1.. SPDX-License-Identifier: BSD-3-Clause 2 Copyright(c) 2010-2016 Intel Corporation. 3 4.. _testpmd_runtime: 5 6Testpmd Runtime Functions 7========================= 8 9Where the testpmd application is started in interactive mode, (``-i|--interactive``), 10it displays a prompt that can be used to start and stop forwarding, 11configure the application, display statistics (including the extended NIC 12statistics aka xstats) , set the Flow Director and other tasks:: 13 14 testpmd> 15 16The testpmd prompt has some, limited, readline support. 17Common bash command-line functions such as ``Ctrl+a`` and ``Ctrl+e`` to go to the start and end of the prompt line are supported 18as well as access to the command history via the up-arrow. 19 20There is also support for tab completion. 21If you type a partial command and hit ``<TAB>`` you get a list of the available completions: 22 23.. code-block:: console 24 25 testpmd> show port <TAB> 26 27 info [Mul-choice STRING]: show|clear port info|stats|xstats|fdir|stat_qmap|dcb_tc|cap X 28 info [Mul-choice STRING]: show|clear port info|stats|xstats|fdir|stat_qmap|dcb_tc|cap all 29 stats [Mul-choice STRING]: show|clear port info|stats|xstats|fdir|stat_qmap|dcb_tc|cap X 30 stats [Mul-choice STRING]: show|clear port info|stats|xstats|fdir|stat_qmap|dcb_tc|cap all 31 ... 32 33 34.. note:: 35 36 Some examples in this document are too long to fit on one line are are shown wrapped at `"\\"` for display purposes:: 37 38 testpmd> set flow_ctrl rx (on|off) tx (on|off) (high_water) (low_water) \ 39 (pause_time) (send_xon) (port_id) 40 41In the real ``testpmd>`` prompt these commands should be on a single line. 42 43Help Functions 44-------------- 45 46The testpmd has on-line help for the functions that are available at runtime. 47These are divided into sections and can be accessed using help, help section or help all: 48 49.. code-block:: console 50 51 testpmd> help 52 53 help control : Start and stop forwarding. 54 help display : Displaying port, stats and config information. 55 help config : Configuration information. 56 help ports : Configuring ports. 57 help registers : Reading and setting port registers. 58 help filters : Filters configuration help. 59 help all : All of the above sections. 60 61 62Command File Functions 63---------------------- 64 65To facilitate loading large number of commands or to avoid cutting and pasting where not 66practical or possible testpmd supports alternative methods for executing commands. 67 68* If started with the ``--cmdline-file=FILENAME`` command line argument testpmd 69 will execute all CLI commands contained within the file immediately before 70 starting packet forwarding or entering interactive mode. 71 72.. code-block:: console 73 74 ./testpmd -n4 -r2 ... -- -i --cmdline-file=/home/ubuntu/flow-create-commands.txt 75 Interactive-mode selected 76 CLI commands to be read from /home/ubuntu/flow-create-commands.txt 77 Configuring Port 0 (socket 0) 78 Port 0: 7C:FE:90:CB:74:CE 79 Configuring Port 1 (socket 0) 80 Port 1: 7C:FE:90:CB:74:CA 81 Checking link statuses... 82 Port 0 Link Up - speed 10000 Mbps - full-duplex 83 Port 1 Link Up - speed 10000 Mbps - full-duplex 84 Done 85 Flow rule #0 created 86 Flow rule #1 created 87 ... 88 ... 89 Flow rule #498 created 90 Flow rule #499 created 91 Read all CLI commands from /home/ubuntu/flow-create-commands.txt 92 testpmd> 93 94 95* At run-time additional commands can be loaded in bulk by invoking the ``load FILENAME`` 96 command. 97 98.. code-block:: console 99 100 testpmd> load /home/ubuntu/flow-create-commands.txt 101 Flow rule #0 created 102 Flow rule #1 created 103 ... 104 ... 105 Flow rule #498 created 106 Flow rule #499 created 107 Read all CLI commands from /home/ubuntu/flow-create-commands.txt 108 testpmd> 109 110 111In all cases output from any included command will be displayed as standard output. 112Execution will continue until the end of the file is reached regardless of 113whether any errors occur. The end user must examine the output to determine if 114any failures occurred. 115 116 117Control Functions 118----------------- 119 120start 121~~~~~ 122 123Start packet forwarding with current configuration:: 124 125 testpmd> start 126 127start tx_first 128~~~~~~~~~~~~~~ 129 130Start packet forwarding with current configuration after sending specified number of bursts of packets:: 131 132 testpmd> start tx_first (""|burst_num) 133 134The default burst number is 1 when ``burst_num`` not presented. 135 136stop 137~~~~ 138 139Stop packet forwarding, and display accumulated statistics:: 140 141 testpmd> stop 142 143quit 144~~~~ 145 146Quit to prompt:: 147 148 testpmd> quit 149 150 151Display Functions 152----------------- 153 154The functions in the following sections are used to display information about the 155testpmd configuration or the NIC status. 156 157show port 158~~~~~~~~~ 159 160Display information for a given port or all ports:: 161 162 testpmd> show port (info|stats|xstats|fdir|stat_qmap|dcb_tc|cap) (port_id|all) 163 164The available information categories are: 165 166* ``info``: General port information such as MAC address. 167 168* ``stats``: RX/TX statistics. 169 170* ``xstats``: RX/TX extended NIC statistics. 171 172* ``fdir``: Flow Director information and statistics. 173 174* ``stat_qmap``: Queue statistics mapping. 175 176* ``dcb_tc``: DCB information such as TC mapping. 177 178* ``cap``: Supported offload capabilities. 179 180For example: 181 182.. code-block:: console 183 184 testpmd> show port info 0 185 186 ********************* Infos for port 0 ********************* 187 188 MAC address: XX:XX:XX:XX:XX:XX 189 Connect to socket: 0 190 memory allocation on the socket: 0 191 Link status: up 192 Link speed: 40000 Mbps 193 Link duplex: full-duplex 194 Promiscuous mode: enabled 195 Allmulticast mode: disabled 196 Maximum number of MAC addresses: 64 197 Maximum number of MAC addresses of hash filtering: 0 198 VLAN offload: 199 strip on 200 filter on 201 qinq(extend) off 202 Redirection table size: 512 203 Supported flow types: 204 ipv4-frag 205 ipv4-tcp 206 ipv4-udp 207 ipv4-sctp 208 ipv4-other 209 ipv6-frag 210 ipv6-tcp 211 ipv6-udp 212 ipv6-sctp 213 ipv6-other 214 l2_payload 215 port 216 vxlan 217 geneve 218 nvgre 219 220show port rss reta 221~~~~~~~~~~~~~~~~~~ 222 223Display the rss redirection table entry indicated by masks on port X:: 224 225 testpmd> show port (port_id) rss reta (size) (mask0, mask1...) 226 227size is used to indicate the hardware supported reta size 228 229show port rss-hash 230~~~~~~~~~~~~~~~~~~ 231 232Display the RSS hash functions and RSS hash key of a port:: 233 234 testpmd> show port (port_id) rss-hash ipv4|ipv4-frag|ipv4-tcp|ipv4-udp|ipv4-sctp|ipv4-other|ipv6|ipv6-frag|ipv6-tcp|ipv6-udp|ipv6-sctp|ipv6-other|l2-payload|ipv6-ex|ipv6-tcp-ex|ipv6-udp-ex [key] 235 236clear port 237~~~~~~~~~~ 238 239Clear the port statistics for a given port or for all ports:: 240 241 testpmd> clear port (info|stats|xstats|fdir|stat_qmap) (port_id|all) 242 243For example:: 244 245 testpmd> clear port stats all 246 247show (rxq|txq) 248~~~~~~~~~~~~~~ 249 250Display information for a given port's RX/TX queue:: 251 252 testpmd> show (rxq|txq) info (port_id) (queue_id) 253 254show config 255~~~~~~~~~~~ 256 257Displays the configuration of the application. 258The configuration comes from the command-line, the runtime or the application defaults:: 259 260 testpmd> show config (rxtx|cores|fwd|txpkts) 261 262The available information categories are: 263 264* ``rxtx``: RX/TX configuration items. 265 266* ``cores``: List of forwarding cores. 267 268* ``fwd``: Packet forwarding configuration. 269 270* ``txpkts``: Packets to TX configuration. 271 272For example: 273 274.. code-block:: console 275 276 testpmd> show config rxtx 277 278 io packet forwarding - CRC stripping disabled - packets/burst=16 279 nb forwarding cores=2 - nb forwarding ports=1 280 RX queues=1 - RX desc=128 - RX free threshold=0 281 RX threshold registers: pthresh=8 hthresh=8 wthresh=4 282 TX queues=1 - TX desc=512 - TX free threshold=0 283 TX threshold registers: pthresh=36 hthresh=0 wthresh=0 284 TX RS bit threshold=0 - TXQ flags=0x0 285 286set fwd 287~~~~~~~ 288 289Set the packet forwarding mode:: 290 291 testpmd> set fwd (io|mac|macswap|flowgen| \ 292 rxonly|txonly|csum|icmpecho) (""|retry) 293 294``retry`` can be specified for forwarding engines except ``rx_only``. 295 296The available information categories are: 297 298* ``io``: Forwards packets "as-is" in I/O mode. 299 This is the fastest possible forwarding operation as it does not access packets data. 300 This is the default mode. 301 302* ``mac``: Changes the source and the destination Ethernet addresses of packets before forwarding them. 303 Default application behaviour is to set source Ethernet address to that of the transmitting interface, and destination 304 address to a dummy value (set during init). The user may specify a target destination Ethernet address via the 'eth-peer' or 305 'eth-peer-configfile' command-line options. It is not currently possible to specify a specific source Ethernet address. 306 307* ``macswap``: MAC swap forwarding mode. 308 Swaps the source and the destination Ethernet addresses of packets before forwarding them. 309 310* ``flowgen``: Multi-flow generation mode. 311 Originates a number of flows (with varying destination IP addresses), and terminate receive traffic. 312 313* ``rxonly``: Receives packets but doesn't transmit them. 314 315* ``txonly``: Generates and transmits packets without receiving any. 316 317* ``csum``: Changes the checksum field with hardware or software methods depending on the offload flags on the packet. 318 319* ``icmpecho``: Receives a burst of packets, lookup for IMCP echo requests and, if any, send back ICMP echo replies. 320 321* ``ieee1588``: Demonstrate L2 IEEE1588 V2 PTP timestamping for RX and TX. Requires ``CONFIG_RTE_LIBRTE_IEEE1588=y``. 322 323* ``tm``: Traffic Management forwarding mode 324 Demonstrates the use of ethdev traffic management APIs and softnic PMD for 325 QoS traffic management. In this mode, 5-level hierarchical QoS scheduler is 326 available as an default option that can be enabled through CLI. The user can 327 also modify the default hierarchy or specify the new hierarchy through CLI for 328 implementing QoS scheduler. Requires ``CONFIG_RTE_LIBRTE_PMD_SOFTNIC=y`` ``CONFIG_RTE_LIBRTE_SCHED=y``. 329 330Example:: 331 332 testpmd> set fwd rxonly 333 334 Set rxonly packet forwarding mode 335 336 337read rxd 338~~~~~~~~ 339 340Display an RX descriptor for a port RX queue:: 341 342 testpmd> read rxd (port_id) (queue_id) (rxd_id) 343 344For example:: 345 346 testpmd> read rxd 0 0 4 347 0x0000000B - 0x001D0180 / 0x0000000B - 0x001D0180 348 349read txd 350~~~~~~~~ 351 352Display a TX descriptor for a port TX queue:: 353 354 testpmd> read txd (port_id) (queue_id) (txd_id) 355 356For example:: 357 358 testpmd> read txd 0 0 4 359 0x00000001 - 0x24C3C440 / 0x000F0000 - 0x2330003C 360 361ddp get list 362~~~~~~~~~~~~ 363 364Get loaded dynamic device personalization (DDP) package info list:: 365 366 testpmd> ddp get list (port_id) 367 368ddp get info 369~~~~~~~~~~~~ 370 371Display information about dynamic device personalization (DDP) profile:: 372 373 testpmd> ddp get info (profile_path) 374 375show vf stats 376~~~~~~~~~~~~~ 377 378Display VF statistics:: 379 380 testpmd> show vf stats (port_id) (vf_id) 381 382clear vf stats 383~~~~~~~~~~~~~~ 384 385Reset VF statistics:: 386 387 testpmd> clear vf stats (port_id) (vf_id) 388 389show port pctype mapping 390~~~~~~~~~~~~~~~~~~~~~~~~ 391 392List all items from the pctype mapping table:: 393 394 testpmd> show port (port_id) pctype mapping 395 396 397Configuration Functions 398----------------------- 399 400The testpmd application can be configured from the runtime as well as from the command-line. 401 402This section details the available configuration functions that are available. 403 404.. note:: 405 406 Configuration changes only become active when forwarding is started/restarted. 407 408set default 409~~~~~~~~~~~ 410 411Reset forwarding to the default configuration:: 412 413 testpmd> set default 414 415set verbose 416~~~~~~~~~~~ 417 418Set the debug verbosity level:: 419 420 testpmd> set verbose (level) 421 422Currently the only available levels are 0 (silent except for error) and 1 (fully verbose). 423 424set log 425~~~~~~~ 426 427Set the log level for a log type:: 428 429 testpmd> set log global|(type) (level) 430 431Where: 432 433* ``type`` is the log name. 434 435* ``level`` is the log level. 436 437For example, to change the global log level:: 438 testpmd> set log global (level) 439 440Regexes can also be used for type. To change log level of user1, user2 and user3:: 441 testpmd> set log user[1-3] (level) 442 443set nbport 444~~~~~~~~~~ 445 446Set the number of ports used by the application: 447 448set nbport (num) 449 450This is equivalent to the ``--nb-ports`` command-line option. 451 452set nbcore 453~~~~~~~~~~ 454 455Set the number of cores used by the application:: 456 457 testpmd> set nbcore (num) 458 459This is equivalent to the ``--nb-cores`` command-line option. 460 461.. note:: 462 463 The number of cores used must not be greater than number of ports used multiplied by the number of queues per port. 464 465set coremask 466~~~~~~~~~~~~ 467 468Set the forwarding cores hexadecimal mask:: 469 470 testpmd> set coremask (mask) 471 472This is equivalent to the ``--coremask`` command-line option. 473 474.. note:: 475 476 The master lcore is reserved for command line parsing only and cannot be masked on for packet forwarding. 477 478set portmask 479~~~~~~~~~~~~ 480 481Set the forwarding ports hexadecimal mask:: 482 483 testpmd> set portmask (mask) 484 485This is equivalent to the ``--portmask`` command-line option. 486 487set burst 488~~~~~~~~~ 489 490Set number of packets per burst:: 491 492 testpmd> set burst (num) 493 494This is equivalent to the ``--burst command-line`` option. 495 496When retry is enabled, the transmit delay time and number of retries can also be set:: 497 498 testpmd> set burst tx delay (microseconds) retry (num) 499 500set txpkts 501~~~~~~~~~~ 502 503Set the length of each segment of the TX-ONLY packets or length of packet for FLOWGEN mode:: 504 505 testpmd> set txpkts (x[,y]*) 506 507Where x[,y]* represents a CSV list of values, without white space. 508 509set txsplit 510~~~~~~~~~~~ 511 512Set the split policy for the TX packets, applicable for TX-ONLY and CSUM forwarding modes:: 513 514 testpmd> set txsplit (off|on|rand) 515 516Where: 517 518* ``off`` disable packet copy & split for CSUM mode. 519 520* ``on`` split outgoing packet into multiple segments. Size of each segment 521 and number of segments per packet is determined by ``set txpkts`` command 522 (see above). 523 524* ``rand`` same as 'on', but number of segments per each packet is a random value between 1 and total number of segments. 525 526set corelist 527~~~~~~~~~~~~ 528 529Set the list of forwarding cores:: 530 531 testpmd> set corelist (x[,y]*) 532 533For example, to change the forwarding cores: 534 535.. code-block:: console 536 537 testpmd> set corelist 3,1 538 testpmd> show config fwd 539 540 io packet forwarding - ports=2 - cores=2 - streams=2 - NUMA support disabled 541 Logical Core 3 (socket 0) forwards packets on 1 streams: 542 RX P=0/Q=0 (socket 0) -> TX P=1/Q=0 (socket 0) peer=02:00:00:00:00:01 543 Logical Core 1 (socket 0) forwards packets on 1 streams: 544 RX P=1/Q=0 (socket 0) -> TX P=0/Q=0 (socket 0) peer=02:00:00:00:00:00 545 546.. note:: 547 548 The cores are used in the same order as specified on the command line. 549 550set portlist 551~~~~~~~~~~~~ 552 553Set the list of forwarding ports:: 554 555 testpmd> set portlist (x[,y]*) 556 557For example, to change the port forwarding: 558 559.. code-block:: console 560 561 testpmd> set portlist 0,2,1,3 562 testpmd> show config fwd 563 564 io packet forwarding - ports=4 - cores=1 - streams=4 565 Logical Core 3 (socket 0) forwards packets on 4 streams: 566 RX P=0/Q=0 (socket 0) -> TX P=2/Q=0 (socket 0) peer=02:00:00:00:00:01 567 RX P=2/Q=0 (socket 0) -> TX P=0/Q=0 (socket 0) peer=02:00:00:00:00:00 568 RX P=1/Q=0 (socket 0) -> TX P=3/Q=0 (socket 0) peer=02:00:00:00:00:03 569 RX P=3/Q=0 (socket 0) -> TX P=1/Q=0 (socket 0) peer=02:00:00:00:00:02 570 571set tx loopback 572~~~~~~~~~~~~~~~ 573 574Enable/disable tx loopback:: 575 576 testpmd> set tx loopback (port_id) (on|off) 577 578set drop enable 579~~~~~~~~~~~~~~~ 580 581set drop enable bit for all queues:: 582 583 testpmd> set all queues drop (port_id) (on|off) 584 585set split drop enable (for VF) 586~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ 587 588set split drop enable bit for VF from PF:: 589 590 testpmd> set vf split drop (port_id) (vf_id) (on|off) 591 592set mac antispoof (for VF) 593~~~~~~~~~~~~~~~~~~~~~~~~~~ 594 595Set mac antispoof for a VF from the PF:: 596 597 testpmd> set vf mac antispoof (port_id) (vf_id) (on|off) 598 599set macsec offload 600~~~~~~~~~~~~~~~~~~ 601 602Enable/disable MACsec offload:: 603 604 testpmd> set macsec offload (port_id) on encrypt (on|off) replay-protect (on|off) 605 testpmd> set macsec offload (port_id) off 606 607set macsec sc 608~~~~~~~~~~~~~ 609 610Configure MACsec secure connection (SC):: 611 612 testpmd> set macsec sc (tx|rx) (port_id) (mac) (pi) 613 614.. note:: 615 616 The pi argument is ignored for tx. 617 Check the NIC Datasheet for hardware limits. 618 619set macsec sa 620~~~~~~~~~~~~~ 621 622Configure MACsec secure association (SA):: 623 624 testpmd> set macsec sa (tx|rx) (port_id) (idx) (an) (pn) (key) 625 626.. note:: 627 628 The IDX value must be 0 or 1. 629 Check the NIC Datasheet for hardware limits. 630 631set broadcast mode (for VF) 632~~~~~~~~~~~~~~~~~~~~~~~~~~~ 633 634Set broadcast mode for a VF from the PF:: 635 636 testpmd> set vf broadcast (port_id) (vf_id) (on|off) 637 638vlan set strip 639~~~~~~~~~~~~~~ 640 641Set the VLAN strip on a port:: 642 643 testpmd> vlan set strip (on|off) (port_id) 644 645vlan set stripq 646~~~~~~~~~~~~~~~ 647 648Set the VLAN strip for a queue on a port:: 649 650 testpmd> vlan set stripq (on|off) (port_id,queue_id) 651 652vlan set stripq (for VF) 653~~~~~~~~~~~~~~~~~~~~~~~~ 654 655Set VLAN strip for all queues in a pool for a VF from the PF:: 656 657 testpmd> set vf vlan stripq (port_id) (vf_id) (on|off) 658 659vlan set insert (for VF) 660~~~~~~~~~~~~~~~~~~~~~~~~ 661 662Set VLAN insert for a VF from the PF:: 663 664 testpmd> set vf vlan insert (port_id) (vf_id) (vlan_id) 665 666vlan set tag (for VF) 667~~~~~~~~~~~~~~~~~~~~~ 668 669Set VLAN tag for a VF from the PF:: 670 671 testpmd> set vf vlan tag (port_id) (vf_id) (on|off) 672 673vlan set antispoof (for VF) 674~~~~~~~~~~~~~~~~~~~~~~~~~~~ 675 676Set VLAN antispoof for a VF from the PF:: 677 678 testpmd> set vf vlan antispoof (port_id) (vf_id) (on|off) 679 680vlan set filter 681~~~~~~~~~~~~~~~ 682 683Set the VLAN filter on a port:: 684 685 testpmd> vlan set filter (on|off) (port_id) 686 687vlan set qinq 688~~~~~~~~~~~~~ 689 690Set the VLAN QinQ (extended queue in queue) on for a port:: 691 692 testpmd> vlan set qinq (on|off) (port_id) 693 694vlan set tpid 695~~~~~~~~~~~~~ 696 697Set the inner or outer VLAN TPID for packet filtering on a port:: 698 699 testpmd> vlan set (inner|outer) tpid (value) (port_id) 700 701.. note:: 702 703 TPID value must be a 16-bit number (value <= 65536). 704 705rx_vlan add 706~~~~~~~~~~~ 707 708Add a VLAN ID, or all identifiers, to the set of VLAN identifiers filtered by port ID:: 709 710 testpmd> rx_vlan add (vlan_id|all) (port_id) 711 712.. note:: 713 714 VLAN filter must be set on that port. VLAN ID < 4096. 715 Depending on the NIC used, number of vlan_ids may be limited to the maximum entries 716 in VFTA table. This is important if enabling all vlan_ids. 717 718rx_vlan rm 719~~~~~~~~~~ 720 721Remove a VLAN ID, or all identifiers, from the set of VLAN identifiers filtered by port ID:: 722 723 testpmd> rx_vlan rm (vlan_id|all) (port_id) 724 725rx_vlan add (for VF) 726~~~~~~~~~~~~~~~~~~~~ 727 728Add a VLAN ID, to the set of VLAN identifiers filtered for VF(s) for port ID:: 729 730 testpmd> rx_vlan add (vlan_id) port (port_id) vf (vf_mask) 731 732rx_vlan rm (for VF) 733~~~~~~~~~~~~~~~~~~~ 734 735Remove a VLAN ID, from the set of VLAN identifiers filtered for VF(s) for port ID:: 736 737 testpmd> rx_vlan rm (vlan_id) port (port_id) vf (vf_mask) 738 739tunnel_filter add 740~~~~~~~~~~~~~~~~~ 741 742Add a tunnel filter on a port:: 743 744 testpmd> tunnel_filter add (port_id) (outer_mac) (inner_mac) (ip_addr) \ 745 (inner_vlan) (vxlan|nvgre|ipingre) (imac-ivlan|imac-ivlan-tenid|\ 746 imac-tenid|imac|omac-imac-tenid|oip|iip) (tenant_id) (queue_id) 747 748The available information categories are: 749 750* ``vxlan``: Set tunnel type as VXLAN. 751 752* ``nvgre``: Set tunnel type as NVGRE. 753 754* ``ipingre``: Set tunnel type as IP-in-GRE. 755 756* ``imac-ivlan``: Set filter type as Inner MAC and VLAN. 757 758* ``imac-ivlan-tenid``: Set filter type as Inner MAC, VLAN and tenant ID. 759 760* ``imac-tenid``: Set filter type as Inner MAC and tenant ID. 761 762* ``imac``: Set filter type as Inner MAC. 763 764* ``omac-imac-tenid``: Set filter type as Outer MAC, Inner MAC and tenant ID. 765 766* ``oip``: Set filter type as Outer IP. 767 768* ``iip``: Set filter type as Inner IP. 769 770Example:: 771 772 testpmd> tunnel_filter add 0 68:05:CA:28:09:82 00:00:00:00:00:00 \ 773 192.168.2.2 0 ipingre oip 1 1 774 775 Set an IP-in-GRE tunnel on port 0, and the filter type is Outer IP. 776 777tunnel_filter remove 778~~~~~~~~~~~~~~~~~~~~ 779 780Remove a tunnel filter on a port:: 781 782 testpmd> tunnel_filter rm (port_id) (outer_mac) (inner_mac) (ip_addr) \ 783 (inner_vlan) (vxlan|nvgre|ipingre) (imac-ivlan|imac-ivlan-tenid|\ 784 imac-tenid|imac|omac-imac-tenid|oip|iip) (tenant_id) (queue_id) 785 786rx_vxlan_port add 787~~~~~~~~~~~~~~~~~ 788 789Add an UDP port for VXLAN packet filter on a port:: 790 791 testpmd> rx_vxlan_port add (udp_port) (port_id) 792 793rx_vxlan_port remove 794~~~~~~~~~~~~~~~~~~~~ 795 796Remove an UDP port for VXLAN packet filter on a port:: 797 798 testpmd> rx_vxlan_port rm (udp_port) (port_id) 799 800tx_vlan set 801~~~~~~~~~~~ 802 803Set hardware insertion of VLAN IDs in packets sent on a port:: 804 805 testpmd> tx_vlan set (port_id) vlan_id[, vlan_id_outer] 806 807For example, set a single VLAN ID (5) insertion on port 0:: 808 809 tx_vlan set 0 5 810 811Or, set double VLAN ID (inner: 2, outer: 3) insertion on port 1:: 812 813 tx_vlan set 1 2 3 814 815 816tx_vlan set pvid 817~~~~~~~~~~~~~~~~ 818 819Set port based hardware insertion of VLAN ID in packets sent on a port:: 820 821 testpmd> tx_vlan set pvid (port_id) (vlan_id) (on|off) 822 823tx_vlan reset 824~~~~~~~~~~~~~ 825 826Disable hardware insertion of a VLAN header in packets sent on a port:: 827 828 testpmd> tx_vlan reset (port_id) 829 830csum set 831~~~~~~~~ 832 833Select hardware or software calculation of the checksum when 834transmitting a packet using the ``csum`` forwarding engine:: 835 836 testpmd> csum set (ip|udp|tcp|sctp|outer-ip) (hw|sw) (port_id) 837 838Where: 839 840* ``ip|udp|tcp|sctp`` always relate to the inner layer. 841 842* ``outer-ip`` relates to the outer IP layer (only for IPv4) in the case where the packet is recognized 843 as a tunnel packet by the forwarding engine (vxlan, gre and ipip are 844 supported). See also the ``csum parse-tunnel`` command. 845 846.. note:: 847 848 Check the NIC Datasheet for hardware limits. 849 850RSS queue region 851~~~~~~~~~~~~~~~~ 852 853Set RSS queue region span on a port:: 854 855 testpmd> set port (port_id) queue-region region_id (value) \ 856 queue_start_index (value) queue_num (value) 857 858Set flowtype mapping on a RSS queue region on a port:: 859 860 testpmd> set port (port_id) queue-region region_id (value) flowtype (value) 861 862where: 863 864* For the flowtype(pctype) of packet,the specific index for each type has 865 been defined in file i40e_type.h as enum i40e_filter_pctype. 866 867Set user priority mapping on a RSS queue region on a port:: 868 869 testpmd> set port (port_id) queue-region UP (value) region_id (value) 870 871Flush all queue region related configuration on a port:: 872 873 testpmd> set port (port_id) queue-region flush (on|off) 874 875where: 876 877* "on"is just an enable function which server for other configuration, 878 it is for all configuration about queue region from up layer, 879 at first will only keep in DPDK softwarestored in driver, 880 only after "flush on", it commit all configuration to HW. 881 "off" is just clean all configuration about queue region just now, 882 and restore all to DPDK i40e driver default config when start up. 883 884Show all queue region related configuration info on a port:: 885 886 testpmd> show port (port_id) queue-region 887 888.. note:: 889 890 Queue region only support on PF by now, so these command is 891 only for configuration of queue region on PF port. 892 893csum parse-tunnel 894~~~~~~~~~~~~~~~~~ 895 896Define how tunneled packets should be handled by the csum forward 897engine:: 898 899 testpmd> csum parse-tunnel (on|off) (tx_port_id) 900 901If enabled, the csum forward engine will try to recognize supported 902tunnel headers (vxlan, gre, ipip). 903 904If disabled, treat tunnel packets as non-tunneled packets (a inner 905header is handled as a packet payload). 906 907.. note:: 908 909 The port argument is the TX port like in the ``csum set`` command. 910 911Example: 912 913Consider a packet in packet like the following:: 914 915 eth_out/ipv4_out/udp_out/vxlan/eth_in/ipv4_in/tcp_in 916 917* If parse-tunnel is enabled, the ``ip|udp|tcp|sctp`` parameters of ``csum set`` 918 command relate to the inner headers (here ``ipv4_in`` and ``tcp_in``), and the 919 ``outer-ip parameter`` relates to the outer headers (here ``ipv4_out``). 920 921* If parse-tunnel is disabled, the ``ip|udp|tcp|sctp`` parameters of ``csum set`` 922 command relate to the outer headers, here ``ipv4_out`` and ``udp_out``. 923 924csum show 925~~~~~~~~~ 926 927Display tx checksum offload configuration:: 928 929 testpmd> csum show (port_id) 930 931tso set 932~~~~~~~ 933 934Enable TCP Segmentation Offload (TSO) in the ``csum`` forwarding engine:: 935 936 testpmd> tso set (segsize) (port_id) 937 938.. note:: 939 940 Check the NIC datasheet for hardware limits. 941 942tso show 943~~~~~~~~ 944 945Display the status of TCP Segmentation Offload:: 946 947 testpmd> tso show (port_id) 948 949set port - gro 950~~~~~~~~~~~~~~ 951 952Enable or disable GRO in ``csum`` forwarding engine:: 953 954 testpmd> set port <port_id> gro on|off 955 956If enabled, the csum forwarding engine will perform GRO on the TCP/IPv4 957packets received from the given port. 958 959If disabled, packets received from the given port won't be performed 960GRO. By default, GRO is disabled for all ports. 961 962.. note:: 963 964 When enable GRO for a port, TCP/IPv4 packets received from the port 965 will be performed GRO. After GRO, all merged packets have bad 966 checksums, since the GRO library doesn't re-calculate checksums for 967 the merged packets. Therefore, if users want the merged packets to 968 have correct checksums, please select HW IP checksum calculation and 969 HW TCP checksum calculation for the port which the merged packets are 970 transmitted to. 971 972show port - gro 973~~~~~~~~~~~~~~~ 974 975Display GRO configuration for a given port:: 976 977 testpmd> show port <port_id> gro 978 979set gro flush 980~~~~~~~~~~~~~ 981 982Set the cycle to flush the GROed packets from reassembly tables:: 983 984 testpmd> set gro flush <cycles> 985 986When enable GRO, the csum forwarding engine performs GRO on received 987packets, and the GROed packets are stored in reassembly tables. Users 988can use this command to determine when the GROed packets are flushed 989from the reassembly tables. 990 991The ``cycles`` is measured in GRO operation times. The csum forwarding 992engine flushes the GROed packets from the tables every ``cycles`` GRO 993operations. 994 995By default, the value of ``cycles`` is 1, which means flush GROed packets 996from the reassembly tables as soon as one GRO operation finishes. The value 997of ``cycles`` should be in the range of 1 to ``GRO_MAX_FLUSH_CYCLES``. 998 999Please note that the large value of ``cycles`` may cause the poor TCP/IP 1000stack performance. Because the GROed packets are delayed to arrive the 1001stack, thus causing more duplicated ACKs and TCP retransmissions. 1002 1003set port - gso 1004~~~~~~~~~~~~~~ 1005 1006Toggle per-port GSO support in ``csum`` forwarding engine:: 1007 1008 testpmd> set port <port_id> gso on|off 1009 1010If enabled, the csum forwarding engine will perform GSO on supported IPv4 1011packets, transmitted on the given port. 1012 1013If disabled, packets transmitted on the given port will not undergo GSO. 1014By default, GSO is disabled for all ports. 1015 1016.. note:: 1017 1018 When GSO is enabled on a port, supported IPv4 packets transmitted on that 1019 port undergo GSO. Afterwards, the segmented packets are represented by 1020 multi-segment mbufs; however, the csum forwarding engine doesn't calculation 1021 of checksums for GSO'd segments in SW. As a result, if users want correct 1022 checksums in GSO segments, they should enable HW checksum calculation for 1023 GSO-enabled ports. 1024 1025 For example, HW checksum calculation for VxLAN GSO'd packets may be enabled 1026 by setting the following options in the csum forwarding engine: 1027 1028 testpmd> csum set outer_ip hw <port_id> 1029 1030 testpmd> csum set ip hw <port_id> 1031 1032 testpmd> csum set tcp hw <port_id> 1033 1034set gso segsz 1035~~~~~~~~~~~~~ 1036 1037Set the maximum GSO segment size (measured in bytes), which includes the 1038packet header and the packet payload for GSO-enabled ports (global):: 1039 1040 testpmd> set gso segsz <length> 1041 1042show port - gso 1043~~~~~~~~~~~~~~~ 1044 1045Display the status of Generic Segmentation Offload for a given port:: 1046 1047 testpmd> show port <port_id> gso 1048 1049mac_addr add 1050~~~~~~~~~~~~ 1051 1052Add an alternative MAC address to a port:: 1053 1054 testpmd> mac_addr add (port_id) (XX:XX:XX:XX:XX:XX) 1055 1056mac_addr remove 1057~~~~~~~~~~~~~~~ 1058 1059Remove a MAC address from a port:: 1060 1061 testpmd> mac_addr remove (port_id) (XX:XX:XX:XX:XX:XX) 1062 1063mac_addr add (for VF) 1064~~~~~~~~~~~~~~~~~~~~~ 1065 1066Add an alternative MAC address for a VF to a port:: 1067 1068 testpmd> mac_add add port (port_id) vf (vf_id) (XX:XX:XX:XX:XX:XX) 1069 1070mac_addr set 1071~~~~~~~~~~~~ 1072 1073Set the default MAC address for a port:: 1074 1075 testpmd> mac_addr set (port_id) (XX:XX:XX:XX:XX:XX) 1076 1077mac_addr set (for VF) 1078~~~~~~~~~~~~~~~~~~~~~ 1079 1080Set the MAC address for a VF from the PF:: 1081 1082 testpmd> set vf mac addr (port_id) (vf_id) (XX:XX:XX:XX:XX:XX) 1083 1084set eth-peer 1085~~~~~~~~~~~~ 1086 1087Set the forwarding peer address for certain port:: 1088 1089 testpmd> set eth-peer (port_id) (perr_addr) 1090 1091This is equivalent to the ``--eth-peer`` command-line option. 1092 1093set port-uta 1094~~~~~~~~~~~~ 1095 1096Set the unicast hash filter(s) on/off for a port:: 1097 1098 testpmd> set port (port_id) uta (XX:XX:XX:XX:XX:XX|all) (on|off) 1099 1100set promisc 1101~~~~~~~~~~~ 1102 1103Set the promiscuous mode on for a port or for all ports. 1104In promiscuous mode packets are not dropped if they aren't for the specified MAC address:: 1105 1106 testpmd> set promisc (port_id|all) (on|off) 1107 1108set allmulti 1109~~~~~~~~~~~~ 1110 1111Set the allmulti mode for a port or for all ports:: 1112 1113 testpmd> set allmulti (port_id|all) (on|off) 1114 1115Same as the ifconfig (8) option. Controls how multicast packets are handled. 1116 1117set promisc (for VF) 1118~~~~~~~~~~~~~~~~~~~~ 1119 1120Set the unicast promiscuous mode for a VF from PF. 1121It's supported by Intel i40e NICs now. 1122In promiscuous mode packets are not dropped if they aren't for the specified MAC address:: 1123 1124 testpmd> set vf promisc (port_id) (vf_id) (on|off) 1125 1126set allmulticast (for VF) 1127~~~~~~~~~~~~~~~~~~~~~~~~~ 1128 1129Set the multicast promiscuous mode for a VF from PF. 1130It's supported by Intel i40e NICs now. 1131In promiscuous mode packets are not dropped if they aren't for the specified MAC address:: 1132 1133 testpmd> set vf allmulti (port_id) (vf_id) (on|off) 1134 1135set tx max bandwidth (for VF) 1136~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ 1137 1138Set TX max absolute bandwidth (Mbps) for a VF from PF:: 1139 1140 testpmd> set vf tx max-bandwidth (port_id) (vf_id) (max_bandwidth) 1141 1142set tc tx min bandwidth (for VF) 1143~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ 1144 1145Set all TCs' TX min relative bandwidth (%) for a VF from PF:: 1146 1147 testpmd> set vf tc tx min-bandwidth (port_id) (vf_id) (bw1, bw2, ...) 1148 1149set tc tx max bandwidth (for VF) 1150~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ 1151 1152Set a TC's TX max absolute bandwidth (Mbps) for a VF from PF:: 1153 1154 testpmd> set vf tc tx max-bandwidth (port_id) (vf_id) (tc_no) (max_bandwidth) 1155 1156set tc strict link priority mode 1157~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ 1158 1159Set some TCs' strict link priority mode on a physical port:: 1160 1161 testpmd> set tx strict-link-priority (port_id) (tc_bitmap) 1162 1163set tc tx min bandwidth 1164~~~~~~~~~~~~~~~~~~~~~~~ 1165 1166Set all TCs' TX min relative bandwidth (%) globally for all PF and VFs:: 1167 1168 testpmd> set tc tx min-bandwidth (port_id) (bw1, bw2, ...) 1169 1170set flow_ctrl rx 1171~~~~~~~~~~~~~~~~ 1172 1173Set the link flow control parameter on a port:: 1174 1175 testpmd> set flow_ctrl rx (on|off) tx (on|off) (high_water) (low_water) \ 1176 (pause_time) (send_xon) mac_ctrl_frame_fwd (on|off) \ 1177 autoneg (on|off) (port_id) 1178 1179Where: 1180 1181* ``high_water`` (integer): High threshold value to trigger XOFF. 1182 1183* ``low_water`` (integer): Low threshold value to trigger XON. 1184 1185* ``pause_time`` (integer): Pause quota in the Pause frame. 1186 1187* ``send_xon`` (0/1): Send XON frame. 1188 1189* ``mac_ctrl_frame_fwd``: Enable receiving MAC control frames. 1190 1191* ``autoneg``: Change the auto-negotiation parameter. 1192 1193set pfc_ctrl rx 1194~~~~~~~~~~~~~~~ 1195 1196Set the priority flow control parameter on a port:: 1197 1198 testpmd> set pfc_ctrl rx (on|off) tx (on|off) (high_water) (low_water) \ 1199 (pause_time) (priority) (port_id) 1200 1201Where: 1202 1203* ``high_water`` (integer): High threshold value. 1204 1205* ``low_water`` (integer): Low threshold value. 1206 1207* ``pause_time`` (integer): Pause quota in the Pause frame. 1208 1209* ``priority`` (0-7): VLAN User Priority. 1210 1211set stat_qmap 1212~~~~~~~~~~~~~ 1213 1214Set statistics mapping (qmapping 0..15) for RX/TX queue on port:: 1215 1216 testpmd> set stat_qmap (tx|rx) (port_id) (queue_id) (qmapping) 1217 1218For example, to set rx queue 2 on port 0 to mapping 5:: 1219 1220 testpmd>set stat_qmap rx 0 2 5 1221 1222set xstats-hide-zero 1223~~~~~~~~~~~~~~~~~~~~ 1224 1225Set the option to hide zero values for xstats display:: 1226 1227 testpmd> set xstats-hide-zero on|off 1228 1229.. note:: 1230 1231 By default, the zero values are displayed for xstats. 1232 1233set port - rx/tx (for VF) 1234~~~~~~~~~~~~~~~~~~~~~~~~~ 1235 1236Set VF receive/transmit from a port:: 1237 1238 testpmd> set port (port_id) vf (vf_id) (rx|tx) (on|off) 1239 1240set port - mac address filter (for VF) 1241~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ 1242 1243Add/Remove unicast or multicast MAC addr filter for a VF:: 1244 1245 testpmd> set port (port_id) vf (vf_id) (mac_addr) \ 1246 (exact-mac|exact-mac-vlan|hashmac|hashmac-vlan) (on|off) 1247 1248set port - rx mode(for VF) 1249~~~~~~~~~~~~~~~~~~~~~~~~~~ 1250 1251Set the VF receive mode of a port:: 1252 1253 testpmd> set port (port_id) vf (vf_id) \ 1254 rxmode (AUPE|ROPE|BAM|MPE) (on|off) 1255 1256The available receive modes are: 1257 1258* ``AUPE``: Accepts untagged VLAN. 1259 1260* ``ROPE``: Accepts unicast hash. 1261 1262* ``BAM``: Accepts broadcast packets. 1263 1264* ``MPE``: Accepts all multicast packets. 1265 1266set port - tx_rate (for Queue) 1267~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ 1268 1269Set TX rate limitation for a queue on a port:: 1270 1271 testpmd> set port (port_id) queue (queue_id) rate (rate_value) 1272 1273set port - tx_rate (for VF) 1274~~~~~~~~~~~~~~~~~~~~~~~~~~~ 1275 1276Set TX rate limitation for queues in VF on a port:: 1277 1278 testpmd> set port (port_id) vf (vf_id) rate (rate_value) queue_mask (queue_mask) 1279 1280set port - mirror rule 1281~~~~~~~~~~~~~~~~~~~~~~ 1282 1283Set pool or vlan type mirror rule for a port:: 1284 1285 testpmd> set port (port_id) mirror-rule (rule_id) \ 1286 (pool-mirror-up|pool-mirror-down|vlan-mirror) \ 1287 (poolmask|vlanid[,vlanid]*) dst-pool (pool_id) (on|off) 1288 1289Set link mirror rule for a port:: 1290 1291 testpmd> set port (port_id) mirror-rule (rule_id) \ 1292 (uplink-mirror|downlink-mirror) dst-pool (pool_id) (on|off) 1293 1294For example to enable mirror traffic with vlan 0,1 to pool 0:: 1295 1296 set port 0 mirror-rule 0 vlan-mirror 0,1 dst-pool 0 on 1297 1298reset port - mirror rule 1299~~~~~~~~~~~~~~~~~~~~~~~~ 1300 1301Reset a mirror rule for a port:: 1302 1303 testpmd> reset port (port_id) mirror-rule (rule_id) 1304 1305set flush_rx 1306~~~~~~~~~~~~ 1307 1308Set the flush on RX streams before forwarding. 1309The default is flush ``on``. 1310Mainly used with PCAP drivers to turn off the default behavior of flushing the first 512 packets on RX streams:: 1311 1312 testpmd> set flush_rx off 1313 1314set bypass mode 1315~~~~~~~~~~~~~~~ 1316 1317Set the bypass mode for the lowest port on bypass enabled NIC:: 1318 1319 testpmd> set bypass mode (normal|bypass|isolate) (port_id) 1320 1321set bypass event 1322~~~~~~~~~~~~~~~~ 1323 1324Set the event required to initiate specified bypass mode for the lowest port on a bypass enabled:: 1325 1326 testpmd> set bypass event (timeout|os_on|os_off|power_on|power_off) \ 1327 mode (normal|bypass|isolate) (port_id) 1328 1329Where: 1330 1331* ``timeout``: Enable bypass after watchdog timeout. 1332 1333* ``os_on``: Enable bypass when OS/board is powered on. 1334 1335* ``os_off``: Enable bypass when OS/board is powered off. 1336 1337* ``power_on``: Enable bypass when power supply is turned on. 1338 1339* ``power_off``: Enable bypass when power supply is turned off. 1340 1341 1342set bypass timeout 1343~~~~~~~~~~~~~~~~~~ 1344 1345Set the bypass watchdog timeout to ``n`` seconds where 0 = instant:: 1346 1347 testpmd> set bypass timeout (0|1.5|2|3|4|8|16|32) 1348 1349show bypass config 1350~~~~~~~~~~~~~~~~~~ 1351 1352Show the bypass configuration for a bypass enabled NIC using the lowest port on the NIC:: 1353 1354 testpmd> show bypass config (port_id) 1355 1356set link up 1357~~~~~~~~~~~ 1358 1359Set link up for a port:: 1360 1361 testpmd> set link-up port (port id) 1362 1363set link down 1364~~~~~~~~~~~~~ 1365 1366Set link down for a port:: 1367 1368 testpmd> set link-down port (port id) 1369 1370E-tag set 1371~~~~~~~~~ 1372 1373Enable E-tag insertion for a VF on a port:: 1374 1375 testpmd> E-tag set insertion on port-tag-id (value) port (port_id) vf (vf_id) 1376 1377Disable E-tag insertion for a VF on a port:: 1378 1379 testpmd> E-tag set insertion off port (port_id) vf (vf_id) 1380 1381Enable/disable E-tag stripping on a port:: 1382 1383 testpmd> E-tag set stripping (on|off) port (port_id) 1384 1385Enable/disable E-tag based forwarding on a port:: 1386 1387 testpmd> E-tag set forwarding (on|off) port (port_id) 1388 1389Add an E-tag forwarding filter on a port:: 1390 1391 testpmd> E-tag set filter add e-tag-id (value) dst-pool (pool_id) port (port_id) 1392 1393Delete an E-tag forwarding filter on a port:: 1394 testpmd> E-tag set filter del e-tag-id (value) port (port_id) 1395 1396ddp add 1397~~~~~~~ 1398 1399Load a dynamic device personalization (DDP) profile and store backup profile:: 1400 1401 testpmd> ddp add (port_id) (profile_path[,backup_profile_path]) 1402 1403ddp del 1404~~~~~~~ 1405 1406Delete a dynamic device personalization profile and restore backup profile:: 1407 1408 testpmd> ddp del (port_id) (backup_profile_path) 1409 1410ptype mapping 1411~~~~~~~~~~~~~ 1412 1413List all items from the ptype mapping table:: 1414 1415 testpmd> ptype mapping get (port_id) (valid_only) 1416 1417Where: 1418 1419* ``valid_only``: A flag indicates if only list valid items(=1) or all itemss(=0). 1420 1421Replace a specific or a group of software defined ptype with a new one:: 1422 1423 testpmd> ptype mapping replace (port_id) (target) (mask) (pkt_type) 1424 1425where: 1426 1427* ``target``: A specific software ptype or a mask to represent a group of software ptypes. 1428 1429* ``mask``: A flag indicate if "target" is a specific software ptype(=0) or a ptype mask(=1). 1430 1431* ``pkt_type``: The new software ptype to replace the old ones. 1432 1433Update hardware defined ptype to software defined packet type mapping table:: 1434 1435 testpmd> ptype mapping update (port_id) (hw_ptype) (sw_ptype) 1436 1437where: 1438 1439* ``hw_ptype``: hardware ptype as the index of the ptype mapping table. 1440 1441* ``sw_ptype``: software ptype as the value of the ptype mapping table. 1442 1443Reset ptype mapping table:: 1444 1445 testpmd> ptype mapping reset (port_id) 1446 1447Port Functions 1448-------------- 1449 1450The following sections show functions for configuring ports. 1451 1452.. note:: 1453 1454 Port configuration changes only become active when forwarding is started/restarted. 1455 1456port attach 1457~~~~~~~~~~~ 1458 1459Attach a port specified by pci address or virtual device args:: 1460 1461 testpmd> port attach (identifier) 1462 1463To attach a new pci device, the device should be recognized by kernel first. 1464Then it should be moved under DPDK management. 1465Finally the port can be attached to testpmd. 1466 1467For example, to move a pci device using ixgbe under DPDK management: 1468 1469.. code-block:: console 1470 1471 # Check the status of the available devices. 1472 ./usertools/dpdk-devbind.py --status 1473 1474 Network devices using DPDK-compatible driver 1475 ============================================ 1476 <none> 1477 1478 Network devices using kernel driver 1479 =================================== 1480 0000:0a:00.0 '82599ES 10-Gigabit' if=eth2 drv=ixgbe unused= 1481 1482 1483 # Bind the device to igb_uio. 1484 sudo ./usertools/dpdk-devbind.py -b igb_uio 0000:0a:00.0 1485 1486 1487 # Recheck the status of the devices. 1488 ./usertools/dpdk-devbind.py --status 1489 Network devices using DPDK-compatible driver 1490 ============================================ 1491 0000:0a:00.0 '82599ES 10-Gigabit' drv=igb_uio unused= 1492 1493To attach a port created by virtual device, above steps are not needed. 1494 1495For example, to attach a port whose pci address is 0000:0a:00.0. 1496 1497.. code-block:: console 1498 1499 testpmd> port attach 0000:0a:00.0 1500 Attaching a new port... 1501 EAL: PCI device 0000:0a:00.0 on NUMA socket -1 1502 EAL: probe driver: 8086:10fb rte_ixgbe_pmd 1503 EAL: PCI memory mapped at 0x7f83bfa00000 1504 EAL: PCI memory mapped at 0x7f83bfa80000 1505 PMD: eth_ixgbe_dev_init(): MAC: 2, PHY: 18, SFP+: 5 1506 PMD: eth_ixgbe_dev_init(): port 0 vendorID=0x8086 deviceID=0x10fb 1507 Port 0 is attached. Now total ports is 1 1508 Done 1509 1510For example, to attach a port created by pcap PMD. 1511 1512.. code-block:: console 1513 1514 testpmd> port attach net_pcap0 1515 Attaching a new port... 1516 PMD: Initializing pmd_pcap for net_pcap0 1517 PMD: Creating pcap-backed ethdev on numa socket 0 1518 Port 0 is attached. Now total ports is 1 1519 Done 1520 1521In this case, identifier is ``net_pcap0``. 1522This identifier format is the same as ``--vdev`` format of DPDK applications. 1523 1524For example, to re-attach a bonded port which has been previously detached, 1525the mode and slave parameters must be given. 1526 1527.. code-block:: console 1528 1529 testpmd> port attach net_bond_0,mode=0,slave=1 1530 Attaching a new port... 1531 EAL: Initializing pmd_bond for net_bond_0 1532 EAL: Create bonded device net_bond_0 on port 0 in mode 0 on socket 0. 1533 Port 0 is attached. Now total ports is 1 1534 Done 1535 1536 1537port detach 1538~~~~~~~~~~~ 1539 1540Detach a specific port:: 1541 1542 testpmd> port detach (port_id) 1543 1544Before detaching a port, the port should be stopped and closed. 1545 1546For example, to detach a pci device port 0. 1547 1548.. code-block:: console 1549 1550 testpmd> port stop 0 1551 Stopping ports... 1552 Done 1553 testpmd> port close 0 1554 Closing ports... 1555 Done 1556 1557 testpmd> port detach 0 1558 Detaching a port... 1559 EAL: PCI device 0000:0a:00.0 on NUMA socket -1 1560 EAL: remove driver: 8086:10fb rte_ixgbe_pmd 1561 EAL: PCI memory unmapped at 0x7f83bfa00000 1562 EAL: PCI memory unmapped at 0x7f83bfa80000 1563 Done 1564 1565 1566For example, to detach a virtual device port 0. 1567 1568.. code-block:: console 1569 1570 testpmd> port stop 0 1571 Stopping ports... 1572 Done 1573 testpmd> port close 0 1574 Closing ports... 1575 Done 1576 1577 testpmd> port detach 0 1578 Detaching a port... 1579 PMD: Closing pcap ethdev on numa socket 0 1580 Port 'net_pcap0' is detached. Now total ports is 0 1581 Done 1582 1583To remove a pci device completely from the system, first detach the port from testpmd. 1584Then the device should be moved under kernel management. 1585Finally the device can be removed using kernel pci hotplug functionality. 1586 1587For example, to move a pci device under kernel management: 1588 1589.. code-block:: console 1590 1591 sudo ./usertools/dpdk-devbind.py -b ixgbe 0000:0a:00.0 1592 1593 ./usertools/dpdk-devbind.py --status 1594 1595 Network devices using DPDK-compatible driver 1596 ============================================ 1597 <none> 1598 1599 Network devices using kernel driver 1600 =================================== 1601 0000:0a:00.0 '82599ES 10-Gigabit' if=eth2 drv=ixgbe unused=igb_uio 1602 1603To remove a port created by a virtual device, above steps are not needed. 1604 1605port start 1606~~~~~~~~~~ 1607 1608Start all ports or a specific port:: 1609 1610 testpmd> port start (port_id|all) 1611 1612port stop 1613~~~~~~~~~ 1614 1615Stop all ports or a specific port:: 1616 1617 testpmd> port stop (port_id|all) 1618 1619port close 1620~~~~~~~~~~ 1621 1622Close all ports or a specific port:: 1623 1624 testpmd> port close (port_id|all) 1625 1626port config - queue ring size 1627~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ 1628 1629Configure a rx/tx queue ring size:: 1630 1631 testpmd> port (port_id) (rxq|txq) (queue_id) ring_size (value) 1632 1633Only take effect after command that (re-)start the port or command that setup specific queue. 1634 1635port start/stop queue 1636~~~~~~~~~~~~~~~~~~~~~ 1637 1638Start/stop a rx/tx queue on a specific port:: 1639 1640 testpmd> port (port_id) (rxq|txq) (queue_id) (start|stop) 1641 1642port setup queue 1643~~~~~~~~~~~~~~~~~~~~~ 1644 1645Setup a rx/tx queue on a specific port:: 1646 1647 testpmd> port (port_id) (rxq|txq) (queue_id) setup 1648 1649Only take effect when port is started. 1650 1651port config - speed 1652~~~~~~~~~~~~~~~~~~~ 1653 1654Set the speed and duplex mode for all ports or a specific port:: 1655 1656 testpmd> port config (port_id|all) speed (10|100|1000|10000|25000|40000|50000|100000|auto) \ 1657 duplex (half|full|auto) 1658 1659port config - queues/descriptors 1660~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ 1661 1662Set number of queues/descriptors for rxq, txq, rxd and txd:: 1663 1664 testpmd> port config all (rxq|txq|rxd|txd) (value) 1665 1666This is equivalent to the ``--rxq``, ``--txq``, ``--rxd`` and ``--txd`` command-line options. 1667 1668port config - max-pkt-len 1669~~~~~~~~~~~~~~~~~~~~~~~~~ 1670 1671Set the maximum packet length:: 1672 1673 testpmd> port config all max-pkt-len (value) 1674 1675This is equivalent to the ``--max-pkt-len`` command-line option. 1676 1677port config - CRC Strip 1678~~~~~~~~~~~~~~~~~~~~~~~ 1679 1680Set hardware CRC stripping on or off for all ports:: 1681 1682 testpmd> port config all crc-strip (on|off) 1683 1684CRC stripping is on by default. 1685 1686The ``off`` option is equivalent to the ``--disable-crc-strip`` command-line option. 1687 1688port config - scatter 1689~~~~~~~~~~~~~~~~~~~~~~~ 1690 1691Set RX scatter mode on or off for all ports:: 1692 1693 testpmd> port config all scatter (on|off) 1694 1695RX scatter mode is off by default. 1696 1697The ``on`` option is equivalent to the ``--enable-scatter`` command-line option. 1698 1699port config - RX Checksum 1700~~~~~~~~~~~~~~~~~~~~~~~~~ 1701 1702Set hardware RX checksum offload to on or off for all ports:: 1703 1704 testpmd> port config all rx-cksum (on|off) 1705 1706Checksum offload is off by default. 1707 1708The ``on`` option is equivalent to the ``--enable-rx-cksum`` command-line option. 1709 1710port config - VLAN 1711~~~~~~~~~~~~~~~~~~ 1712 1713Set hardware VLAN on or off for all ports:: 1714 1715 testpmd> port config all hw-vlan (on|off) 1716 1717Hardware VLAN is off by default. 1718 1719The ``on`` option is equivalent to the ``--enable-hw-vlan`` command-line option. 1720 1721port config - VLAN filter 1722~~~~~~~~~~~~~~~~~~~~~~~~~ 1723 1724Set hardware VLAN filter on or off for all ports:: 1725 1726 testpmd> port config all hw-vlan-filter (on|off) 1727 1728Hardware VLAN filter is off by default. 1729 1730The ``on`` option is equivalent to the ``--enable-hw-vlan-filter`` command-line option. 1731 1732port config - VLAN strip 1733~~~~~~~~~~~~~~~~~~~~~~~~ 1734 1735Set hardware VLAN strip on or off for all ports:: 1736 1737 testpmd> port config all hw-vlan-strip (on|off) 1738 1739Hardware VLAN strip is off by default. 1740 1741The ``on`` option is equivalent to the ``--enable-hw-vlan-strip`` command-line option. 1742 1743port config - VLAN extend 1744~~~~~~~~~~~~~~~~~~~~~~~~~ 1745 1746Set hardware VLAN extend on or off for all ports:: 1747 1748 testpmd> port config all hw-vlan-extend (on|off) 1749 1750Hardware VLAN extend is off by default. 1751 1752The ``on`` option is equivalent to the ``--enable-hw-vlan-extend`` command-line option. 1753 1754port config - Drop Packets 1755~~~~~~~~~~~~~~~~~~~~~~~~~~ 1756 1757Set packet drop for packets with no descriptors on or off for all ports:: 1758 1759 testpmd> port config all drop-en (on|off) 1760 1761Packet dropping for packets with no descriptors is off by default. 1762 1763The ``on`` option is equivalent to the ``--enable-drop-en`` command-line option. 1764 1765port config - RSS 1766~~~~~~~~~~~~~~~~~ 1767 1768Set the RSS (Receive Side Scaling) mode on or off:: 1769 1770 testpmd> port config all rss (all|default|ip|tcp|udp|sctp|ether|port|vxlan|geneve|nvgre|none) 1771 1772RSS is on by default. 1773 1774The ``all`` option is equivalent to ip|tcp|udp|sctp|ether. 1775The ``default`` option enables all supported RSS types reported by device info. 1776The ``none`` option is equivalent to the ``--disable-rss`` command-line option. 1777 1778port config - RSS Reta 1779~~~~~~~~~~~~~~~~~~~~~~ 1780 1781Set the RSS (Receive Side Scaling) redirection table:: 1782 1783 testpmd> port config all rss reta (hash,queue)[,(hash,queue)] 1784 1785port config - DCB 1786~~~~~~~~~~~~~~~~~ 1787 1788Set the DCB mode for an individual port:: 1789 1790 testpmd> port config (port_id) dcb vt (on|off) (traffic_class) pfc (on|off) 1791 1792The traffic class should be 4 or 8. 1793 1794port config - Burst 1795~~~~~~~~~~~~~~~~~~~ 1796 1797Set the number of packets per burst:: 1798 1799 testpmd> port config all burst (value) 1800 1801This is equivalent to the ``--burst`` command-line option. 1802 1803port config - Threshold 1804~~~~~~~~~~~~~~~~~~~~~~~ 1805 1806Set thresholds for TX/RX queues:: 1807 1808 testpmd> port config all (threshold) (value) 1809 1810Where the threshold type can be: 1811 1812* ``txpt:`` Set the prefetch threshold register of the TX rings, 0 <= value <= 255. 1813 1814* ``txht:`` Set the host threshold register of the TX rings, 0 <= value <= 255. 1815 1816* ``txwt:`` Set the write-back threshold register of the TX rings, 0 <= value <= 255. 1817 1818* ``rxpt:`` Set the prefetch threshold register of the RX rings, 0 <= value <= 255. 1819 1820* ``rxht:`` Set the host threshold register of the RX rings, 0 <= value <= 255. 1821 1822* ``rxwt:`` Set the write-back threshold register of the RX rings, 0 <= value <= 255. 1823 1824* ``txfreet:`` Set the transmit free threshold of the TX rings, 0 <= value <= txd. 1825 1826* ``rxfreet:`` Set the transmit free threshold of the RX rings, 0 <= value <= rxd. 1827 1828* ``txrst:`` Set the transmit RS bit threshold of TX rings, 0 <= value <= txd. 1829 1830These threshold options are also available from the command-line. 1831 1832port config - E-tag 1833~~~~~~~~~~~~~~~~~~~ 1834 1835Set the value of ether-type for E-tag:: 1836 1837 testpmd> port config (port_id|all) l2-tunnel E-tag ether-type (value) 1838 1839Enable/disable the E-tag support:: 1840 1841 testpmd> port config (port_id|all) l2-tunnel E-tag (enable|disable) 1842 1843port config pctype mapping 1844~~~~~~~~~~~~~~~~~~~~~~~~~~ 1845 1846Reset pctype mapping table:: 1847 1848 testpmd> port config (port_id) pctype mapping reset 1849 1850Update hardware defined pctype to software defined flow type mapping table:: 1851 1852 testpmd> port config (port_id) pctype mapping update (pctype_id_0[,pctype_id_1]*) (flow_type_id) 1853 1854where: 1855 1856* ``pctype_id_x``: hardware pctype id as index of bit in bitmask value of the pctype mapping table. 1857 1858* ``flow_type_id``: software flow type id as the index of the pctype mapping table. 1859 1860port config input set 1861~~~~~~~~~~~~~~~~~~~~~ 1862 1863Config RSS/FDIR/FDIR flexible payload input set for some pctype:: 1864 testpmd> port config (port_id) pctype (pctype_id) \ 1865 (hash_inset|fdir_inset|fdir_flx_inset) \ 1866 (get|set|clear) field (field_idx) 1867 1868Clear RSS/FDIR/FDIR flexible payload input set for some pctype:: 1869 testpmd> port config (port_id) pctype (pctype_id) \ 1870 (hash_inset|fdir_inset|fdir_flx_inset) clear all 1871 1872where: 1873 1874* ``pctype_id``: hardware packet classification types. 1875* ``field_idx``: hardware field index. 1876 1877Link Bonding Functions 1878---------------------- 1879 1880The Link Bonding functions make it possible to dynamically create and 1881manage link bonding devices from within testpmd interactive prompt. 1882 1883create bonded device 1884~~~~~~~~~~~~~~~~~~~~ 1885 1886Create a new bonding device:: 1887 1888 testpmd> create bonded device (mode) (socket) 1889 1890For example, to create a bonded device in mode 1 on socket 0:: 1891 1892 testpmd> create bonded 1 0 1893 created new bonded device (port X) 1894 1895add bonding slave 1896~~~~~~~~~~~~~~~~~ 1897 1898Adds Ethernet device to a Link Bonding device:: 1899 1900 testpmd> add bonding slave (slave id) (port id) 1901 1902For example, to add Ethernet device (port 6) to a Link Bonding device (port 10):: 1903 1904 testpmd> add bonding slave 6 10 1905 1906 1907remove bonding slave 1908~~~~~~~~~~~~~~~~~~~~ 1909 1910Removes an Ethernet slave device from a Link Bonding device:: 1911 1912 testpmd> remove bonding slave (slave id) (port id) 1913 1914For example, to remove Ethernet slave device (port 6) to a Link Bonding device (port 10):: 1915 1916 testpmd> remove bonding slave 6 10 1917 1918set bonding mode 1919~~~~~~~~~~~~~~~~ 1920 1921Set the Link Bonding mode of a Link Bonding device:: 1922 1923 testpmd> set bonding mode (value) (port id) 1924 1925For example, to set the bonding mode of a Link Bonding device (port 10) to broadcast (mode 3):: 1926 1927 testpmd> set bonding mode 3 10 1928 1929set bonding primary 1930~~~~~~~~~~~~~~~~~~~ 1931 1932Set an Ethernet slave device as the primary device on a Link Bonding device:: 1933 1934 testpmd> set bonding primary (slave id) (port id) 1935 1936For example, to set the Ethernet slave device (port 6) as the primary port of a Link Bonding device (port 10):: 1937 1938 testpmd> set bonding primary 6 10 1939 1940set bonding mac 1941~~~~~~~~~~~~~~~ 1942 1943Set the MAC address of a Link Bonding device:: 1944 1945 testpmd> set bonding mac (port id) (mac) 1946 1947For example, to set the MAC address of a Link Bonding device (port 10) to 00:00:00:00:00:01:: 1948 1949 testpmd> set bonding mac 10 00:00:00:00:00:01 1950 1951set bonding xmit_balance_policy 1952~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ 1953 1954Set the transmission policy for a Link Bonding device when it is in Balance XOR mode:: 1955 1956 testpmd> set bonding xmit_balance_policy (port_id) (l2|l23|l34) 1957 1958For example, set a Link Bonding device (port 10) to use a balance policy of layer 3+4 (IP addresses & UDP ports):: 1959 1960 testpmd> set bonding xmit_balance_policy 10 l34 1961 1962 1963set bonding mon_period 1964~~~~~~~~~~~~~~~~~~~~~~ 1965 1966Set the link status monitoring polling period in milliseconds for a bonding device. 1967 1968This adds support for PMD slave devices which do not support link status interrupts. 1969When the mon_period is set to a value greater than 0 then all PMD's which do not support 1970link status ISR will be queried every polling interval to check if their link status has changed:: 1971 1972 testpmd> set bonding mon_period (port_id) (value) 1973 1974For example, to set the link status monitoring polling period of bonded device (port 5) to 150ms:: 1975 1976 testpmd> set bonding mon_period 5 150 1977 1978 1979set bonding lacp dedicated_queue 1980~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ 1981 1982Enable dedicated tx/rx queues on bonding devices slaves to handle LACP control plane traffic 1983when in mode 4 (link-aggregration-802.3ad):: 1984 1985 testpmd> set bonding lacp dedicated_queues (port_id) (enable|disable) 1986 1987 1988set bonding agg_mode 1989~~~~~~~~~~~~~~~~~~~~ 1990 1991Enable one of the specific aggregators mode when in mode 4 (link-aggregration-802.3ad):: 1992 1993 testpmd> set bonding agg_mode (port_id) (bandwidth|count|stable) 1994 1995 1996show bonding config 1997~~~~~~~~~~~~~~~~~~~ 1998 1999Show the current configuration of a Link Bonding device:: 2000 2001 testpmd> show bonding config (port id) 2002 2003For example, 2004to show the configuration a Link Bonding device (port 9) with 3 slave devices (1, 3, 4) 2005in balance mode with a transmission policy of layer 2+3:: 2006 2007 testpmd> show bonding config 9 2008 Bonding mode: 2 2009 Balance Xmit Policy: BALANCE_XMIT_POLICY_LAYER23 2010 Slaves (3): [1 3 4] 2011 Active Slaves (3): [1 3 4] 2012 Primary: [3] 2013 2014 2015Register Functions 2016------------------ 2017 2018The Register Functions can be used to read from and write to registers on the network card referenced by a port number. 2019This is mainly useful for debugging purposes. 2020Reference should be made to the appropriate datasheet for the network card for details on the register addresses 2021and fields that can be accessed. 2022 2023read reg 2024~~~~~~~~ 2025 2026Display the value of a port register:: 2027 2028 testpmd> read reg (port_id) (address) 2029 2030For example, to examine the Flow Director control register (FDIRCTL, 0x0000EE000) on an Intel 82599 10 GbE Controller:: 2031 2032 testpmd> read reg 0 0xEE00 2033 port 0 PCI register at offset 0xEE00: 0x4A060029 (1241907241) 2034 2035read regfield 2036~~~~~~~~~~~~~ 2037 2038Display a port register bit field:: 2039 2040 testpmd> read regfield (port_id) (address) (bit_x) (bit_y) 2041 2042For example, reading the lowest two bits from the register in the example above:: 2043 2044 testpmd> read regfield 0 0xEE00 0 1 2045 port 0 PCI register at offset 0xEE00: bits[0, 1]=0x1 (1) 2046 2047read regbit 2048~~~~~~~~~~~ 2049 2050Display a single port register bit:: 2051 2052 testpmd> read regbit (port_id) (address) (bit_x) 2053 2054For example, reading the lowest bit from the register in the example above:: 2055 2056 testpmd> read regbit 0 0xEE00 0 2057 port 0 PCI register at offset 0xEE00: bit 0=1 2058 2059write reg 2060~~~~~~~~~ 2061 2062Set the value of a port register:: 2063 2064 testpmd> write reg (port_id) (address) (value) 2065 2066For example, to clear a register:: 2067 2068 testpmd> write reg 0 0xEE00 0x0 2069 port 0 PCI register at offset 0xEE00: 0x00000000 (0) 2070 2071write regfield 2072~~~~~~~~~~~~~~ 2073 2074Set bit field of a port register:: 2075 2076 testpmd> write regfield (port_id) (address) (bit_x) (bit_y) (value) 2077 2078For example, writing to the register cleared in the example above:: 2079 2080 testpmd> write regfield 0 0xEE00 0 1 2 2081 port 0 PCI register at offset 0xEE00: 0x00000002 (2) 2082 2083write regbit 2084~~~~~~~~~~~~ 2085 2086Set single bit value of a port register:: 2087 2088 testpmd> write regbit (port_id) (address) (bit_x) (value) 2089 2090For example, to set the high bit in the register from the example above:: 2091 2092 testpmd> write regbit 0 0xEE00 31 1 2093 port 0 PCI register at offset 0xEE00: 0x8000000A (2147483658) 2094 2095Traffic Metering and Policing 2096----------------------------- 2097 2098The following section shows functions for configuring traffic metering and 2099policing on the ethernet device through the use of generic ethdev API. 2100 2101show port traffic management capability 2102~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ 2103 2104Show traffic metering and policing capability of the port:: 2105 2106 testpmd> show port meter cap (port_id) 2107 2108add port meter profile (srTCM rfc2967) 2109~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ 2110 2111Add meter profile (srTCM rfc2697) to the ethernet device:: 2112 2113 testpmd> add port meter profile srtcm_rfc2697 (port_id) (profile_id) \ 2114 (cir) (cbs) (ebs) 2115 2116where: 2117 2118* ``profile_id``: ID for the meter profile. 2119* ``cir``: Committed Information Rate (CIR) (bytes/second). 2120* ``cbs``: Committed Burst Size (CBS) (bytes). 2121* ``ebs``: Excess Burst Size (EBS) (bytes). 2122 2123add port meter profile (trTCM rfc2968) 2124~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ 2125 2126Add meter profile (srTCM rfc2698) to the ethernet device:: 2127 2128 testpmd> add port meter profile trtcm_rfc2698 (port_id) (profile_id) \ 2129 (cir) (pir) (cbs) (pbs) 2130 2131where: 2132 2133* ``profile_id``: ID for the meter profile. 2134* ``cir``: Committed information rate (bytes/second). 2135* ``pir``: Peak information rate (bytes/second). 2136* ``cbs``: Committed burst size (bytes). 2137* ``pbs``: Peak burst size (bytes). 2138 2139add port meter profile (trTCM rfc4115) 2140~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ 2141 2142Add meter profile (trTCM rfc4115) to the ethernet device:: 2143 2144 testpmd> add port meter profile trtcm_rfc4115 (port_id) (profile_id) \ 2145 (cir) (eir) (cbs) (ebs) 2146 2147where: 2148 2149* ``profile_id``: ID for the meter profile. 2150* ``cir``: Committed information rate (bytes/second). 2151* ``eir``: Excess information rate (bytes/second). 2152* ``cbs``: Committed burst size (bytes). 2153* ``ebs``: Excess burst size (bytes). 2154 2155delete port meter profile 2156~~~~~~~~~~~~~~~~~~~~~~~~~ 2157 2158Delete meter profile from the ethernet device:: 2159 2160 testpmd> del port meter profile (port_id) (profile_id) 2161 2162create port meter 2163~~~~~~~~~~~~~~~~~ 2164 2165Create new meter object for the ethernet device:: 2166 2167 testpmd> create port meter (port_id) (mtr_id) (profile_id) \ 2168 (meter_enable) (g_action) (y_action) (r_action) (stats_mask) (shared) \ 2169 (use_pre_meter_color) [(dscp_tbl_entry0) (dscp_tbl_entry1)...\ 2170 (dscp_tbl_entry63)] 2171 2172where: 2173 2174* ``mtr_id``: meter object ID. 2175* ``profile_id``: ID for the meter profile. 2176* ``meter_enable``: When this parameter has a non-zero value, the meter object 2177 gets enabled at the time of creation, otherwise remains disabled. 2178* ``g_action``: Policer action for the packet with green color. 2179* ``y_action``: Policer action for the packet with yellow color. 2180* ``r_action``: Policer action for the packet with red color. 2181* ``stats_mask``: Mask of statistics counter types to be enabled for the 2182 meter object. 2183* ``shared``: When this parameter has a non-zero value, the meter object is 2184 shared by multiple flows. Otherwise, meter object is used by single flow. 2185* ``use_pre_meter_color``: When this parameter has a non-zero value, the 2186 input color for the current meter object is determined by the latest meter 2187 object in the same flow. Otherwise, the current meter object uses the 2188 *dscp_table* to determine the input color. 2189* ``dscp_tbl_entryx``: DSCP table entry x providing meter providing input 2190 color, 0 <= x <= 63. 2191 2192enable port meter 2193~~~~~~~~~~~~~~~~~ 2194 2195Enable meter for the ethernet device:: 2196 2197 testpmd> enable port meter (port_id) (mtr_id) 2198 2199disable port meter 2200~~~~~~~~~~~~~~~~~~ 2201 2202Disable meter for the ethernet device:: 2203 2204 testpmd> disable port meter (port_id) (mtr_id) 2205 2206delete port meter 2207~~~~~~~~~~~~~~~~~ 2208 2209Delete meter for the ethernet device:: 2210 2211 testpmd> del port meter (port_id) (mtr_id) 2212 2213Set port meter profile 2214~~~~~~~~~~~~~~~~~~~~~~ 2215 2216Set meter profile for the ethernet device:: 2217 2218 testpmd> set port meter profile (port_id) (mtr_id) (profile_id) 2219 2220set port meter dscp table 2221~~~~~~~~~~~~~~~~~~~~~~~~~ 2222 2223Set meter dscp table for the ethernet device:: 2224 2225 testpmd> set port meter dscp table (port_id) (mtr_id) [(dscp_tbl_entry0) \ 2226 (dscp_tbl_entry1)...(dscp_tbl_entry63)] 2227 2228set port meter policer action 2229~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ 2230 2231Set meter policer action for the ethernet device:: 2232 2233 testpmd> set port meter policer action (port_id) (mtr_id) (action_mask) \ 2234 (action0) [(action1) (action1)] 2235 2236where: 2237 2238* ``action_mask``: Bit mask indicating which policer actions need to be 2239 updated. One or more policer actions can be updated in a single function 2240 invocation. To update the policer action associated with color C, bit 2241 (1 << C) needs to be set in *action_mask* and element at position C 2242 in the *actions* array needs to be valid. 2243* ``actionx``: Policer action for the color x, 2244 RTE_MTR_GREEN <= x < RTE_MTR_COLORS 2245 2246set port meter stats mask 2247~~~~~~~~~~~~~~~~~~~~~~~~~ 2248 2249Set meter stats mask for the ethernet device:: 2250 2251 testpmd> set port meter stats mask (port_id) (mtr_id) (stats_mask) 2252 2253where: 2254 2255* ``stats_mask``: Bit mask indicating statistics counter types to be enabled. 2256 2257show port meter stats 2258~~~~~~~~~~~~~~~~~~~~~ 2259 2260Show meter stats of the ethernet device:: 2261 2262 testpmd> show port meter stats (port_id) (mtr_id) (clear) 2263 2264where: 2265 2266* ``clear``: Flag that indicates whether the statistics counters should 2267 be cleared (i.e. set to zero) immediately after they have been read or not. 2268 2269Traffic Management 2270------------------ 2271 2272The following section shows functions for configuring traffic management on 2273on the ethernet device through the use of generic TM API. 2274 2275show port traffic management capability 2276~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ 2277 2278Show traffic management capability of the port:: 2279 2280 testpmd> show port tm cap (port_id) 2281 2282show port traffic management capability (hierarchy level) 2283~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ 2284 2285Show traffic management hierarchy level capability of the port:: 2286 2287 testpmd> show port tm level cap (port_id) (level_id) 2288 2289show port traffic management capability (hierarchy node level) 2290~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ 2291 2292Show the traffic management hierarchy node capability of the port:: 2293 2294 testpmd> show port tm node cap (port_id) (node_id) 2295 2296show port traffic management hierarchy node type 2297~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ 2298 2299Show the port traffic management hierarchy node type:: 2300 2301 testpmd> show port tm node type (port_id) (node_id) 2302 2303show port traffic management hierarchy node stats 2304~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ 2305 2306Show the port traffic management hierarchy node statistics:: 2307 2308 testpmd> show port tm node stats (port_id) (node_id) (clear) 2309 2310where: 2311 2312* ``clear``: When this parameter has a non-zero value, the statistics counters 2313 are cleared (i.e. set to zero) immediately after they have been read, 2314 otherwise the statistics counters are left untouched. 2315 2316Add port traffic management private shaper profile 2317~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ 2318 2319Add the port traffic management private shaper profile:: 2320 2321 testpmd> add port tm node shaper profile (port_id) (shaper_profile_id) \ 2322 (tb_rate) (tb_size) (packet_length_adjust) 2323 2324where: 2325 2326* ``shaper_profile id``: Shaper profile ID for the new profile. 2327* ``tb_rate``: Token bucket rate (bytes per second). 2328* ``tb_size``: Token bucket size (bytes). 2329* ``packet_length_adjust``: The value (bytes) to be added to the length of 2330 each packet for the purpose of shaping. This parameter value can be used to 2331 correct the packet length with the framing overhead bytes that are consumed 2332 on the wire. 2333 2334Delete port traffic management private shaper profile 2335~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ 2336 2337Delete the port traffic management private shaper:: 2338 2339 testpmd> del port tm node shaper profile (port_id) (shaper_profile_id) 2340 2341where: 2342 2343* ``shaper_profile id``: Shaper profile ID that needs to be deleted. 2344 2345Add port traffic management shared shaper 2346~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ 2347 2348Create the port traffic management shared shaper:: 2349 2350 testpmd> add port tm node shared shaper (port_id) (shared_shaper_id) \ 2351 (shaper_profile_id) 2352 2353where: 2354 2355* ``shared_shaper_id``: Shared shaper ID to be created. 2356* ``shaper_profile id``: Shaper profile ID for shared shaper. 2357 2358Set port traffic management shared shaper 2359~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ 2360 2361Update the port traffic management shared shaper:: 2362 2363 testpmd> set port tm node shared shaper (port_id) (shared_shaper_id) \ 2364 (shaper_profile_id) 2365 2366where: 2367 2368* ``shared_shaper_id``: Shared shaper ID to be update. 2369* ``shaper_profile id``: Shaper profile ID for shared shaper. 2370 2371Delete port traffic management shared shaper 2372~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ 2373 2374Delete the port traffic management shared shaper:: 2375 2376 testpmd> del port tm node shared shaper (port_id) (shared_shaper_id) 2377 2378where: 2379 2380* ``shared_shaper_id``: Shared shaper ID to be deleted. 2381 2382Set port traffic management hiearchy node private shaper 2383~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ 2384 2385set the port traffic management hierarchy node private shaper:: 2386 2387 testpmd> set port tm node shaper profile (port_id) (node_id) \ 2388 (shaper_profile_id) 2389 2390where: 2391 2392* ``shaper_profile id``: Private shaper profile ID to be enabled on the 2393 hierarchy node. 2394 2395Add port traffic management WRED profile 2396~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ 2397 2398Create a new WRED profile:: 2399 2400 testpmd> add port tm node wred profile (port_id) (wred_profile_id) \ 2401 (color_g) (min_th_g) (max_th_g) (maxp_inv_g) (wq_log2_g) \ 2402 (color_y) (min_th_y) (max_th_y) (maxp_inv_y) (wq_log2_y) \ 2403 (color_r) (min_th_r) (max_th_r) (maxp_inv_r) (wq_log2_r) 2404 2405where: 2406 2407* ``wred_profile id``: Identifier for the newly create WRED profile 2408* ``color_g``: Packet color (green) 2409* ``min_th_g``: Minimum queue threshold for packet with green color 2410* ``max_th_g``: Minimum queue threshold for packet with green color 2411* ``maxp_inv_g``: Inverse of packet marking probability maximum value (maxp) 2412* ``wq_log2_g``: Negated log2 of queue weight (wq) 2413* ``color_y``: Packet color (yellow) 2414* ``min_th_y``: Minimum queue threshold for packet with yellow color 2415* ``max_th_y``: Minimum queue threshold for packet with yellow color 2416* ``maxp_inv_y``: Inverse of packet marking probability maximum value (maxp) 2417* ``wq_log2_y``: Negated log2 of queue weight (wq) 2418* ``color_r``: Packet color (red) 2419* ``min_th_r``: Minimum queue threshold for packet with yellow color 2420* ``max_th_r``: Minimum queue threshold for packet with yellow color 2421* ``maxp_inv_r``: Inverse of packet marking probability maximum value (maxp) 2422* ``wq_log2_r``: Negated log2 of queue weight (wq) 2423 2424Delete port traffic management WRED profile 2425~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ 2426 2427Delete the WRED profile:: 2428 2429 testpmd> del port tm node wred profile (port_id) (wred_profile_id) 2430 2431Add port traffic management hierarchy nonleaf node 2432~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ 2433 2434Add nonleaf node to port traffic management hiearchy:: 2435 2436 testpmd> add port tm nonleaf node (port_id) (node_id) (parent_node_id) \ 2437 (priority) (weight) (level_id) (shaper_profile_id) \ 2438 (n_sp_priorities) (stats_mask) (n_shared_shapers) \ 2439 [(shared_shaper_0) (shared_shaper_1) ...] \ 2440 2441where: 2442 2443* ``parent_node_id``: Node ID of the parent. 2444* ``priority``: Node priority (highest node priority is zero). This is used by 2445 the SP algorithm running on the parent node for scheduling this node. 2446* ``weight``: Node weight (lowest weight is one). The node weight is relative 2447 to the weight sum of all siblings that have the same priority. It is used by 2448 the WFQ algorithm running on the parent node for scheduling this node. 2449* ``level_id``: Hiearchy level of the node. 2450* ``shaper_profile_id``: Shaper profile ID of the private shaper to be used by 2451 the node. 2452* ``n_sp_priorities``: Number of strict priorities. 2453* ``stats_mask``: Mask of statistics counter types to be enabled for this node. 2454* ``n_shared_shapers``: Number of shared shapers. 2455* ``shared_shaper_id``: Shared shaper id. 2456 2457Add port traffic management hierarchy leaf node 2458~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ 2459 2460Add leaf node to port traffic management hiearchy:: 2461 2462 testpmd> add port tm leaf node (port_id) (node_id) (parent_node_id) \ 2463 (priority) (weight) (level_id) (shaper_profile_id) \ 2464 (cman_mode) (wred_profile_id) (stats_mask) (n_shared_shapers) \ 2465 [(shared_shaper_id) (shared_shaper_id) ...] \ 2466 2467where: 2468 2469* ``parent_node_id``: Node ID of the parent. 2470* ``priority``: Node priority (highest node priority is zero). This is used by 2471 the SP algorithm running on the parent node for scheduling this node. 2472* ``weight``: Node weight (lowest weight is one). The node weight is relative 2473 to the weight sum of all siblings that have the same priority. It is used by 2474 the WFQ algorithm running on the parent node for scheduling this node. 2475* ``level_id``: Hiearchy level of the node. 2476* ``shaper_profile_id``: Shaper profile ID of the private shaper to be used by 2477 the node. 2478* ``cman_mode``: Congestion management mode to be enabled for this node. 2479* ``wred_profile_id``: WRED profile id to be enabled for this node. 2480* ``stats_mask``: Mask of statistics counter types to be enabled for this node. 2481* ``n_shared_shapers``: Number of shared shapers. 2482* ``shared_shaper_id``: Shared shaper id. 2483 2484Delete port traffic management hierarchy node 2485~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ 2486 2487Delete node from port traffic management hiearchy:: 2488 2489 testpmd> del port tm node (port_id) (node_id) 2490 2491Update port traffic management hierarchy parent node 2492~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ 2493 2494Update port traffic management hierarchy parent node:: 2495 2496 testpmd> set port tm node parent (port_id) (node_id) (parent_node_id) \ 2497 (priority) (weight) 2498 2499This function can only be called after the hierarchy commit invocation. Its 2500success depends on the port support for this operation, as advertised through 2501the port capability set. This function is valid for all nodes of the traffic 2502management hierarchy except root node. 2503 2504Commit port traffic management hierarchy 2505~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ 2506 2507Commit the traffic management hierarchy on the port:: 2508 2509 testpmd> port tm hierarchy commit (port_id) (clean_on_fail) 2510 2511where: 2512 2513* ``clean_on_fail``: When set to non-zero, hierarchy is cleared on function 2514 call failure. On the other hand, hierarchy is preserved when this parameter 2515 is equal to zero. 2516 2517Set port traffic management default hierarchy (tm forwarding mode) 2518~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ 2519 2520set the traffic management default hierarchy on the port:: 2521 2522 testpmd> set port tm hierarchy default (port_id) 2523 2524Filter Functions 2525---------------- 2526 2527This section details the available filter functions that are available. 2528 2529Note these functions interface the deprecated legacy filtering framework, 2530superseded by *rte_flow*. See `Flow rules management`_. 2531 2532ethertype_filter 2533~~~~~~~~~~~~~~~~~~~~ 2534 2535Add or delete a L2 Ethertype filter, which identify packets by their L2 Ethertype mainly assign them to a receive queue:: 2536 2537 ethertype_filter (port_id) (add|del) (mac_addr|mac_ignr) (mac_address) \ 2538 ethertype (ether_type) (drop|fwd) queue (queue_id) 2539 2540The available information parameters are: 2541 2542* ``port_id``: The port which the Ethertype filter assigned on. 2543 2544* ``mac_addr``: Compare destination mac address. 2545 2546* ``mac_ignr``: Ignore destination mac address match. 2547 2548* ``mac_address``: Destination mac address to match. 2549 2550* ``ether_type``: The EtherType value want to match, 2551 for example 0x0806 for ARP packet. 0x0800 (IPv4) and 0x86DD (IPv6) are invalid. 2552 2553* ``queue_id``: The receive queue associated with this EtherType filter. 2554 It is meaningless when deleting or dropping. 2555 2556Example, to add/remove an ethertype filter rule:: 2557 2558 testpmd> ethertype_filter 0 add mac_ignr 00:11:22:33:44:55 \ 2559 ethertype 0x0806 fwd queue 3 2560 2561 testpmd> ethertype_filter 0 del mac_ignr 00:11:22:33:44:55 \ 2562 ethertype 0x0806 fwd queue 3 2563 25642tuple_filter 2565~~~~~~~~~~~~~~~~~ 2566 2567Add or delete a 2-tuple filter, 2568which identifies packets by specific protocol and destination TCP/UDP port 2569and forwards packets into one of the receive queues:: 2570 2571 2tuple_filter (port_id) (add|del) dst_port (dst_port_value) \ 2572 protocol (protocol_value) mask (mask_value) \ 2573 tcp_flags (tcp_flags_value) priority (prio_value) \ 2574 queue (queue_id) 2575 2576The available information parameters are: 2577 2578* ``port_id``: The port which the 2-tuple filter assigned on. 2579 2580* ``dst_port_value``: Destination port in L4. 2581 2582* ``protocol_value``: IP L4 protocol. 2583 2584* ``mask_value``: Participates in the match or not by bit for field above, 1b means participate. 2585 2586* ``tcp_flags_value``: TCP control bits. The non-zero value is invalid, when the pro_value is not set to 0x06 (TCP). 2587 2588* ``prio_value``: Priority of this filter. 2589 2590* ``queue_id``: The receive queue associated with this 2-tuple filter. 2591 2592Example, to add/remove an 2tuple filter rule:: 2593 2594 testpmd> 2tuple_filter 0 add dst_port 32 protocol 0x06 mask 0x03 \ 2595 tcp_flags 0x02 priority 3 queue 3 2596 2597 testpmd> 2tuple_filter 0 del dst_port 32 protocol 0x06 mask 0x03 \ 2598 tcp_flags 0x02 priority 3 queue 3 2599 26005tuple_filter 2601~~~~~~~~~~~~~~~~~ 2602 2603Add or delete a 5-tuple filter, 2604which consists of a 5-tuple (protocol, source and destination IP addresses, source and destination TCP/UDP/SCTP port) 2605and routes packets into one of the receive queues:: 2606 2607 5tuple_filter (port_id) (add|del) dst_ip (dst_address) src_ip \ 2608 (src_address) dst_port (dst_port_value) \ 2609 src_port (src_port_value) protocol (protocol_value) \ 2610 mask (mask_value) tcp_flags (tcp_flags_value) \ 2611 priority (prio_value) queue (queue_id) 2612 2613The available information parameters are: 2614 2615* ``port_id``: The port which the 5-tuple filter assigned on. 2616 2617* ``dst_address``: Destination IP address. 2618 2619* ``src_address``: Source IP address. 2620 2621* ``dst_port_value``: TCP/UDP destination port. 2622 2623* ``src_port_value``: TCP/UDP source port. 2624 2625* ``protocol_value``: L4 protocol. 2626 2627* ``mask_value``: Participates in the match or not by bit for field above, 1b means participate 2628 2629* ``tcp_flags_value``: TCP control bits. The non-zero value is invalid, when the protocol_value is not set to 0x06 (TCP). 2630 2631* ``prio_value``: The priority of this filter. 2632 2633* ``queue_id``: The receive queue associated with this 5-tuple filter. 2634 2635Example, to add/remove an 5tuple filter rule:: 2636 2637 testpmd> 5tuple_filter 0 add dst_ip 2.2.2.5 src_ip 2.2.2.4 \ 2638 dst_port 64 src_port 32 protocol 0x06 mask 0x1F \ 2639 flags 0x0 priority 3 queue 3 2640 2641 testpmd> 5tuple_filter 0 del dst_ip 2.2.2.5 src_ip 2.2.2.4 \ 2642 dst_port 64 src_port 32 protocol 0x06 mask 0x1F \ 2643 flags 0x0 priority 3 queue 3 2644 2645syn_filter 2646~~~~~~~~~~ 2647 2648Using the SYN filter, TCP packets whose *SYN* flag is set can be forwarded to a separate queue:: 2649 2650 syn_filter (port_id) (add|del) priority (high|low) queue (queue_id) 2651 2652The available information parameters are: 2653 2654* ``port_id``: The port which the SYN filter assigned on. 2655 2656* ``high``: This SYN filter has higher priority than other filters. 2657 2658* ``low``: This SYN filter has lower priority than other filters. 2659 2660* ``queue_id``: The receive queue associated with this SYN filter 2661 2662Example:: 2663 2664 testpmd> syn_filter 0 add priority high queue 3 2665 2666flex_filter 2667~~~~~~~~~~~ 2668 2669With flex filter, packets can be recognized by any arbitrary pattern within the first 128 bytes of the packet 2670and routed into one of the receive queues:: 2671 2672 flex_filter (port_id) (add|del) len (len_value) bytes (bytes_value) \ 2673 mask (mask_value) priority (prio_value) queue (queue_id) 2674 2675The available information parameters are: 2676 2677* ``port_id``: The port which the Flex filter is assigned on. 2678 2679* ``len_value``: Filter length in bytes, no greater than 128. 2680 2681* ``bytes_value``: A string in hexadecimal, means the value the flex filter needs to match. 2682 2683* ``mask_value``: A string in hexadecimal, bit 1 means corresponding byte participates in the match. 2684 2685* ``prio_value``: The priority of this filter. 2686 2687* ``queue_id``: The receive queue associated with this Flex filter. 2688 2689Example:: 2690 2691 testpmd> flex_filter 0 add len 16 bytes 0x00000000000000000000000008060000 \ 2692 mask 000C priority 3 queue 3 2693 2694 testpmd> flex_filter 0 del len 16 bytes 0x00000000000000000000000008060000 \ 2695 mask 000C priority 3 queue 3 2696 2697 2698.. _testpmd_flow_director: 2699 2700flow_director_filter 2701~~~~~~~~~~~~~~~~~~~~ 2702 2703The Flow Director works in receive mode to identify specific flows or sets of flows and route them to specific queues. 2704 2705Four types of filtering are supported which are referred to as Perfect Match, Signature, Perfect-mac-vlan and 2706Perfect-tunnel filters, the match mode is set by the ``--pkt-filter-mode`` command-line parameter: 2707 2708* Perfect match filters. 2709 The hardware checks a match between the masked fields of the received packets and the programmed filters. 2710 The masked fields are for IP flow. 2711 2712* Signature filters. 2713 The hardware checks a match between a hash-based signature of the masked fields of the received packet. 2714 2715* Perfect-mac-vlan match filters. 2716 The hardware checks a match between the masked fields of the received packets and the programmed filters. 2717 The masked fields are for MAC VLAN flow. 2718 2719* Perfect-tunnel match filters. 2720 The hardware checks a match between the masked fields of the received packets and the programmed filters. 2721 The masked fields are for tunnel flow. 2722 2723* Perfect-raw-flow-type match filters. 2724 The hardware checks a match between the masked fields of the received packets and pre-loaded raw (template) packet. 2725 The masked fields are specified by input sets. 2726 2727The Flow Director filters can match the different fields for different type of packet: flow type, specific input set 2728per flow type and the flexible payload. 2729 2730The Flow Director can also mask out parts of all of these fields so that filters 2731are only applied to certain fields or parts of the fields. 2732 2733Note that for raw flow type mode the source and destination fields in the 2734raw packet buffer need to be presented in a reversed order with respect 2735to the expected received packets. 2736For example: IP source and destination addresses or TCP/UDP/SCTP 2737source and destination ports 2738 2739Different NICs may have different capabilities, command show port fdir (port_id) can be used to acquire the information. 2740 2741# Commands to add flow director filters of different flow types:: 2742 2743 flow_director_filter (port_id) mode IP (add|del|update) \ 2744 flow (ipv4-other|ipv4-frag|ipv6-other|ipv6-frag) \ 2745 src (src_ip_address) dst (dst_ip_address) \ 2746 tos (tos_value) proto (proto_value) ttl (ttl_value) \ 2747 vlan (vlan_value) flexbytes (flexbytes_value) \ 2748 (drop|fwd) pf|vf(vf_id) queue (queue_id) \ 2749 fd_id (fd_id_value) 2750 2751 flow_director_filter (port_id) mode IP (add|del|update) \ 2752 flow (ipv4-tcp|ipv4-udp|ipv6-tcp|ipv6-udp) \ 2753 src (src_ip_address) (src_port) \ 2754 dst (dst_ip_address) (dst_port) \ 2755 tos (tos_value) ttl (ttl_value) \ 2756 vlan (vlan_value) flexbytes (flexbytes_value) \ 2757 (drop|fwd) queue pf|vf(vf_id) (queue_id) \ 2758 fd_id (fd_id_value) 2759 2760 flow_director_filter (port_id) mode IP (add|del|update) \ 2761 flow (ipv4-sctp|ipv6-sctp) \ 2762 src (src_ip_address) (src_port) \ 2763 dst (dst_ip_address) (dst_port) \ 2764 tos (tos_value) ttl (ttl_value) \ 2765 tag (verification_tag) vlan (vlan_value) \ 2766 flexbytes (flexbytes_value) (drop|fwd) \ 2767 pf|vf(vf_id) queue (queue_id) fd_id (fd_id_value) 2768 2769 flow_director_filter (port_id) mode IP (add|del|update) flow l2_payload \ 2770 ether (ethertype) flexbytes (flexbytes_value) \ 2771 (drop|fwd) pf|vf(vf_id) queue (queue_id) 2772 fd_id (fd_id_value) 2773 2774 flow_director_filter (port_id) mode MAC-VLAN (add|del|update) \ 2775 mac (mac_address) vlan (vlan_value) \ 2776 flexbytes (flexbytes_value) (drop|fwd) \ 2777 queue (queue_id) fd_id (fd_id_value) 2778 2779 flow_director_filter (port_id) mode Tunnel (add|del|update) \ 2780 mac (mac_address) vlan (vlan_value) \ 2781 tunnel (NVGRE|VxLAN) tunnel-id (tunnel_id_value) \ 2782 flexbytes (flexbytes_value) (drop|fwd) \ 2783 queue (queue_id) fd_id (fd_id_value) 2784 2785 flow_director_filter (port_id) mode raw (add|del|update) flow (flow_id) \ 2786 (drop|fwd) queue (queue_id) fd_id (fd_id_value) \ 2787 packet (packet file name) 2788 2789For example, to add an ipv4-udp flow type filter:: 2790 2791 testpmd> flow_director_filter 0 mode IP add flow ipv4-udp src 2.2.2.3 32 \ 2792 dst 2.2.2.5 33 tos 2 ttl 40 vlan 0x1 flexbytes (0x88,0x48) \ 2793 fwd pf queue 1 fd_id 1 2794 2795For example, add an ipv4-other flow type filter:: 2796 2797 testpmd> flow_director_filter 0 mode IP add flow ipv4-other src 2.2.2.3 \ 2798 dst 2.2.2.5 tos 2 proto 20 ttl 40 vlan 0x1 \ 2799 flexbytes (0x88,0x48) fwd pf queue 1 fd_id 1 2800 2801flush_flow_director 2802~~~~~~~~~~~~~~~~~~~ 2803 2804Flush all flow director filters on a device:: 2805 2806 testpmd> flush_flow_director (port_id) 2807 2808Example, to flush all flow director filter on port 0:: 2809 2810 testpmd> flush_flow_director 0 2811 2812flow_director_mask 2813~~~~~~~~~~~~~~~~~~ 2814 2815Set flow director's input masks:: 2816 2817 flow_director_mask (port_id) mode IP vlan (vlan_value) \ 2818 src_mask (ipv4_src) (ipv6_src) (src_port) \ 2819 dst_mask (ipv4_dst) (ipv6_dst) (dst_port) 2820 2821 flow_director_mask (port_id) mode MAC-VLAN vlan (vlan_value) 2822 2823 flow_director_mask (port_id) mode Tunnel vlan (vlan_value) \ 2824 mac (mac_value) tunnel-type (tunnel_type_value) \ 2825 tunnel-id (tunnel_id_value) 2826 2827Example, to set flow director mask on port 0:: 2828 2829 testpmd> flow_director_mask 0 mode IP vlan 0xefff \ 2830 src_mask 255.255.255.255 \ 2831 FFFF:FFFF:FFFF:FFFF:FFFF:FFFF:FFFF:FFFF 0xFFFF \ 2832 dst_mask 255.255.255.255 \ 2833 FFFF:FFFF:FFFF:FFFF:FFFF:FFFF:FFFF:FFFF 0xFFFF 2834 2835flow_director_flex_mask 2836~~~~~~~~~~~~~~~~~~~~~~~ 2837 2838set masks of flow director's flexible payload based on certain flow type:: 2839 2840 testpmd> flow_director_flex_mask (port_id) \ 2841 flow (none|ipv4-other|ipv4-frag|ipv4-tcp|ipv4-udp|ipv4-sctp| \ 2842 ipv6-other|ipv6-frag|ipv6-tcp|ipv6-udp|ipv6-sctp| \ 2843 l2_payload|all) (mask) 2844 2845Example, to set flow director's flex mask for all flow type on port 0:: 2846 2847 testpmd> flow_director_flex_mask 0 flow all \ 2848 (0xff,0xff,0,0,0,0,0,0,0,0,0,0,0,0,0,0) 2849 2850 2851flow_director_flex_payload 2852~~~~~~~~~~~~~~~~~~~~~~~~~~ 2853 2854Configure flexible payload selection:: 2855 2856 flow_director_flex_payload (port_id) (raw|l2|l3|l4) (config) 2857 2858For example, to select the first 16 bytes from the offset 4 (bytes) of packet's payload as flexible payload:: 2859 2860 testpmd> flow_director_flex_payload 0 l4 \ 2861 (4,5,6,7,8,9,10,11,12,13,14,15,16,17,18,19) 2862 2863get_sym_hash_ena_per_port 2864~~~~~~~~~~~~~~~~~~~~~~~~~ 2865 2866Get symmetric hash enable configuration per port:: 2867 2868 get_sym_hash_ena_per_port (port_id) 2869 2870For example, to get symmetric hash enable configuration of port 1:: 2871 2872 testpmd> get_sym_hash_ena_per_port 1 2873 2874set_sym_hash_ena_per_port 2875~~~~~~~~~~~~~~~~~~~~~~~~~ 2876 2877Set symmetric hash enable configuration per port to enable or disable:: 2878 2879 set_sym_hash_ena_per_port (port_id) (enable|disable) 2880 2881For example, to set symmetric hash enable configuration of port 1 to enable:: 2882 2883 testpmd> set_sym_hash_ena_per_port 1 enable 2884 2885get_hash_global_config 2886~~~~~~~~~~~~~~~~~~~~~~ 2887 2888Get the global configurations of hash filters:: 2889 2890 get_hash_global_config (port_id) 2891 2892For example, to get the global configurations of hash filters of port 1:: 2893 2894 testpmd> get_hash_global_config 1 2895 2896set_hash_global_config 2897~~~~~~~~~~~~~~~~~~~~~~ 2898 2899Set the global configurations of hash filters:: 2900 2901 set_hash_global_config (port_id) (toeplitz|simple_xor|default) \ 2902 (ipv4|ipv4-frag|ipv4-tcp|ipv4-udp|ipv4-sctp|ipv4-other|ipv6|ipv6-frag| \ 2903 ipv6-tcp|ipv6-udp|ipv6-sctp|ipv6-other|l2_payload|<flow_id>) \ 2904 (enable|disable) 2905 2906For example, to enable simple_xor for flow type of ipv6 on port 2:: 2907 2908 testpmd> set_hash_global_config 2 simple_xor ipv6 enable 2909 2910set_hash_input_set 2911~~~~~~~~~~~~~~~~~~ 2912 2913Set the input set for hash:: 2914 2915 set_hash_input_set (port_id) (ipv4-frag|ipv4-tcp|ipv4-udp|ipv4-sctp| \ 2916 ipv4-other|ipv6-frag|ipv6-tcp|ipv6-udp|ipv6-sctp|ipv6-other| \ 2917 l2_payload|<flow_id>) (ovlan|ivlan|src-ipv4|dst-ipv4|src-ipv6|dst-ipv6| \ 2918 ipv4-tos|ipv4-proto|ipv6-tc|ipv6-next-header|udp-src-port|udp-dst-port| \ 2919 tcp-src-port|tcp-dst-port|sctp-src-port|sctp-dst-port|sctp-veri-tag| \ 2920 udp-key|gre-key|fld-1st|fld-2nd|fld-3rd|fld-4th|fld-5th|fld-6th|fld-7th| \ 2921 fld-8th|none) (select|add) 2922 2923For example, to add source IP to hash input set for flow type of ipv4-udp on port 0:: 2924 2925 testpmd> set_hash_input_set 0 ipv4-udp src-ipv4 add 2926 2927set_fdir_input_set 2928~~~~~~~~~~~~~~~~~~ 2929 2930The Flow Director filters can match the different fields for different type of packet, i.e. specific input set 2931on per flow type and the flexible payload. This command can be used to change input set for each flow type. 2932 2933Set the input set for flow director:: 2934 2935 set_fdir_input_set (port_id) (ipv4-frag|ipv4-tcp|ipv4-udp|ipv4-sctp| \ 2936 ipv4-other|ipv6|ipv6-frag|ipv6-tcp|ipv6-udp|ipv6-sctp|ipv6-other| \ 2937 l2_payload|<flow_id>) (ivlan|ethertype|src-ipv4|dst-ipv4|src-ipv6|dst-ipv6| \ 2938 ipv4-tos|ipv4-proto|ipv4-ttl|ipv6-tc|ipv6-next-header|ipv6-hop-limits| \ 2939 tudp-src-port|udp-dst-port|cp-src-port|tcp-dst-port|sctp-src-port| \ 2940 sctp-dst-port|sctp-veri-tag|none) (select|add) 2941 2942For example to add source IP to FD input set for flow type of ipv4-udp on port 0:: 2943 2944 testpmd> set_fdir_input_set 0 ipv4-udp src-ipv4 add 2945 2946global_config 2947~~~~~~~~~~~~~ 2948 2949Set different GRE key length for input set:: 2950 2951 global_config (port_id) gre-key-len (number in bytes) 2952 2953For example to set GRE key length for input set to 4 bytes on port 0:: 2954 2955 testpmd> global_config 0 gre-key-len 4 2956 2957 2958.. _testpmd_rte_flow: 2959 2960Flow rules management 2961--------------------- 2962 2963Control of the generic flow API (*rte_flow*) is fully exposed through the 2964``flow`` command (validation, creation, destruction, queries and operation 2965modes). 2966 2967Considering *rte_flow* overlaps with all `Filter Functions`_, using both 2968features simultaneously may cause undefined side-effects and is therefore 2969not recommended. 2970 2971``flow`` syntax 2972~~~~~~~~~~~~~~~ 2973 2974Because the ``flow`` command uses dynamic tokens to handle the large number 2975of possible flow rules combinations, its behavior differs slightly from 2976other commands, in particular: 2977 2978- Pressing *?* or the *<tab>* key displays contextual help for the current 2979 token, not that of the entire command. 2980 2981- Optional and repeated parameters are supported (provided they are listed 2982 in the contextual help). 2983 2984The first parameter stands for the operation mode. Possible operations and 2985their general syntax are described below. They are covered in detail in the 2986following sections. 2987 2988- Check whether a flow rule can be created:: 2989 2990 flow validate {port_id} 2991 [group {group_id}] [priority {level}] [ingress] [egress] 2992 pattern {item} [/ {item} [...]] / end 2993 actions {action} [/ {action} [...]] / end 2994 2995- Create a flow rule:: 2996 2997 flow create {port_id} 2998 [group {group_id}] [priority {level}] [ingress] [egress] 2999 pattern {item} [/ {item} [...]] / end 3000 actions {action} [/ {action} [...]] / end 3001 3002- Destroy specific flow rules:: 3003 3004 flow destroy {port_id} rule {rule_id} [...] 3005 3006- Destroy all flow rules:: 3007 3008 flow flush {port_id} 3009 3010- Query an existing flow rule:: 3011 3012 flow query {port_id} {rule_id} {action} 3013 3014- List existing flow rules sorted by priority, filtered by group 3015 identifiers:: 3016 3017 flow list {port_id} [group {group_id}] [...] 3018 3019- Restrict ingress traffic to the defined flow rules:: 3020 3021 flow isolate {port_id} {boolean} 3022 3023Validating flow rules 3024~~~~~~~~~~~~~~~~~~~~~ 3025 3026``flow validate`` reports whether a flow rule would be accepted by the 3027underlying device in its current state but stops short of creating it. It is 3028bound to ``rte_flow_validate()``:: 3029 3030 flow validate {port_id} 3031 [group {group_id}] [priority {level}] [ingress] [egress] 3032 pattern {item} [/ {item} [...]] / end 3033 actions {action} [/ {action} [...]] / end 3034 3035If successful, it will show:: 3036 3037 Flow rule validated 3038 3039Otherwise it will show an error message of the form:: 3040 3041 Caught error type [...] ([...]): [...] 3042 3043This command uses the same parameters as ``flow create``, their format is 3044described in `Creating flow rules`_. 3045 3046Check whether redirecting any Ethernet packet received on port 0 to RX queue 3047index 6 is supported:: 3048 3049 testpmd> flow validate 0 ingress pattern eth / end 3050 actions queue index 6 / end 3051 Flow rule validated 3052 testpmd> 3053 3054Port 0 does not support TCPv6 rules:: 3055 3056 testpmd> flow validate 0 ingress pattern eth / ipv6 / tcp / end 3057 actions drop / end 3058 Caught error type 9 (specific pattern item): Invalid argument 3059 testpmd> 3060 3061Creating flow rules 3062~~~~~~~~~~~~~~~~~~~ 3063 3064``flow create`` validates and creates the specified flow rule. It is bound 3065to ``rte_flow_create()``:: 3066 3067 flow create {port_id} 3068 [group {group_id}] [priority {level}] [ingress] [egress] 3069 pattern {item} [/ {item} [...]] / end 3070 actions {action} [/ {action} [...]] / end 3071 3072If successful, it will return a flow rule ID usable with other commands:: 3073 3074 Flow rule #[...] created 3075 3076Otherwise it will show an error message of the form:: 3077 3078 Caught error type [...] ([...]): [...] 3079 3080Parameters describe in the following order: 3081 3082- Attributes (*group*, *priority*, *ingress*, *egress* tokens). 3083- A matching pattern, starting with the *pattern* token and terminated by an 3084 *end* pattern item. 3085- Actions, starting with the *actions* token and terminated by an *end* 3086 action. 3087 3088These translate directly to *rte_flow* objects provided as-is to the 3089underlying functions. 3090 3091The shortest valid definition only comprises mandatory tokens:: 3092 3093 testpmd> flow create 0 pattern end actions end 3094 3095Note that PMDs may refuse rules that essentially do nothing such as this 3096one. 3097 3098**All unspecified object values are automatically initialized to 0.** 3099 3100Attributes 3101^^^^^^^^^^ 3102 3103These tokens affect flow rule attributes (``struct rte_flow_attr``) and are 3104specified before the ``pattern`` token. 3105 3106- ``group {group id}``: priority group. 3107- ``priority {level}``: priority level within group. 3108- ``ingress``: rule applies to ingress traffic. 3109- ``egress``: rule applies to egress traffic. 3110 3111Each instance of an attribute specified several times overrides the previous 3112value as shown below (group 4 is used):: 3113 3114 testpmd> flow create 0 group 42 group 24 group 4 [...] 3115 3116Note that once enabled, ``ingress`` and ``egress`` cannot be disabled. 3117 3118While not specifying a direction is an error, some rules may allow both 3119simultaneously. 3120 3121Most rules affect RX therefore contain the ``ingress`` token:: 3122 3123 testpmd> flow create 0 ingress pattern [...] 3124 3125Matching pattern 3126^^^^^^^^^^^^^^^^ 3127 3128A matching pattern starts after the ``pattern`` token. It is made of pattern 3129items and is terminated by a mandatory ``end`` item. 3130 3131Items are named after their type (*RTE_FLOW_ITEM_TYPE_* from ``enum 3132rte_flow_item_type``). 3133 3134The ``/`` token is used as a separator between pattern items as shown 3135below:: 3136 3137 testpmd> flow create 0 ingress pattern eth / ipv4 / udp / end [...] 3138 3139Note that protocol items like these must be stacked from lowest to highest 3140layer to make sense. For instance, the following rule is either invalid or 3141unlikely to match any packet:: 3142 3143 testpmd> flow create 0 ingress pattern eth / udp / ipv4 / end [...] 3144 3145More information on these restrictions can be found in the *rte_flow* 3146documentation. 3147 3148Several items support additional specification structures, for example 3149``ipv4`` allows specifying source and destination addresses as follows:: 3150 3151 testpmd> flow create 0 ingress pattern eth / ipv4 src is 10.1.1.1 3152 dst is 10.2.0.0 / end [...] 3153 3154This rule matches all IPv4 traffic with the specified properties. 3155 3156In this example, ``src`` and ``dst`` are field names of the underlying 3157``struct rte_flow_item_ipv4`` object. All item properties can be specified 3158in a similar fashion. 3159 3160The ``is`` token means that the subsequent value must be matched exactly, 3161and assigns ``spec`` and ``mask`` fields in ``struct rte_flow_item`` 3162accordingly. Possible assignment tokens are: 3163 3164- ``is``: match value perfectly (with full bit-mask). 3165- ``spec``: match value according to configured bit-mask. 3166- ``last``: specify upper bound to establish a range. 3167- ``mask``: specify bit-mask with relevant bits set to one. 3168- ``prefix``: generate bit-mask from a prefix length. 3169 3170These yield identical results:: 3171 3172 ipv4 src is 10.1.1.1 3173 3174:: 3175 3176 ipv4 src spec 10.1.1.1 src mask 255.255.255.255 3177 3178:: 3179 3180 ipv4 src spec 10.1.1.1 src prefix 32 3181 3182:: 3183 3184 ipv4 src is 10.1.1.1 src last 10.1.1.1 # range with a single value 3185 3186:: 3187 3188 ipv4 src is 10.1.1.1 src last 0 # 0 disables range 3189 3190Inclusive ranges can be defined with ``last``:: 3191 3192 ipv4 src is 10.1.1.1 src last 10.2.3.4 # 10.1.1.1 to 10.2.3.4 3193 3194Note that ``mask`` affects both ``spec`` and ``last``:: 3195 3196 ipv4 src is 10.1.1.1 src last 10.2.3.4 src mask 255.255.0.0 3197 # matches 10.1.0.0 to 10.2.255.255 3198 3199Properties can be modified multiple times:: 3200 3201 ipv4 src is 10.1.1.1 src is 10.1.2.3 src is 10.2.3.4 # matches 10.2.3.4 3202 3203:: 3204 3205 ipv4 src is 10.1.1.1 src prefix 24 src prefix 16 # matches 10.1.0.0/16 3206 3207Pattern items 3208^^^^^^^^^^^^^ 3209 3210This section lists supported pattern items and their attributes, if any. 3211 3212- ``end``: end list of pattern items. 3213 3214- ``void``: no-op pattern item. 3215 3216- ``invert``: perform actions when pattern does not match. 3217 3218- ``any``: match any protocol for the current layer. 3219 3220 - ``num {unsigned}``: number of layers covered. 3221 3222- ``pf``: match packets addressed to the physical function. 3223 3224- ``vf``: match packets addressed to a virtual function ID. 3225 3226 - ``id {unsigned}``: destination VF ID. 3227 3228- ``port``: device-specific physical port index to use. 3229 3230 - ``index {unsigned}``: physical port index. 3231 3232- ``raw``: match an arbitrary byte string. 3233 3234 - ``relative {boolean}``: look for pattern after the previous item. 3235 - ``search {boolean}``: search pattern from offset (see also limit). 3236 - ``offset {integer}``: absolute or relative offset for pattern. 3237 - ``limit {unsigned}``: search area limit for start of pattern. 3238 - ``pattern {string}``: byte string to look for. 3239 3240- ``eth``: match Ethernet header. 3241 3242 - ``dst {MAC-48}``: destination MAC. 3243 - ``src {MAC-48}``: source MAC. 3244 - ``type {unsigned}``: EtherType. 3245 3246- ``vlan``: match 802.1Q/ad VLAN tag. 3247 3248 - ``tpid {unsigned}``: tag protocol identifier. 3249 - ``tci {unsigned}``: tag control information. 3250 - ``pcp {unsigned}``: priority code point. 3251 - ``dei {unsigned}``: drop eligible indicator. 3252 - ``vid {unsigned}``: VLAN identifier. 3253 3254- ``ipv4``: match IPv4 header. 3255 3256 - ``tos {unsigned}``: type of service. 3257 - ``ttl {unsigned}``: time to live. 3258 - ``proto {unsigned}``: next protocol ID. 3259 - ``src {ipv4 address}``: source address. 3260 - ``dst {ipv4 address}``: destination address. 3261 3262- ``ipv6``: match IPv6 header. 3263 3264 - ``tc {unsigned}``: traffic class. 3265 - ``flow {unsigned}``: flow label. 3266 - ``proto {unsigned}``: protocol (next header). 3267 - ``hop {unsigned}``: hop limit. 3268 - ``src {ipv6 address}``: source address. 3269 - ``dst {ipv6 address}``: destination address. 3270 3271- ``icmp``: match ICMP header. 3272 3273 - ``type {unsigned}``: ICMP packet type. 3274 - ``code {unsigned}``: ICMP packet code. 3275 3276- ``udp``: match UDP header. 3277 3278 - ``src {unsigned}``: UDP source port. 3279 - ``dst {unsigned}``: UDP destination port. 3280 3281- ``tcp``: match TCP header. 3282 3283 - ``src {unsigned}``: TCP source port. 3284 - ``dst {unsigned}``: TCP destination port. 3285 3286- ``sctp``: match SCTP header. 3287 3288 - ``src {unsigned}``: SCTP source port. 3289 - ``dst {unsigned}``: SCTP destination port. 3290 - ``tag {unsigned}``: validation tag. 3291 - ``cksum {unsigned}``: checksum. 3292 3293- ``vxlan``: match VXLAN header. 3294 3295 - ``vni {unsigned}``: VXLAN identifier. 3296 3297- ``e_tag``: match IEEE 802.1BR E-Tag header. 3298 3299 - ``grp_ecid_b {unsigned}``: GRP and E-CID base. 3300 3301- ``nvgre``: match NVGRE header. 3302 3303 - ``tni {unsigned}``: virtual subnet ID. 3304 3305- ``mpls``: match MPLS header. 3306 3307 - ``label {unsigned}``: MPLS label. 3308 3309- ``gre``: match GRE header. 3310 3311 - ``protocol {unsigned}``: protocol type. 3312 3313- ``fuzzy``: fuzzy pattern match, expect faster than default. 3314 3315 - ``thresh {unsigned}``: accuracy threshold. 3316 3317- ``gtp``, ``gtpc``, ``gtpu``: match GTPv1 header. 3318 3319 - ``teid {unsigned}``: tunnel endpoint identifier. 3320 3321- ``geneve``: match GENEVE header. 3322 3323 - ``vni {unsigned}``: virtual network identifier. 3324 - ``protocol {unsigned}``: protocol type. 3325 3326Actions list 3327^^^^^^^^^^^^ 3328 3329A list of actions starts after the ``actions`` token in the same fashion as 3330`Matching pattern`_; actions are separated by ``/`` tokens and the list is 3331terminated by a mandatory ``end`` action. 3332 3333Actions are named after their type (*RTE_FLOW_ACTION_TYPE_* from ``enum 3334rte_flow_action_type``). 3335 3336Dropping all incoming UDPv4 packets can be expressed as follows:: 3337 3338 testpmd> flow create 0 ingress pattern eth / ipv4 / udp / end 3339 actions drop / end 3340 3341Several actions have configurable properties which must be specified when 3342there is no valid default value. For example, ``queue`` requires a target 3343queue index. 3344 3345This rule redirects incoming UDPv4 traffic to queue index 6:: 3346 3347 testpmd> flow create 0 ingress pattern eth / ipv4 / udp / end 3348 actions queue index 6 / end 3349 3350While this one could be rejected by PMDs (unspecified queue index):: 3351 3352 testpmd> flow create 0 ingress pattern eth / ipv4 / udp / end 3353 actions queue / end 3354 3355As defined by *rte_flow*, the list is not ordered, all actions of a given 3356rule are performed simultaneously. These are equivalent:: 3357 3358 queue index 6 / void / mark id 42 / end 3359 3360:: 3361 3362 void / mark id 42 / queue index 6 / end 3363 3364All actions in a list should have different types, otherwise only the last 3365action of a given type is taken into account:: 3366 3367 queue index 4 / queue index 5 / queue index 6 / end # will use queue 6 3368 3369:: 3370 3371 drop / drop / drop / end # drop is performed only once 3372 3373:: 3374 3375 mark id 42 / queue index 3 / mark id 24 / end # mark will be 24 3376 3377Considering they are performed simultaneously, opposite and overlapping 3378actions can sometimes be combined when the end result is unambiguous:: 3379 3380 drop / queue index 6 / end # drop has no effect 3381 3382:: 3383 3384 drop / dup index 6 / end # same as above 3385 3386:: 3387 3388 queue index 6 / rss queues 6 7 8 / end # queue has no effect 3389 3390:: 3391 3392 drop / passthru / end # drop has no effect 3393 3394Note that PMDs may still refuse such combinations. 3395 3396Actions 3397^^^^^^^ 3398 3399This section lists supported actions and their attributes, if any. 3400 3401- ``end``: end list of actions. 3402 3403- ``void``: no-op action. 3404 3405- ``passthru``: let subsequent rule process matched packets. 3406 3407- ``mark``: attach 32 bit value to packets. 3408 3409 - ``id {unsigned}``: 32 bit value to return with packets. 3410 3411- ``flag``: flag packets. 3412 3413- ``queue``: assign packets to a given queue index. 3414 3415 - ``index {unsigned}``: queue index to use. 3416 3417- ``drop``: drop packets (note: passthru has priority). 3418 3419- ``count``: enable counters for this rule. 3420 3421- ``dup``: duplicate packets to a given queue index. 3422 3423 - ``index {unsigned}``: queue index to duplicate packets to. 3424 3425- ``rss``: spread packets among several queues. 3426 3427 - ``types [{RSS hash type} [...]] end``: RSS hash types, allowed tokens 3428 are the same as `set_hash_input_set`_, an empty list means none (0). 3429 3430 - ``key {string}``: RSS hash key, overrides ``key_len``. 3431 3432 - ``key_len {unsigned}``: RSS hash key length in bytes, can be used in 3433 conjunction with ``key`` to pad or truncate it. 3434 3435 - ``queues [{unsigned} [...]] end``: queue indices to use. 3436 3437- ``pf``: redirect packets to physical device function. 3438 3439- ``vf``: redirect packets to virtual device function. 3440 3441 - ``original {boolean}``: use original VF ID if possible. 3442 - ``id {unsigned}``: VF ID to redirect packets to. 3443 3444Destroying flow rules 3445~~~~~~~~~~~~~~~~~~~~~ 3446 3447``flow destroy`` destroys one or more rules from their rule ID (as returned 3448by ``flow create``), this command calls ``rte_flow_destroy()`` as many 3449times as necessary:: 3450 3451 flow destroy {port_id} rule {rule_id} [...] 3452 3453If successful, it will show:: 3454 3455 Flow rule #[...] destroyed 3456 3457It does not report anything for rule IDs that do not exist. The usual error 3458message is shown when a rule cannot be destroyed:: 3459 3460 Caught error type [...] ([...]): [...] 3461 3462``flow flush`` destroys all rules on a device and does not take extra 3463arguments. It is bound to ``rte_flow_flush()``:: 3464 3465 flow flush {port_id} 3466 3467Any errors are reported as above. 3468 3469Creating several rules and destroying them:: 3470 3471 testpmd> flow create 0 ingress pattern eth / ipv6 / end 3472 actions queue index 2 / end 3473 Flow rule #0 created 3474 testpmd> flow create 0 ingress pattern eth / ipv4 / end 3475 actions queue index 3 / end 3476 Flow rule #1 created 3477 testpmd> flow destroy 0 rule 0 rule 1 3478 Flow rule #1 destroyed 3479 Flow rule #0 destroyed 3480 testpmd> 3481 3482The same result can be achieved using ``flow flush``:: 3483 3484 testpmd> flow create 0 ingress pattern eth / ipv6 / end 3485 actions queue index 2 / end 3486 Flow rule #0 created 3487 testpmd> flow create 0 ingress pattern eth / ipv4 / end 3488 actions queue index 3 / end 3489 Flow rule #1 created 3490 testpmd> flow flush 0 3491 testpmd> 3492 3493Non-existent rule IDs are ignored:: 3494 3495 testpmd> flow create 0 ingress pattern eth / ipv6 / end 3496 actions queue index 2 / end 3497 Flow rule #0 created 3498 testpmd> flow create 0 ingress pattern eth / ipv4 / end 3499 actions queue index 3 / end 3500 Flow rule #1 created 3501 testpmd> flow destroy 0 rule 42 rule 10 rule 2 3502 testpmd> 3503 testpmd> flow destroy 0 rule 0 3504 Flow rule #0 destroyed 3505 testpmd> 3506 3507Querying flow rules 3508~~~~~~~~~~~~~~~~~~~ 3509 3510``flow query`` queries a specific action of a flow rule having that 3511ability. Such actions collect information that can be reported using this 3512command. It is bound to ``rte_flow_query()``:: 3513 3514 flow query {port_id} {rule_id} {action} 3515 3516If successful, it will display either the retrieved data for known actions 3517or the following message:: 3518 3519 Cannot display result for action type [...] ([...]) 3520 3521Otherwise, it will complain either that the rule does not exist or that some 3522error occurred:: 3523 3524 Flow rule #[...] not found 3525 3526:: 3527 3528 Caught error type [...] ([...]): [...] 3529 3530Currently only the ``count`` action is supported. This action reports the 3531number of packets that hit the flow rule and the total number of bytes. Its 3532output has the following format:: 3533 3534 count: 3535 hits_set: [...] # whether "hits" contains a valid value 3536 bytes_set: [...] # whether "bytes" contains a valid value 3537 hits: [...] # number of packets 3538 bytes: [...] # number of bytes 3539 3540Querying counters for TCPv6 packets redirected to queue 6:: 3541 3542 testpmd> flow create 0 ingress pattern eth / ipv6 / tcp / end 3543 actions queue index 6 / count / end 3544 Flow rule #4 created 3545 testpmd> flow query 0 4 count 3546 count: 3547 hits_set: 1 3548 bytes_set: 0 3549 hits: 386446 3550 bytes: 0 3551 testpmd> 3552 3553Listing flow rules 3554~~~~~~~~~~~~~~~~~~ 3555 3556``flow list`` lists existing flow rules sorted by priority and optionally 3557filtered by group identifiers:: 3558 3559 flow list {port_id} [group {group_id}] [...] 3560 3561This command only fails with the following message if the device does not 3562exist:: 3563 3564 Invalid port [...] 3565 3566Output consists of a header line followed by a short description of each 3567flow rule, one per line. There is no output at all when no flow rules are 3568configured on the device:: 3569 3570 ID Group Prio Attr Rule 3571 [...] [...] [...] [...] [...] 3572 3573``Attr`` column flags: 3574 3575- ``i`` for ``ingress``. 3576- ``e`` for ``egress``. 3577 3578Creating several flow rules and listing them:: 3579 3580 testpmd> flow create 0 ingress pattern eth / ipv4 / end 3581 actions queue index 6 / end 3582 Flow rule #0 created 3583 testpmd> flow create 0 ingress pattern eth / ipv6 / end 3584 actions queue index 2 / end 3585 Flow rule #1 created 3586 testpmd> flow create 0 priority 5 ingress pattern eth / ipv4 / udp / end 3587 actions rss queues 6 7 8 end / end 3588 Flow rule #2 created 3589 testpmd> flow list 0 3590 ID Group Prio Attr Rule 3591 0 0 0 i- ETH IPV4 => QUEUE 3592 1 0 0 i- ETH IPV6 => QUEUE 3593 2 0 5 i- ETH IPV4 UDP => RSS 3594 testpmd> 3595 3596Rules are sorted by priority (i.e. group ID first, then priority level):: 3597 3598 testpmd> flow list 1 3599 ID Group Prio Attr Rule 3600 0 0 0 i- ETH => COUNT 3601 6 0 500 i- ETH IPV6 TCP => DROP COUNT 3602 5 0 1000 i- ETH IPV6 ICMP => QUEUE 3603 1 24 0 i- ETH IPV4 UDP => QUEUE 3604 4 24 10 i- ETH IPV4 TCP => DROP 3605 3 24 20 i- ETH IPV4 => DROP 3606 2 24 42 i- ETH IPV4 UDP => QUEUE 3607 7 63 0 i- ETH IPV6 UDP VXLAN => MARK QUEUE 3608 testpmd> 3609 3610Output can be limited to specific groups:: 3611 3612 testpmd> flow list 1 group 0 group 63 3613 ID Group Prio Attr Rule 3614 0 0 0 i- ETH => COUNT 3615 6 0 500 i- ETH IPV6 TCP => DROP COUNT 3616 5 0 1000 i- ETH IPV6 ICMP => QUEUE 3617 7 63 0 i- ETH IPV6 UDP VXLAN => MARK QUEUE 3618 testpmd> 3619 3620Toggling isolated mode 3621~~~~~~~~~~~~~~~~~~~~~~ 3622 3623``flow isolate`` can be used to tell the underlying PMD that ingress traffic 3624must only be injected from the defined flow rules; that no default traffic 3625is expected outside those rules and the driver is free to assign more 3626resources to handle them. It is bound to ``rte_flow_isolate()``:: 3627 3628 flow isolate {port_id} {boolean} 3629 3630If successful, enabling or disabling isolated mode shows either:: 3631 3632 Ingress traffic on port [...] 3633 is now restricted to the defined flow rules 3634 3635Or:: 3636 3637 Ingress traffic on port [...] 3638 is not restricted anymore to the defined flow rules 3639 3640Otherwise, in case of error:: 3641 3642 Caught error type [...] ([...]): [...] 3643 3644Mainly due to its side effects, PMDs supporting this mode may not have the 3645ability to toggle it more than once without reinitializing affected ports 3646first (e.g. by exiting testpmd). 3647 3648Enabling isolated mode:: 3649 3650 testpmd> flow isolate 0 true 3651 Ingress traffic on port 0 is now restricted to the defined flow rules 3652 testpmd> 3653 3654Disabling isolated mode:: 3655 3656 testpmd> flow isolate 0 false 3657 Ingress traffic on port 0 is not restricted anymore to the defined flow rules 3658 testpmd> 3659 3660Sample QinQ flow rules 3661~~~~~~~~~~~~~~~~~~~~~~ 3662 3663Before creating QinQ rule(s) the following commands should be issued to enable QinQ:: 3664 3665 testpmd> port stop 0 3666 testpmd> vlan set qinq on 0 3667 3668The above command sets the inner and outer TPID's to 0x8100. 3669 3670To change the TPID's the following commands should be used:: 3671 3672 testpmd> vlan set outer tpid 0xa100 0 3673 testpmd> vlan set inner tpid 0x9100 0 3674 testpmd> port start 0 3675 3676Validate and create a QinQ rule on port 0 to steer traffic to a VF queue in a VM. 3677 3678:: 3679 3680 testpmd> flow validate 0 ingress pattern eth / vlan tci is 123 / 3681 vlan tci is 456 / end actions vf id 1 / queue index 0 / end 3682 Flow rule #0 validated 3683 3684 testpmd> flow create 0 ingress pattern eth / vlan tci is 4 / 3685 vlan tci is 456 / end actions vf id 123 / queue index 0 / end 3686 Flow rule #0 created 3687 3688 testpmd> flow list 0 3689 ID Group Prio Attr Rule 3690 0 0 0 i- ETH VLAN VLAN=>VF QUEUE 3691 3692Validate and create a QinQ rule on port 0 to steer traffic to a queue on the host. 3693 3694:: 3695 3696 testpmd> flow validate 0 ingress pattern eth / vlan tci is 321 / 3697 vlan tci is 654 / end actions pf / queue index 0 / end 3698 Flow rule #1 validated 3699 3700 testpmd> flow create 0 ingress pattern eth / vlan tci is 321 / 3701 vlan tci is 654 / end actions pf / queue index 1 / end 3702 Flow rule #1 created 3703 3704 testpmd> flow list 0 3705 ID Group Prio Attr Rule 3706 0 0 0 i- ETH VLAN VLAN=>VF QUEUE 3707 1 0 0 i- ETH VLAN VLAN=>PF QUEUE 3708