000000000000000000000008060000 \ 2683 mask 000C priority 3 queue 3 2684 2685 testpmd> flex_filter 0 del len 16 bytes 0x00000000000000000000000008060000 \ 2686 mask 000C priority 3 queue 3 2687 2688 2689.. _testpmd_flow_director: 2690 2691flow_director_filter 2692~~~~~~~~~~~~~~~~~~~~ 2693 2694The Flow Director works in receive mode to identify specific flows or sets of flows and route them to specific queues. 2695 2696Four types of filtering are supported which are referred to as Perfect Match, Signature, Perfect-mac-vlan and 2697Perfect-tunnel filters, the match mode is set by the ``--pkt-filter-mode`` command-line parameter: 2698 2699* Perfect match filters. 2700 The hardware checks a match between the masked fields of the received packets and the programmed filters. 2701 The masked fields are for IP flow. 2702 2703* Signature filters. 2704 The hardware checks a match between a hash-based signature of the masked fields of the received packet. 2705 2706* Perfect-mac-vlan match filters. 2707 The hardware checks a match between the masked fields of the received packets and the programmed filters. 2708 The masked fields are for MAC VLAN flow. 2709 2710* Perfect-tunnel match filters. 2711 The hardware checks a match between the masked fields of the received packets and the programmed filters. 2712 The masked fields are for tunnel flow. 2713 2714* Perfect-raw-flow-type match filters. 2715 The hardware checks a match between the masked fields of the received packets and pre-loaded raw (template) packet. 2716 The masked fields are specified by input sets. 2717 2718The Flow Director filters can match the different fields for different type of packet: flow type, specific input set 2719per flow type and the flexible payload. 2720 2721The Flow Director can also mask out parts of all of these fields so that filters 2722are only applied to certain fields or parts of the fields. 2723 2724Note that for raw flow type mode the source and destination fields in the 2725raw packet buffer need to be presented in a reversed order with respect 2726to the expected received packets. 2727For example: IP source and destination addresses or TCP/UDP/SCTP 2728source and destination ports 2729 2730Different NICs may have different capabilities, command show port fdir (port_id) can be used to acquire the information. 2731 2732# Commands to add flow director filters of different flow types:: 2733 2734 flow_director_filter (port_id) mode IP (add|del|update) \ 2735 flow (ipv4-other|ipv4-frag|ipv6-other|ipv6-frag) \ 2736 src (src_ip_address) dst (dst_ip_address) \ 2737 tos (tos_value) proto (proto_value) ttl (ttl_value) \ 2738 vlan (vlan_value) flexbytes (flexbytes_value) \ 2739 (drop|fwd) pf|vf(vf_id) queue (queue_id) \ 2740 fd_id (fd_id_value) 2741 2742 flow_director_filter (port_id) mode IP (add|del|update) \ 2743 flow (ipv4-tcp|ipv4-udp|ipv6-tcp|ipv6-udp) \ 2744 src (src_ip_address) (src_port) \ 2745 dst (dst_ip_address) (dst_port) \ 2746 tos (tos_value) ttl (ttl_value) \ 2747 vlan (vlan_value) flexbytes (flexbytes_value) \ 2748 (drop|fwd) queue pf|vf(vf_id) (queue_id) \ 2749 fd_id (fd_id_value) 2750 2751 flow_director_filter (port_id) mode IP (add|del|update) \ 2752 flow (ipv4-sctp|ipv6-sctp) \ 2753 src (src_ip_address) (src_port) \ 2754 dst (dst_ip_address) (dst_port) \ 2755 tos (tos_value) ttl (ttl_value) \ 2756 tag (verification_tag) vlan (vlan_value) \ 2757 flexbytes (flexbytes_value) (drop|fwd) \ 2758 pf|vf(vf_id) queue (queue_id) fd_id (fd_id_value) 2759 2760 flow_director_filter (port_id) mode IP (add|del|update) flow l2_payload \ 2761 ether (ethertype) flexbytes (flexbytes_value) \ 2762 (drop|fwd) pf|vf(vf_id) queue (queue_id) 2763 fd_id (fd_id_value) 2764 2765 flow_director_filter (port_id) mode MAC-VLAN (add|del|update) \ 2766 mac (mac_address) vlan (vlan_value) \ 2767 flexbytes (flexbytes_value) (drop|fwd) \ 2768 queue (queue_id) fd_id (fd_id_value) 2769 2770 flow_director_filter (port_id) mode Tunnel (add|del|update) \ 2771 mac (mac_address) vlan (vlan_value) \ 2772 tunnel (NVGRE|VxLAN) tunnel-id (tunnel_id_value) \ 2773 flexbytes (flexbytes_value) (drop|fwd) \ 2774 queue (queue_id) fd_id (fd_id_value) 2775 2776 flow_director_filter (port_id) mode raw (add|del|update) flow (flow_id) \ 2777 (drop|fwd) queue (queue_id) fd_id (fd_id_value) \ 2778 packet (packet file name) 2779 2780For example, to add an ipv4-udp flow type filter:: 2781 2782 testpmd> flow_director_filter 0 mode IP add flow ipv4-udp src 2.2.2.3 32 \ 2783 dst 2.2.2.5 33 tos 2 ttl 40 vlan 0x1 flexbytes (0x88,0x48) \ 2784 fwd pf queue 1 fd_id 1 2785 2786For example, add an ipv4-other flow type filter:: 2787 2788 testpmd> flow_director_filter 0 mode IP add flow ipv4-other src 2.2.2.3 \ 2789 dst 2.2.2.5 tos 2 proto 20 ttl 40 vlan 0x1 \ 2790 flexbytes (0x88,0x48) fwd pf queue 1 fd_id 1 2791 2792flush_flow_director 2793~~~~~~~~~~~~~~~~~~~ 2794 2795Flush all flow director filters on a device:: 2796 2797 testpmd> flush_flow_director (port_id) 2798 2799Example, to flush all flow director filter on port 0:: 2800 2801 testpmd> flush_flow_director 0 2802 2803flow_director_mask 2804~~~~~~~~~~~~~~~~~~ 2805 2806Set flow director's input masks:: 2807 2808 flow_director_mask (port_id) mode IP vlan (vlan_value) \ 2809 src_mask (ipv4_src) (ipv6_src) (src_port) \ 2810 dst_mask (ipv4_dst) (ipv6_dst) (dst_port) 2811 2812 flow_director_mask (port_id) mode MAC-VLAN vlan (vlan_value) 2813 2814 flow_director_mask (port_id) mode Tunnel vlan (vlan_value) \ 2815 mac (mac_value) tunnel-type (tunnel_type_value) \ 2816 tunnel-id (tunnel_id_value) 2817 2818Example, to set flow director mask on port 0:: 2819 2820 testpmd> flow_director_mask 0 mode IP vlan 0xefff \ 2821 src_mask 255.255.255.255 \ 2822 FFFF:FFFF:FFFF:FFFF:FFFF:FFFF:FFFF:FFFF 0xFFFF \ 2823 dst_mask 255.255.255.255 \ 2824 FFFF:FFFF:FFFF:FFFF:FFFF:FFFF:FFFF:FFFF 0xFFFF 2825 2826flow_director_flex_mask 2827~~~~~~~~~~~~~~~~~~~~~~~ 2828 2829set masks of flow director's flexible payload based on certain flow type:: 2830 2831 testpmd> flow_director_flex_mask (port_id) \ 2832 flow (none|ipv4-other|ipv4-frag|ipv4-tcp|ipv4-udp|ipv4-sctp| \ 2833 ipv6-other|ipv6-frag|ipv6-tcp|ipv6-udp|ipv6-sctp| \ 2834 l2_payload|all) (mask) 2835 2836Example, to set flow director's flex mask for all flow type on port 0:: 2837 2838 testpmd> flow_director_flex_mask 0 flow all \ 2839 (0xff,0xff,0,0,0,0,0,0,0,0,0,0,0,0,0,0) 2840 2841 2842flow_director_flex_payload 2843~~~~~~~~~~~~~~~~~~~~~~~~~~ 2844 2845Configure flexible payload selection:: 2846 2847 flow_director_flex_payload (port_id) (raw|l2|l3|l4) (config) 2848 2849For example, to select the first 16 bytes from the offset 4 (bytes) of packet's payload as flexible payload:: 2850 2851 testpmd> flow_director_flex_payload 0 l4 \ 2852 (4,5,6,7,8,9,10,11,12,13,14,15,16,17,18,19) 2853 2854get_sym_hash_ena_per_port 2855~~~~~~~~~~~~~~~~~~~~~~~~~ 2856 2857Get symmetric hash enable configuration per port:: 2858 2859 get_sym_hash_ena_per_port (port_id) 2860 2861For example, to get symmetric hash enable configuration of port 1:: 2862 2863 testpmd> get_sym_hash_ena_per_port 1 2864 2865set_sym_hash_ena_per_port 2866~~~~~~~~~~~~~~~~~~~~~~~~~ 2867 2868Set symmetric hash enable configuration per port to enable or disable:: 2869 2870 set_sym_hash_ena_per_port (port_id) (enable|disable) 2871 2872For example, to set symmetric hash enable configuration of port 1 to enable:: 2873 2874 testpmd> set_sym_hash_ena_per_port 1 enable 2875 2876get_hash_global_config 2877~~~~~~~~~~~~~~~~~~~~~~ 2878 2879Get the global configurations of hash filters:: 2880 2881 get_hash_global_config (port_id) 2882 2883For example, to get the global configurations of hash filters of port 1:: 2884 2885 testpmd> get_hash_global_config 1 2886 2887set_hash_global_config 2888~~~~~~~~~~~~~~~~~~~~~~ 2889 2890Set the global configurations of hash filters:: 2891 2892 set_hash_global_config (port_id) (toeplitz|simple_xor|default) \ 2893 (ipv4|ipv4-frag|ipv4-tcp|ipv4-udp|ipv4-sctp|ipv4-other|ipv6|ipv6-frag| \ 2894 ipv6-tcp|ipv6-udp|ipv6-sctp|ipv6-other|l2_payload|<flow_id>) \ 2895 (enable|disable) 2896 2897For example, to enable simple_xor for flow type of ipv6 on port 2:: 2898 2899 testpmd> set_hash_global_config 2 simple_xor ipv6 enable 2900 2901set_hash_input_set 2902~~~~~~~~~~~~~~~~~~ 2903 2904Set the input set for hash:: 2905 2906 set_hash_input_set (port_id) (ipv4-frag|ipv4-tcp|ipv4-udp|ipv4-sctp| \ 2907 ipv4-other|ipv6-frag|ipv6-tcp|ipv6-udp|ipv6-sctp|ipv6-other| \ 2908 l2_payload|<flow_id>) (ovlan|ivlan|src-ipv4|dst-ipv4|src-ipv6|dst-ipv6| \ 2909 ipv4-tos|ipv4-proto|ipv6-tc|ipv6-next-header|udp-src-port|udp-dst-port| \ 2910 tcp-src-port|tcp-dst-port|sctp-src-port|sctp-dst-port|sctp-veri-tag| \ 2911 udp-key|gre-key|fld-1st|fld-2nd|fld-3rd|fld-4th|fld-5th|fld-6th|fld-7th| \ 2912 fld-8th|none) (select|add) 2913 2914For example, to add source IP to hash input set for flow type of ipv4-udp on port 0:: 2915 2916 testpmd> set_hash_input_set 0 ipv4-udp src-ipv4 add 2917 2918set_fdir_input_set 2919~~~~~~~~~~~~~~~~~~ 2920 2921The Flow Director filters can match the different fields for different type of packet, i.e. specific input set 2922on per flow type and the flexible payload. This command can be used to change input set for each flow type. 2923 2924Set the input set for flow director:: 2925 2926 set_fdir_input_set (port_id) (ipv4-frag|ipv4-tcp|ipv4-udp|ipv4-sctp| \ 2927 ipv4-other|ipv6|ipv6-frag|ipv6-tcp|ipv6-udp|ipv6-sctp|ipv6-other| \ 2928 l2_payload|<flow_id>) (ivlan|ethertype|src-ipv4|dst-ipv4|src-ipv6|dst-ipv6| \ 2929 ipv4-tos|ipv4-proto|ipv4-ttl|ipv6-tc|ipv6-next-header|ipv6-hop-limits| \ 2930 tudp-src-port|udp-dst-port|cp-src-port|tcp-dst-port|sctp-src-port| \ 2931 sctp-dst-port|sctp-veri-tag|none) (select|add) 2932 2933For example to add source IP to FD input set for flow type of ipv4-udp on port 0:: 2934 2935 testpmd> set_fdir_input_set 0 ipv4-udp src-ipv4 add 2936 2937global_config 2938~~~~~~~~~~~~~ 2939 2940Set different GRE key length for input set:: 2941 2942 global_config (port_id) gre-key-len (number in bytes) 2943 2944For example to set GRE key length for input set to 4 bytes on port 0:: 2945 2946 testpmd> global_config 0 gre-key-len 4 2947 2948 2949.. _testpmd_rte_flow: 2950 2951Flow rules management 2952--------------------- 2953 2954Control of the generic flow API (*rte_flow*) is fully exposed through the 2955``flow`` command (validation, creation, destruction, queries and operation 2956modes). 2957 2958Considering *rte_flow* overlaps with all `Filter Functions`_, using both 2959features simultaneously may cause undefined side-effects and is therefore 2960not recommended. 2961 2962``flow`` syntax 2963~~~~~~~~~~~~~~~ 2964 2965Because the ``flow`` command uses dynamic tokens to handle the large number 2966of possible flow rules combinations, its behavior differs slightly from 2967other commands, in particular: 2968 2969- Pressing *?* or the *<tab>* key displays contextual help for the current 2970 token, not that of the entire command. 2971 2972- Optional and repeated parameters are supported (provided they are listed 2973 in the contextual help). 2974 2975The first parameter stands for the operation mode. Possible operations and 2976their general syntax are described below. They are covered in detail in the 2977following sections. 2978 2979- Check whether a flow rule can be created:: 2980 2981 flow validate {port_id} 2982 [group {group_id}] [priority {level}] [ingress] [egress] 2983 pattern {item} [/ {item} [...]] / end 2984 actions {action} [/ {action} [...]] / end 2985 2986- Create a flow rule:: 2987 2988 flow create {port_id} 2989 [group {group_id}] [priority {level}] [ingress] [egress] 2990 pattern {item} [/ {item} [...]] / end 2991 actions {action} [/ {action} [...]] / end 2992 2993- Destroy specific flow rules:: 2994 2995 flow destroy {port_id} rule {rule_id} [...] 2996 2997- Destroy all flow rules:: 2998 2999 flow flush {port_id} 3000 3001- Query an existing flow rule:: 3002 3003 flow query {port_id} {rule_id} {action} 3004 3005- List existing flow rules sorted by priority, filtered by group 3006 identifiers:: 3007 3008 flow list {port_id} [group {group_id}] [...] 3009 3010- Restrict ingress traffic to the defined flow rules:: 3011 3012 flow isolate {port_id} {boolean} 3013 3014Validating flow rules 3015~~~~~~~~~~~~~~~~~~~~~ 3016 3017``flow validate`` reports whether a flow rule would be accepted by the 3018underlying device in its current state but stops short of creating it. It is 3019bound to ``rte_flow_validate()``:: 3020 3021 flow validate {port_id} 3022 [group {group_id}] [priority {level}] [ingress] [egress] 3023 pattern {item} [/ {item} [...]] / end 3024 actions {action} [/ {action} [...]] / end 3025 3026If successful, it will show:: 3027 3028 Flow rule validated 3029 3030Otherwise it will show an error message of the form:: 3031 3032 Caught error type [...] ([...]): [...] 3033 3034This command uses the same parameters as ``flow create``, their format is 3035described in `Creating flow rules`_. 3036 3037Check whether redirecting any Ethernet packet received on port 0 to RX queue 3038index 6 is supported:: 3039 3040 testpmd> flow validate 0 ingress pattern eth / end 3041 actions queue index 6 / end 3042 Flow rule validated 3043 testpmd> 3044 3045Port 0 does not support TCPv6 rules:: 3046 3047 testpmd> flow validate 0 ingress pattern eth / ipv6 / tcp / end 3048 actions drop / end 3049 Caught error type 9 (specific pattern item): Invalid argument 3050 testpmd> 3051 3052Creating flow rules 3053~~~~~~~~~~~~~~~~~~~ 3054 3055``flow create`` validates and creates the specified flow rule. It is bound 3056to ``rte_flow_create()``:: 3057 3058 flow create {port_id} 3059 [group {group_id}] [priority {level}] [ingress] [egress] 3060 pattern {item} [/ {item} [...]] / end 3061 actions {action} [/ {action} [...]] / end 3062 3063If successful, it will return a flow rule ID usable with other commands:: 3064 3065 Flow rule #[...] created 3066 3067Otherwise it will show an error message of the form:: 3068 3069 Caught error type [...] ([...]): [...] 3070 3071Parameters describe in the following order: 3072 3073- Attributes (*group*, *priority*, *ingress*, *egress* tokens). 3074- A matching pattern, starting with the *pattern* token and terminated by an 3075 *end* pattern item. 3076- Actions, starting with the *actions* token and terminated by an *end* 3077 action. 3078 3079These translate directly to *rte_flow* objects provided as-is to the 3080underlying functions. 3081 3082The shortest valid definition only comprises mandatory tokens:: 3083 3084 testpmd> flow create 0 pattern end actions end 3085 3086Note that PMDs may refuse rules that essentially do nothing such as this 3087one. 3088 3089**All unspecified object values are automatically initialized to 0.** 3090 3091Attributes 3092^^^^^^^^^^ 3093 3094These tokens affect flow rule attributes (``struct rte_flow_attr``) and are 3095specified before the ``pattern`` token. 3096 3097- ``group {group id}``: priority group. 3098- ``priority {level}``: priority level within group. 3099- ``ingress``: rule applies to ingress traffic. 3100- ``egress``: rule applies to egress traffic. 3101 3102Each instance of an attribute specified several times overrides the previous 3103value as shown below (group 4 is used):: 3104 3105 testpmd> flow create 0 group 42 group 24 group 4 [...] 3106 3107Note that once enabled, ``ingress`` and ``egress`` cannot be disabled. 3108 3109While not specifying a direction is an error, some rules may allow both 3110simultaneously. 3111 3112Most rules affect RX therefore contain the ``ingress`` token:: 3113 3114 testpmd> flow create 0 ingress pattern [...] 3115 3116Matching pattern 3117^^^^^^^^^^^^^^^^ 3118 3119A matching pattern starts after the ``pattern`` token. It is made of pattern 3120items and is terminated by a mandatory ``end`` item. 3121 3122Items are named after their type (*RTE_FLOW_ITEM_TYPE_* from ``enum 3123rte_flow_item_type``). 3124 3125The ``/`` token is used as a separator between pattern items as shown 3126below:: 3127 3128 testpmd> flow create 0 ingress pattern eth / ipv4 / udp / end [...] 3129 3130Note that protocol items like these must be stacked from lowest to highest 3131layer to make sense. For instance, the following rule is either invalid or 3132unlikely to match any packet:: 3133 3134 testpmd> flow create 0 ingress pattern eth / udp / ipv4 / end [...] 3135 3136More information on these restrictions can be found in the *rte_flow* 3137documentation. 3138 3139Several items support additional specification structures, for example 3140``ipv4`` allows specifying source and destination addresses as follows:: 3141 3142 testpmd> flow create 0 ingress pattern eth / ipv4 src is 10.1.1.1 3143 dst is 10.2.0.0 / end [...] 3144 3145This rule matches all IPv4 traffic with the specified properties. 3146 3147In this example, ``src`` and ``dst`` are field names of the underlying 3148``struct rte_flow_item_ipv4`` object. All item properties can be specified 3149in a similar fashion. 3150 3151The ``is`` token means that the subsequent value must be matched exactly, 3152and assigns ``spec`` and ``mask`` fields in ``struct rte_flow_item`` 3153accordingly. Possible assignment tokens are: 3154 3155- ``is``: match value perfectly (with full bit-mask). 3156- ``spec``: match value according to configured bit-mask. 3157- ``last``: specify upper bound to establish a range. 3158- ``mask``: specify bit-mask with relevant bits set to one. 3159- ``prefix``: generate bit-mask from a prefix length. 3160 3161These yield identical results:: 3162 3163 ipv4 src is 10.1.1.1 3164 3165:: 3166 3167 ipv4 src spec 10.1.1.1 src mask 255.255.255.255 3168 3169:: 3170 3171 ipv4 src spec 10.1.1.1 src prefix 32 3172 3173:: 3174 3175 ipv4 src is 10.1.1.1 src last 10.1.1.1 # range with a single value 3176 3177:: 3178 3179 ipv4 src is 10.1.1.1 src last 0 # 0 disables range 3180 3181Inclusive ranges can be defined with ``last``:: 3182 3183 ipv4 src is 10.1.1.1 src last 10.2.3.4 # 10.1.1.1 to 10.2.3.4 3184 3185Note that ``mask`` affects both ``spec`` and ``last``:: 3186 3187 ipv4 src is 10.1.1.1 src last 10.2.3.4 src mask 255.255.0.0 3188 # matches 10.1.0.0 to 10.2.255.255 3189 3190Properties can be modified multiple times:: 3191 3192 ipv4 src is 10.1.1.1 src is 10.1.2.3 src is 10.2.3.4 # matches 10.2.3.4 3193 3194:: 3195 3196 ipv4 src is 10.1.1.1 src prefix 24 src prefix 16 # matches 10.1.0.0/16 3197 3198Pattern items 3199^^^^^^^^^^^^^ 3200 3201This section lists supported pattern items and their attributes, if any. 3202 3203- ``end``: end list of pattern items. 3204 3205- ``void``: no-op pattern item. 3206 3207- ``invert``: perform actions when pattern does not match. 3208 3209- ``any``: match any protocol for the current layer. 3210 3211 - ``num {unsigned}``: number of layers covered. 3212 3213- ``pf``: match packets addressed to the physical function. 3214 3215- ``vf``: match packets addressed to a virtual function ID. 3216 3217 - ``id {unsigned}``: destination VF ID. 3218 3219- ``port``: device-specific physical port index to use. 3220 3221 - ``index {unsigned}``: physical port index. 3222 3223- ``raw``: match an arbitrary byte string. 3224 3225 - ``relative {boolean}``: look for pattern after the previous item. 3226 - ``search {boolean}``: search pattern from offset (see also limit). 3227 - ``offset {integer}``: absolute or relative offset for pattern. 3228 - ``limit {unsigned}``: search area limit for start of pattern. 3229 - ``pattern {string}``: byte string to look for. 3230 3231- ``eth``: match Ethernet header. 3232 3233 - ``dst {MAC-48}``: destination MAC. 3234 - ``src {MAC-48}``: source MAC. 3235 - ``type {unsigned}``: EtherType. 3236 3237- ``vlan``: match 802.1Q/ad VLAN tag. 3238 3239 - ``tpid {unsigned}``: tag protocol identifier. 3240 - ``tci {unsigned}``: tag control information. 3241 - ``pcp {unsigned}``: priority code point. 3242 - ``dei {unsigned}``: drop eligible indicator. 3243 - ``vid {unsigned}``: VLAN identifier. 3244 3245- ``ipv4``: match IPv4 header. 3246 3247 - ``tos {unsigned}``: type of service. 3248 - ``ttl {unsigned}``: time to live. 3249 - ``proto {unsigned}``: next protocol ID. 3250 - ``src {ipv4 address}``: source address. 3251 - ``dst {ipv4 address}``: destination address. 3252 3253- ``ipv6``: match IPv6 header. 3254 3255 - ``tc {unsigned}``: traffic class. 3256 - ``flow {unsigned}``: flow label. 3257 - ``proto {unsigned}``: protocol (next header). 3258 - ``hop {unsigned}``: hop limit. 3259 - ``src {ipv6 address}``: source address. 3260 - ``dst {ipv6 address}``: destination address. 3261 3262- ``icmp``: match ICMP header. 3263 3264 - ``type {unsigned}``: ICMP packet type. 3265 - ``code {unsigned}``: ICMP packet code. 3266 3267- ``udp``: match UDP header. 3268 3269 - ``src {unsigned}``: UDP source port. 3270 - ``dst {unsigned}``: UDP destination port. 3271 3272- ``tcp``: match TCP header. 3273 3274 - ``src {unsigned}``: TCP source port. 3275 - ``dst {unsigned}``: TCP destination port. 3276 3277- ``sctp``: match SCTP header. 3278 3279 - ``src {unsigned}``: SCTP source port. 3280 - ``dst {unsigned}``: SCTP destination port. 3281 - ``tag {unsigned}``: validation tag. 3282 - ``cksum {unsigned}``: checksum. 3283 3284- ``vxlan``: match VXLAN header. 3285 3286 - ``vni {unsigned}``: VXLAN identifier. 3287 3288- ``e_tag``: match IEEE 802.1BR E-Tag header. 3289 3290 - ``grp_ecid_b {unsigned}``: GRP and E-CID base. 3291 3292- ``nvgre``: match NVGRE header. 3293 3294 - ``tni {unsigned}``: virtual subnet ID. 3295 3296- ``mpls``: match MPLS header. 3297 3298 - ``label {unsigned}``: MPLS label. 3299 3300- ``gre``: match GRE header. 3301 3302 - ``protocol {unsigned}``: protocol type. 3303 3304- ``fuzzy``: fuzzy pattern match, expect faster than default. 3305 3306 - ``thresh {unsigned}``: accuracy threshold. 3307 3308- ``gtp``, ``gtpc``, ``gtpu``: match GTPv1 header. 3309 3310 - ``teid {unsigned}``: tunnel endpoint identifier. 3311 3312- ``geneve``: match GENEVE header. 3313 3314 - ``vni {unsigned}``: virtual network identifier. 3315 - ``protocol {unsigned}``: protocol type. 3316 3317Actions list 3318^^^^^^^^^^^^ 3319 3320A list of actions starts after the ``actions`` token in the same fashion as 3321`Matching pattern`_; actions are separated by ``/`` tokens and the list is 3322terminated by a mandatory ``end`` action. 3323 3324Actions are named after their type (*RTE_FLOW_ACTION_TYPE_* from ``enum 3325rte_flow_action_type``). 3326 3327Dropping all incoming UDPv4 packets can be expressed as follows:: 3328 3329 testpmd> flow create 0 ingress pattern eth / ipv4 / udp / end 3330 actions drop / end 3331 3332Several actions have configurable properties which must be specified when 3333there is no valid default value. For example, ``queue`` requires a target 3334queue index. 3335 3336This rule redirects incoming UDPv4 traffic to queue index 6:: 3337 3338 testpmd> flow create 0 ingress pattern eth / ipv4 / udp / end 3339 actions queue index 6 / end 3340 3341While this one could be rejected by PMDs (unspecified queue index):: 3342 3343 testpmd> flow create 0 ingress pattern eth / ipv4 / udp / end 3344 actions queue / end 3345 3346As defined by *rte_flow*, the list is not ordered, all actions of a given 3347rule are performed simultaneously. These are equivalent:: 3348 3349 queue index 6 / void / mark id 42 / end 3350 3351:: 3352 3353 void / mark id 42 / queue index 6 / end 3354 3355All actions in a list should have different types, otherwise only the last 3356action of a given type is taken into account:: 3357 3358 queue index 4 / queue index 5 / queue index 6 / end # will use queue 6 3359 3360:: 3361 3362 drop / drop / drop / end # drop is performed only once 3363 3364:: 3365 3366 mark id 42 / queue index 3 / mark id 24 / end # mark will be 24 3367 3368Considering they are performed simultaneously, opposite and overlapping 3369actions can sometimes be combined when the end result is unambiguous:: 3370 3371 drop / queue index 6 / end # drop has no effect 3372 3373:: 3374 3375 drop / dup index 6 / end # same as above 3376 3377:: 3378 3379 queue index 6 / rss queues 6 7 8 / end # queue has no effect 3380 3381:: 3382 3383 drop / passthru / end # drop has no effect 3384 3385Note that PMDs may still refuse such combinations. 3386 3387Actions 3388^^^^^^^ 3389 3390This section lists supported actions and their attributes, if any. 3391 3392- ``end``: end list of actions. 3393 3394- ``void``: no-op action. 3395 3396- ``passthru``: let subsequent rule process matched packets. 3397 3398- ``mark``: attach 32 bit value to packets. 3399 3400 - ``id {unsigned}``: 32 bit value to return with packets. 3401 3402- ``flag``: flag packets. 3403 3404- ``queue``: assign packets to a given queue index. 3405 3406 - ``index {unsigned}``: queue index to use. 3407 3408- ``drop``: drop packets (note: passthru has priority). 3409 3410- ``count``: enable counters for this rule. 3411 3412- ``dup``: duplicate packets to a given queue index. 3413 3414 - ``index {unsigned}``: queue index to duplicate packets to. 3415 3416- ``rss``: spread packets among several queues. 3417 3418 - ``types [{RSS hash type} [...]] end``: RSS hash types, allowed tokens 3419 are the same as `set_hash_input_set`_, an empty list means none (0). 3420 3421 - ``key {string}``: RSS hash key, overrides ``key_len``. 3422 3423 - ``key_len {unsigned}``: RSS hash key length in bytes, can be used in 3424 conjunction with ``key`` to pad or truncate it. 3425 3426 - ``queues [{unsigned} [...]] end``: queue indices to use. 3427 3428- ``pf``: redirect packets to physical device function. 3429 3430- ``vf``: redirect packets to virtual device function. 3431 3432 - ``original {boolean}``: use original VF ID if possible. 3433 - ``id {unsigned}``: VF ID to redirect packets to. 3434 3435Destroying flow rules 3436~~~~~~~~~~~~~~~~~~~~~ 3437 3438``flow destroy`` destroys one or more rules from their rule ID (as returned 3439by ``flow create``), this command calls ``rte_flow_destroy()`` as many 3440times as necessary:: 3441 3442 flow destroy {port_id} rule {rule_id} [...] 3443 3444If successful, it will show:: 3445 3446 Flow rule #[...] destroyed 3447 3448It does not report anything for rule IDs that do not exist. The usual error 3449message is shown when a rule cannot be destroyed:: 3450 3451 Caught error type [...] ([...]): [...] 3452 3453``flow flush`` destroys all rules on a device and does not take extra 3454arguments. It is bound to ``rte_flow_flush()``:: 3455 3456 flow flush {port_id} 3457 3458Any errors are reported as above. 3459 3460Creating several rules and destroying them:: 3461 3462 testpmd> flow create 0 ingress pattern eth / ipv6 / end 3463 actions queue index 2 / end 3464 Flow rule #0 created 3465 testpmd> flow create 0 ingress pattern eth / ipv4 / end 3466 actions queue index 3 / end 3467 Flow rule #1 created 3468 testpmd> flow destroy 0 rule 0 rule 1 3469 Flow rule #1 destroyed 3470 Flow rule #0 destroyed 3471 testpmd> 3472 3473The same result can be achieved using ``flow flush``:: 3474 3475 testpmd> flow create 0 ingress pattern eth / ipv6 / end 3476 actions queue index 2 / end 3477 Flow rule #0 created 3478 testpmd> flow create 0 ingress pattern eth / ipv4 / end 3479 actions queue index 3 / end 3480 Flow rule #1 created 3481 testpmd> flow flush 0 3482 testpmd> 3483 3484Non-existent rule IDs are ignored:: 3485 3486 testpmd> flow create 0 ingress pattern eth / ipv6 / end 3487 actions queue index 2 / end 3488 Flow rule #0 created 3489 testpmd> flow create 0 ingress pattern eth / ipv4 / end 3490 actions queue index 3 / end 3491 Flow rule #1 created 3492 testpmd> flow destroy 0 rule 42 rule 10 rule 2 3493 testpmd> 3494 testpmd> flow destroy 0 rule 0 3495 Flow rule #0 destroyed 3496 testpmd> 3497 3498Querying flow rules 3499~~~~~~~~~~~~~~~~~~~ 3500 3501``flow query`` queries a specific action of a flow rule having that 3502ability. Such actions collect information that can be reported using this 3503command. It is bound to ``rte_flow_query()``:: 3504 3505 flow query {port_id} {rule_id} {action} 3506 3507If successful, it will display either the retrieved data for known actions 3508or the following message:: 3509 3510 Cannot display result for action type [...] ([...]) 3511 3512Otherwise, it will complain either that the rule does not exist or that some 3513error occurred:: 3514 3515 Flow rule #[...] not found 3516 3517:: 3518 3519 Caught error type [...] ([...]): [...] 3520 3521Currently only the ``count`` action is supported. This action reports the 3522number of packets that hit the flow rule and the total number of bytes. Its 3523output has the following format:: 3524 3525 count: 3526 hits_set: [...] # whether "hits" contains a valid value 3527 bytes_set: [...] # whether "bytes" contains a valid value 3528 hits: [...] # number of packets 3529 bytes: [...] # number of bytes 3530 3531Querying counters for TCPv6 packets redirected to queue 6:: 3532 3533 testpmd> flow create 0 ingress pattern eth / ipv6 / tcp / end 3534 actions queue index 6 / count / end 3535 Flow rule #4 created 3536 testpmd> flow query 0 4 count 3537 count: 3538 hits_set: 1 3539 bytes_set: 0 3540 hits: 386446 3541 bytes: 0 3542 testpmd> 3543 3544Listing flow rules 3545~~~~~~~~~~~~~~~~~~ 3546 3547``flow list`` lists existing flow rules sorted by priority and optionally 3548filtered by group identifiers:: 3549 3550 flow list {port_id} [group {group_id}] [...] 3551 3552This command only fails with the following message if the device does not 3553exist:: 3554 3555 Invalid port [...] 3556 3557Output consists of a header line followed by a short description of each 3558flow rule, one per line. There is no output at all when no flow rules are 3559configured on the device:: 3560 3561 ID Group Prio Attr Rule 3562 [...] [...] [...] [...] [...] 3563 3564``Attr`` column flags: 3565 3566- ``i`` for ``ingress``. 3567- ``e`` for ``egress``. 3568 3569Creating several flow rules and listing them:: 3570 3571 testpmd> flow create 0 ingress pattern eth / ipv4 / end 3572 actions queue index 6 / end 3573 Flow rule #0 created 3574 testpmd> flow create 0 ingress pattern eth / ipv6 / end 3575 actions queue index 2 / end 3576 Flow rule #1 created 3577 testpmd> flow create 0 priority 5 ingress pattern eth / ipv4 / udp / end 3578 actions rss queues 6 7 8 end / end 3579 Flow rule #2 created 3580 testpmd> flow list 0 3581 ID Group Prio Attr Rule 3582 0 0 0 i- ETH IPV4 => QUEUE 3583 1 0 0 i- ETH IPV6 => QUEUE 3584 2 0 5 i- ETH IPV4 UDP => RSS 3585 testpmd> 3586 3587Rules are sorted by priority (i.e. group ID first, then priority level):: 3588 3589 testpmd> flow list 1 3590 ID Group Prio Attr Rule 3591 0 0 0 i- ETH => COUNT 3592 6 0 500 i- ETH IPV6 TCP => DROP COUNT 3593 5 0 1000 i- ETH IPV6 ICMP => QUEUE 3594 1 24 0 i- ETH IPV4 UDP => QUEUE 3595 4 24 10 i- ETH IPV4 TCP => DROP 3596 3 24 20 i- ETH IPV4 => DROP 3597 2 24 42 i- ETH IPV4 UDP => QUEUE 3598 7 63 0 i- ETH IPV6 UDP VXLAN => MARK QUEUE 3599 testpmd> 3600 3601Output can be limited to specific groups:: 3602 3603 testpmd> flow list 1 group 0 group 63 3604 ID Group Prio Attr Rule 3605 0 0 0 i- ETH => COUNT 3606 6 0 500 i- ETH IPV6 TCP => DROP COUNT 3607 5 0 1000 i- ETH IPV6 ICMP => QUEUE 3608 7 63 0 i- ETH IPV6 UDP VXLAN => MARK QUEUE 3609 testpmd> 3610 3611Toggling isolated mode 3612~~~~~~~~~~~~~~~~~~~~~~ 3613 3614``flow isolate`` can be used to tell the underlying PMD that ingress traffic 3615must only be injected from the defined flow rules; that no default traffic 3616is expected outside those rules and the driver is free to assign more 3617resources to handle them. It is bound to ``rte_flow_isolate()``:: 3618 3619 flow isolate {port_id} {boolean} 3620 3621If successful, enabling or disabling isolated mode shows either:: 3622 3623 Ingress traffic on port [...] 3624 is now restricted to the defined flow rules 3625 3626Or:: 3627 3628 Ingress traffic on port [...] 3629 is not restricted anymore to the defined flow rules 3630 3631Otherwise, in case of error:: 3632 3633 Caught error type [...] ([...]): [...] 3634 3635Mainly due to its side effects, PMDs supporting this mode may not have the 3636ability to toggle it more than once without reinitializing affected ports 3637first (e.g. by exiting testpmd). 3638 3639Enabling isolated mode:: 3640 3641 testpmd> flow isolate 0 true 3642 Ingress traffic on port 0 is now restricted to the defined flow rules 3643 testpmd> 3644 3645Disabling isolated mode:: 3646 3647 testpmd> flow isolate 0 false 3648 Ingress traffic on port 0 is not restricted anymore to the defined flow rules 3649 testpmd> 3650 3651Sample QinQ flow rules 3652~~~~~~~~~~~~~~~~~~~~~~ 3653 3654Before creating QinQ rule(s) the following commands should be issued to enable QinQ:: 3655 3656 testpmd> port stop 0 3657 testpmd> vlan set qinq on 0 3658 3659The above command sets the inner and outer TPID's to 0x8100. 3660 3661To change the TPID's the following commands should be used:: 3662 3663 testpmd> vlan set outer tpid 0xa100 0 3664 testpmd> vlan set inner tpid 0x9100 0 3665 testpmd> port start 0 3666 3667Validate and create a QinQ rule on port 0 to steer traffic to a VF queue in a VM. 3668 3669:: 3670 3671 testpmd> flow validate 0 ingress pattern eth / vlan tci is 123 / 3672 vlan tci is 456 / end actions vf id 1 / queue index 0 / end 3673 Flow rule #0 validated 3674 3675 testpmd> flow create 0 ingress pattern eth / vlan tci is 4 / 3676 vlan tci is 456 / end actions vf id 123 / queue index 0 / end 3677 Flow rule #0 created 3678 3679 testpmd> flow list 0 3680 ID Group Prio Attr Rule 3681 0 0 0 i- ETH VLAN VLAN=>VF QUEUE 3682 3683Validate and create a QinQ rule on port 0 to steer traffic to a queue on the host. 3684 3685:: 3686 3687 testpmd> flow validate 0 ingress pattern eth / vlan tci is 321 / 3688 vlan tci is 654 / end actions pf / queue index 0 / end 3689 Flow rule #1 validated 3690 3691 testpmd> flow create 0 ingress pattern eth / vlan tci is 321 / 3692 vlan tci is 654 / end actions pf / queue index 1 / end 3693 Flow rule #1 created 3694 3695 testpmd> flow list 0 3696 ID Group Prio Attr Rule 3697 0 0 0 i- ETH VLAN VLAN=>VF QUEUE 3698 1 0 0 i- ETH VLAN VLAN=>PF QUEUE 3699