1.. SPDX-License-Identifier: BSD-3-Clause 2 Copyright(c) 2010-2016 Intel Corporation. 3 4.. _testpmd_runtime: 5 6Testpmd Runtime Functions 7========================= 8 9Where the testpmd application is started in interactive mode, (``-i|--interactive``), 10it displays a prompt that can be used to start and stop forwarding, 11configure the application, display statistics (including the extended NIC 12statistics aka xstats) , set the Flow Director and other tasks:: 13 14 testpmd> 15 16The testpmd prompt has some, limited, readline support. 17Common bash command-line functions such as ``Ctrl+a`` and ``Ctrl+e`` to go to the start and end of the prompt line are supported 18as well as access to the command history via the up-arrow. 19 20There is also support for tab completion. 21If you type a partial command and hit ``<TAB>`` you get a list of the available completions: 22 23.. code-block:: console 24 25 testpmd> show port <TAB> 26 27 info [Mul-choice STRING]: show|clear port info|stats|xstats|fdir|stat_qmap|dcb_tc|cap X 28 info [Mul-choice STRING]: show|clear port info|stats|xstats|fdir|stat_qmap|dcb_tc|cap all 29 stats [Mul-choice STRING]: show|clear port info|stats|xstats|fdir|stat_qmap|dcb_tc|cap X 30 stats [Mul-choice STRING]: show|clear port info|stats|xstats|fdir|stat_qmap|dcb_tc|cap all 31 ... 32 33 34.. note:: 35 36 Some examples in this document are too long to fit on one line are are shown wrapped at `"\\"` for display purposes:: 37 38 testpmd> set flow_ctrl rx (on|off) tx (on|off) (high_water) (low_water) \ 39 (pause_time) (send_xon) (port_id) 40 41In the real ``testpmd>`` prompt these commands should be on a single line. 42 43Help Functions 44-------------- 45 46The testpmd has on-line help for the functions that are available at runtime. 47These are divided into sections and can be accessed using help, help section or help all: 48 49.. code-block:: console 50 51 testpmd> help 52 53 help control : Start and stop forwarding. 54 help display : Displaying port, stats and config information. 55 help config : Configuration information. 56 help ports : Configuring ports. 57 help registers : Reading and setting port registers. 58 help filters : Filters configuration help. 59 help all : All of the above sections. 60 61 62Command File Functions 63---------------------- 64 65To facilitate loading large number of commands or to avoid cutting and pasting where not 66practical or possible testpmd supports alternative methods for executing commands. 67 68* If started with the ``--cmdline-file=FILENAME`` command line argument testpmd 69 will execute all CLI commands contained within the file immediately before 70 starting packet forwarding or entering interactive mode. 71 72.. code-block:: console 73 74 ./testpmd -n4 -r2 ... -- -i --cmdline-file=/home/ubuntu/flow-create-commands.txt 75 Interactive-mode selected 76 CLI commands to be read from /home/ubuntu/flow-create-commands.txt 77 Configuring Port 0 (socket 0) 78 Port 0: 7C:FE:90:CB:74:CE 79 Configuring Port 1 (socket 0) 80 Port 1: 7C:FE:90:CB:74:CA 81 Checking link statuses... 82 Port 0 Link Up - speed 10000 Mbps - full-duplex 83 Port 1 Link Up - speed 10000 Mbps - full-duplex 84 Done 85 Flow rule #0 created 86 Flow rule #1 created 87 ... 88 ... 89 Flow rule #498 created 90 Flow rule #499 created 91 Read all CLI commands from /home/ubuntu/flow-create-commands.txt 92 testpmd> 93 94 95* At run-time additional commands can be loaded in bulk by invoking the ``load FILENAME`` 96 command. 97 98.. code-block:: console 99 100 testpmd> load /home/ubuntu/flow-create-commands.txt 101 Flow rule #0 created 102 Flow rule #1 created 103 ... 104 ... 105 Flow rule #498 created 106 Flow rule #499 created 107 Read all CLI commands from /home/ubuntu/flow-create-commands.txt 108 testpmd> 109 110 111In all cases output from any included command will be displayed as standard output. 112Execution will continue until the end of the file is reached regardless of 113whether any errors occur. The end user must examine the output to determine if 114any failures occurred. 115 116 117Control Functions 118----------------- 119 120start 121~~~~~ 122 123Start packet forwarding with current configuration:: 124 125 testpmd> start 126 127start tx_first 128~~~~~~~~~~~~~~ 129 130Start packet forwarding with current configuration after sending specified number of bursts of packets:: 131 132 testpmd> start tx_first (""|burst_num) 133 134The default burst number is 1 when ``burst_num`` not presented. 135 136stop 137~~~~ 138 139Stop packet forwarding, and display accumulated statistics:: 140 141 testpmd> stop 142 143quit 144~~~~ 145 146Quit to prompt:: 147 148 testpmd> quit 149 150 151Display Functions 152----------------- 153 154The functions in the following sections are used to display information about the 155testpmd configuration or the NIC status. 156 157show port 158~~~~~~~~~ 159 160Display information for a given port or all ports:: 161 162 testpmd> show port (info|stats|xstats|fdir|stat_qmap|dcb_tc|cap) (port_id|all) 163 164The available information categories are: 165 166* ``info``: General port information such as MAC address. 167 168* ``stats``: RX/TX statistics. 169 170* ``xstats``: RX/TX extended NIC statistics. 171 172* ``fdir``: Flow Director information and statistics. 173 174* ``stat_qmap``: Queue statistics mapping. 175 176* ``dcb_tc``: DCB information such as TC mapping. 177 178* ``cap``: Supported offload capabilities. 179 180For example: 181 182.. code-block:: console 183 184 testpmd> show port info 0 185 186 ********************* Infos for port 0 ********************* 187 188 MAC address: XX:XX:XX:XX:XX:XX 189 Connect to socket: 0 190 memory allocation on the socket: 0 191 Link status: up 192 Link speed: 40000 Mbps 193 Link duplex: full-duplex 194 Promiscuous mode: enabled 195 Allmulticast mode: disabled 196 Maximum number of MAC addresses: 64 197 Maximum number of MAC addresses of hash filtering: 0 198 VLAN offload: 199 strip on 200 filter on 201 qinq(extend) off 202 Redirection table size: 512 203 Supported flow types: 204 ipv4-frag 205 ipv4-tcp 206 ipv4-udp 207 ipv4-sctp 208 ipv4-other 209 ipv6-frag 210 ipv6-tcp 211 ipv6-udp 212 ipv6-sctp 213 ipv6-other 214 l2_payload 215 port 216 vxlan 217 geneve 218 nvgre 219 220show port rss reta 221~~~~~~~~~~~~~~~~~~ 222 223Display the rss redirection table entry indicated by masks on port X:: 224 225 testpmd> show port (port_id) rss reta (size) (mask0, mask1...) 226 227size is used to indicate the hardware supported reta size 228 229show port rss-hash 230~~~~~~~~~~~~~~~~~~ 231 232Display the RSS hash functions and RSS hash key of a port:: 233 234 testpmd> show port (port_id) rss-hash ipv4|ipv4-frag|ipv4-tcp|ipv4-udp|ipv4-sctp|ipv4-other|ipv6|ipv6-frag|ipv6-tcp|ipv6-udp|ipv6-sctp|ipv6-other|l2-payload|ipv6-ex|ipv6-tcp-ex|ipv6-udp-ex [key] 235 236clear port 237~~~~~~~~~~ 238 239Clear the port statistics for a given port or for all ports:: 240 241 testpmd> clear port (info|stats|xstats|fdir|stat_qmap) (port_id|all) 242 243For example:: 244 245 testpmd> clear port stats all 246 247show (rxq|txq) 248~~~~~~~~~~~~~~ 249 250Display information for a given port's RX/TX queue:: 251 252 testpmd> show (rxq|txq) info (port_id) (queue_id) 253 254show config 255~~~~~~~~~~~ 256 257Displays the configuration of the application. 258The configuration comes from the command-line, the runtime or the application defaults:: 259 260 testpmd> show config (rxtx|cores|fwd|txpkts) 261 262The available information categories are: 263 264* ``rxtx``: RX/TX configuration items. 265 266* ``cores``: List of forwarding cores. 267 268* ``fwd``: Packet forwarding configuration. 269 270* ``txpkts``: Packets to TX configuration. 271 272For example: 273 274.. code-block:: console 275 276 testpmd> show config rxtx 277 278 io packet forwarding - CRC stripping disabled - packets/burst=16 279 nb forwarding cores=2 - nb forwarding ports=1 280 RX queues=1 - RX desc=128 - RX free threshold=0 281 RX threshold registers: pthresh=8 hthresh=8 wthresh=4 282 TX queues=1 - TX desc=512 - TX free threshold=0 283 TX threshold registers: pthresh=36 hthresh=0 wthresh=0 284 TX RS bit threshold=0 - TXQ flags=0x0 285 286set fwd 287~~~~~~~ 288 289Set the packet forwarding mode:: 290 291 testpmd> set fwd (io|mac|macswap|flowgen| \ 292 rxonly|txonly|csum|icmpecho) (""|retry) 293 294``retry`` can be specified for forwarding engines except ``rx_only``. 295 296The available information categories are: 297 298* ``io``: Forwards packets "as-is" in I/O mode. 299 This is the fastest possible forwarding operation as it does not access packets data. 300 This is the default mode. 301 302* ``mac``: Changes the source and the destination Ethernet addresses of packets before forwarding them. 303 Default application behaviour is to set source Ethernet address to that of the transmitting interface, and destination 304 address to a dummy value (set during init). The user may specify a target destination Ethernet address via the 'eth-peer' or 305 'eth-peer-configfile' command-line options. It is not currently possible to specify a specific source Ethernet address. 306 307* ``macswap``: MAC swap forwarding mode. 308 Swaps the source and the destination Ethernet addresses of packets before forwarding them. 309 310* ``flowgen``: Multi-flow generation mode. 311 Originates a number of flows (with varying destination IP addresses), and terminate receive traffic. 312 313* ``rxonly``: Receives packets but doesn't transmit them. 314 315* ``txonly``: Generates and transmits packets without receiving any. 316 317* ``csum``: Changes the checksum field with hardware or software methods depending on the offload flags on the packet. 318 319* ``icmpecho``: Receives a burst of packets, lookup for IMCP echo requests and, if any, send back ICMP echo replies. 320 321* ``ieee1588``: Demonstrate L2 IEEE1588 V2 PTP timestamping for RX and TX. Requires ``CONFIG_RTE_LIBRTE_IEEE1588=y``. 322 323* ``tm``: Traffic Management forwarding mode 324 Demonstrates the use of ethdev traffic management APIs and softnic PMD for 325 QoS traffic management. In this mode, 5-level hierarchical QoS scheduler is 326 available as an default option that can be enabled through CLI. The user can 327 also modify the default hierarchy or specify the new hierarchy through CLI for 328 implementing QoS scheduler. Requires ``CONFIG_RTE_LIBRTE_PMD_SOFTNIC=y`` ``CONFIG_RTE_LIBRTE_SCHED=y``. 329 330Example:: 331 332 testpmd> set fwd rxonly 333 334 Set rxonly packet forwarding mode 335 336 337read rxd 338~~~~~~~~ 339 340Display an RX descriptor for a port RX queue:: 341 342 testpmd> read rxd (port_id) (queue_id) (rxd_id) 343 344For example:: 345 346 testpmd> read rxd 0 0 4 347 0x0000000B - 0x001D0180 / 0x0000000B - 0x001D0180 348 349read txd 350~~~~~~~~ 351 352Display a TX descriptor for a port TX queue:: 353 354 testpmd> read txd (port_id) (queue_id) (txd_id) 355 356For example:: 357 358 testpmd> read txd 0 0 4 359 0x00000001 - 0x24C3C440 / 0x000F0000 - 0x2330003C 360 361ddp get list 362~~~~~~~~~~~~ 363 364Get loaded dynamic device personalization (DDP) package info list:: 365 366 testpmd> ddp get list (port_id) 367 368ddp get info 369~~~~~~~~~~~~ 370 371Display information about dynamic device personalization (DDP) profile:: 372 373 testpmd> ddp get info (profile_path) 374 375show vf stats 376~~~~~~~~~~~~~ 377 378Display VF statistics:: 379 380 testpmd> show vf stats (port_id) (vf_id) 381 382clear vf stats 383~~~~~~~~~~~~~~ 384 385Reset VF statistics:: 386 387 testpmd> clear vf stats (port_id) (vf_id) 388 389show port pctype mapping 390~~~~~~~~~~~~~~~~~~~~~~~~ 391 392List all items from the pctype mapping table:: 393 394 testpmd> show port (port_id) pctype mapping 395 396 397Configuration Functions 398----------------------- 399 400The testpmd application can be configured from the runtime as well as from the command-line. 401 402This section details the available configuration functions that are available. 403 404.. note:: 405 406 Configuration changes only become active when forwarding is started/restarted. 407 408set default 409~~~~~~~~~~~ 410 411Reset forwarding to the default configuration:: 412 413 testpmd> set default 414 415set verbose 416~~~~~~~~~~~ 417 418Set the debug verbosity level:: 419 420 testpmd> set verbose (level) 421 422Currently the only available levels are 0 (silent except for error) and 1 (fully verbose). 423 424set log 425~~~~~~~ 426 427Set the log level for a log type:: 428 429 testpmd> set log global|(type) (level) 430 431Where: 432 433* ``type`` is the log name. 434 435* ``level`` is the log level. 436 437For example, to change the global log level:: 438 testpmd> set log global (level) 439 440Regexes can also be used for type. To change log level of user1, user2 and user3:: 441 testpmd> set log user[1-3] (level) 442 443set nbport 444~~~~~~~~~~ 445 446Set the number of ports used by the application: 447 448set nbport (num) 449 450This is equivalent to the ``--nb-ports`` command-line option. 451 452set nbcore 453~~~~~~~~~~ 454 455Set the number of cores used by the application:: 456 457 testpmd> set nbcore (num) 458 459This is equivalent to the ``--nb-cores`` command-line option. 460 461.. note:: 462 463 The number of cores used must not be greater than number of ports used multiplied by the number of queues per port. 464 465set coremask 466~~~~~~~~~~~~ 467 468Set the forwarding cores hexadecimal mask:: 469 470 testpmd> set coremask (mask) 471 472This is equivalent to the ``--coremask`` command-line option. 473 474.. note:: 475 476 The master lcore is reserved for command line parsing only and cannot be masked on for packet forwarding. 477 478set portmask 479~~~~~~~~~~~~ 480 481Set the forwarding ports hexadecimal mask:: 482 483 testpmd> set portmask (mask) 484 485This is equivalent to the ``--portmask`` command-line option. 486 487set burst 488~~~~~~~~~ 489 490Set number of packets per burst:: 491 492 testpmd> set burst (num) 493 494This is equivalent to the ``--burst command-line`` option. 495 496When retry is enabled, the transmit delay time and number of retries can also be set:: 497 498 testpmd> set burst tx delay (microseconds) retry (num) 499 500set txpkts 501~~~~~~~~~~ 502 503Set the length of each segment of the TX-ONLY packets or length of packet for FLOWGEN mode:: 504 505 testpmd> set txpkts (x[,y]*) 506 507Where x[,y]* represents a CSV list of values, without white space. 508 509set txsplit 510~~~~~~~~~~~ 511 512Set the split policy for the TX packets, applicable for TX-ONLY and CSUM forwarding modes:: 513 514 testpmd> set txsplit (off|on|rand) 515 516Where: 517 518* ``off`` disable packet copy & split for CSUM mode. 519 520* ``on`` split outgoing packet into multiple segments. Size of each segment 521 and number of segments per packet is determined by ``set txpkts`` command 522 (see above). 523 524* ``rand`` same as 'on', but number of segments per each packet is a random value between 1 and total number of segments. 525 526set corelist 527~~~~~~~~~~~~ 528 529Set the list of forwarding cores:: 530 531 testpmd> set corelist (x[,y]*) 532 533For example, to change the forwarding cores: 534 535.. code-block:: console 536 537 testpmd> set corelist 3,1 538 testpmd> show config fwd 539 540 io packet forwarding - ports=2 - cores=2 - streams=2 - NUMA support disabled 541 Logical Core 3 (socket 0) forwards packets on 1 streams: 542 RX P=0/Q=0 (socket 0) -> TX P=1/Q=0 (socket 0) peer=02:00:00:00:00:01 543 Logical Core 1 (socket 0) forwards packets on 1 streams: 544 RX P=1/Q=0 (socket 0) -> TX P=0/Q=0 (socket 0) peer=02:00:00:00:00:00 545 546.. note:: 547 548 The cores are used in the same order as specified on the command line. 549 550set portlist 551~~~~~~~~~~~~ 552 553Set the list of forwarding ports:: 554 555 testpmd> set portlist (x[,y]*) 556 557For example, to change the port forwarding: 558 559.. code-block:: console 560 561 testpmd> set portlist 0,2,1,3 562 testpmd> show config fwd 563 564 io packet forwarding - ports=4 - cores=1 - streams=4 565 Logical Core 3 (socket 0) forwards packets on 4 streams: 566 RX P=0/Q=0 (socket 0) -> TX P=2/Q=0 (socket 0) peer=02:00:00:00:00:01 567 RX P=2/Q=0 (socket 0) -> TX P=0/Q=0 (socket 0) peer=02:00:00:00:00:00 568 RX P=1/Q=0 (socket 0) -> TX P=3/Q=0 (socket 0) peer=02:00:00:00:00:03 569 RX P=3/Q=0 (socket 0) -> TX P=1/Q=0 (socket 0) peer=02:00:00:00:00:02 570 571set tx loopback 572~~~~~~~~~~~~~~~ 573 574Enable/disable tx loopback:: 575 576 testpmd> set tx loopback (port_id) (on|off) 577 578set drop enable 579~~~~~~~~~~~~~~~ 580 581set drop enable bit for all queues:: 582 583 testpmd> set all queues drop (port_id) (on|off) 584 585set split drop enable (for VF) 586~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ 587 588set split drop enable bit for VF from PF:: 589 590 testpmd> set vf split drop (port_id) (vf_id) (on|off) 591 592set mac antispoof (for VF) 593~~~~~~~~~~~~~~~~~~~~~~~~~~ 594 595Set mac antispoof for a VF from the PF:: 596 597 testpmd> set vf mac antispoof (port_id) (vf_id) (on|off) 598 599set macsec offload 600~~~~~~~~~~~~~~~~~~ 601 602Enable/disable MACsec offload:: 603 604 testpmd> set macsec offload (port_id) on encrypt (on|off) replay-protect (on|off) 605 testpmd> set macsec offload (port_id) off 606 607set macsec sc 608~~~~~~~~~~~~~ 609 610Configure MACsec secure connection (SC):: 611 612 testpmd> set macsec sc (tx|rx) (port_id) (mac) (pi) 613 614.. note:: 615 616 The pi argument is ignored for tx. 617 Check the NIC Datasheet for hardware limits. 618 619set macsec sa 620~~~~~~~~~~~~~ 621 622Configure MACsec secure association (SA):: 623 624 testpmd> set macsec sa (tx|rx) (port_id) (idx) (an) (pn) (key) 625 626.. note:: 627 628 The IDX value must be 0 or 1. 629 Check the NIC Datasheet for hardware limits. 630 631set broadcast mode (for VF) 632~~~~~~~~~~~~~~~~~~~~~~~~~~~ 633 634Set broadcast mode for a VF from the PF:: 635 636 testpmd> set vf broadcast (port_id) (vf_id) (on|off) 637 638vlan set strip 639~~~~~~~~~~~~~~ 640 641Set the VLAN strip on a port:: 642 643 testpmd> vlan set strip (on|off) (port_id) 644 645vlan set stripq 646~~~~~~~~~~~~~~~ 647 648Set the VLAN strip for a queue on a port:: 649 650 testpmd> vlan set stripq (on|off) (port_id,queue_id) 651 652vlan set stripq (for VF) 653~~~~~~~~~~~~~~~~~~~~~~~~ 654 655Set VLAN strip for all queues in a pool for a VF from the PF:: 656 657 testpmd> set vf vlan stripq (port_id) (vf_id) (on|off) 658 659vlan set insert (for VF) 660~~~~~~~~~~~~~~~~~~~~~~~~ 661 662Set VLAN insert for a VF from the PF:: 663 664 testpmd> set vf vlan insert (port_id) (vf_id) (vlan_id) 665 666vlan set tag (for VF) 667~~~~~~~~~~~~~~~~~~~~~ 668 669Set VLAN tag for a VF from the PF:: 670 671 testpmd> set vf vlan tag (port_id) (vf_id) (on|off) 672 673vlan set antispoof (for VF) 674~~~~~~~~~~~~~~~~~~~~~~~~~~~ 675 676Set VLAN antispoof for a VF from the PF:: 677 678 testpmd> set vf vlan antispoof (port_id) (vf_id) (on|off) 679 680vlan set filter 681~~~~~~~~~~~~~~~ 682 683Set the VLAN filter on a port:: 684 685 testpmd> vlan set filter (on|off) (port_id) 686 687vlan set qinq 688~~~~~~~~~~~~~ 689 690Set the VLAN QinQ (extended queue in queue) on for a port:: 691 692 testpmd> vlan set qinq (on|off) (port_id) 693 694vlan set tpid 695~~~~~~~~~~~~~ 696 697Set the inner or outer VLAN TPID for packet filtering on a port:: 698 699 testpmd> vlan set (inner|outer) tpid (value) (port_id) 700 701.. note:: 702 703 TPID value must be a 16-bit number (value <= 65536). 704 705rx_vlan add 706~~~~~~~~~~~ 707 708Add a VLAN ID, or all identifiers, to the set of VLAN identifiers filtered by port ID:: 709 710 testpmd> rx_vlan add (vlan_id|all) (port_id) 711 712.. note:: 713 714 VLAN filter must be set on that port. VLAN ID < 4096. 715 Depending on the NIC used, number of vlan_ids may be limited to the maximum entries 716 in VFTA table. This is important if enabling all vlan_ids. 717 718rx_vlan rm 719~~~~~~~~~~ 720 721Remove a VLAN ID, or all identifiers, from the set of VLAN identifiers filtered by port ID:: 722 723 testpmd> rx_vlan rm (vlan_id|all) (port_id) 724 725rx_vlan add (for VF) 726~~~~~~~~~~~~~~~~~~~~ 727 728Add a VLAN ID, to the set of VLAN identifiers filtered for VF(s) for port ID:: 729 730 testpmd> rx_vlan add (vlan_id) port (port_id) vf (vf_mask) 731 732rx_vlan rm (for VF) 733~~~~~~~~~~~~~~~~~~~ 734 735Remove a VLAN ID, from the set of VLAN identifiers filtered for VF(s) for port ID:: 736 737 testpmd> rx_vlan rm (vlan_id) port (port_id) vf (vf_mask) 738 739tunnel_filter add 740~~~~~~~~~~~~~~~~~ 741 742Add a tunnel filter on a port:: 743 744 testpmd> tunnel_filter add (port_id) (outer_mac) (inner_mac) (ip_addr) \ 745 (inner_vlan) (vxlan|nvgre|ipingre) (imac-ivlan|imac-ivlan-tenid|\ 746 imac-tenid|imac|omac-imac-tenid|oip|iip) (tenant_id) (queue_id) 747 748The available information categories are: 749 750* ``vxlan``: Set tunnel type as VXLAN. 751 752* ``nvgre``: Set tunnel type as NVGRE. 753 754* ``ipingre``: Set tunnel type as IP-in-GRE. 755 756* ``imac-ivlan``: Set filter type as Inner MAC and VLAN. 757 758* ``imac-ivlan-tenid``: Set filter type as Inner MAC, VLAN and tenant ID. 759 760* ``imac-tenid``: Set filter type as Inner MAC and tenant ID. 761 762* ``imac``: Set filter type as Inner MAC. 763 764* ``omac-imac-tenid``: Set filter type as Outer MAC, Inner MAC and tenant ID. 765 766* ``oip``: Set filter type as Outer IP. 767 768* ``iip``: Set filter type as Inner IP. 769 770Example:: 771 772 testpmd> tunnel_filter add 0 68:05:CA:28:09:82 00:00:00:00:00:00 \ 773 192.168.2.2 0 ipingre oip 1 1 774 775 Set an IP-in-GRE tunnel on port 0, and the filter type is Outer IP. 776 777tunnel_filter remove 778~~~~~~~~~~~~~~~~~~~~ 779 780Remove a tunnel filter on a port:: 781 782 testpmd> tunnel_filter rm (port_id) (outer_mac) (inner_mac) (ip_addr) \ 783 (inner_vlan) (vxlan|nvgre|ipingre) (imac-ivlan|imac-ivlan-tenid|\ 784 imac-tenid|imac|omac-imac-tenid|oip|iip) (tenant_id) (queue_id) 785 786rx_vxlan_port add 787~~~~~~~~~~~~~~~~~ 788 789Add an UDP port for VXLAN packet filter on a port:: 790 791 testpmd> rx_vxlan_port add (udp_port) (port_id) 792 793rx_vxlan_port remove 794~~~~~~~~~~~~~~~~~~~~ 795 796Remove an UDP port for VXLAN packet filter on a port:: 797 798 testpmd> rx_vxlan_port rm (udp_port) (port_id) 799 800tx_vlan set 801~~~~~~~~~~~ 802 803Set hardware insertion of VLAN IDs in packets sent on a port:: 804 805 testpmd> tx_vlan set (port_id) vlan_id[, vlan_id_outer] 806 807For example, set a single VLAN ID (5) insertion on port 0:: 808 809 tx_vlan set 0 5 810 811Or, set double VLAN ID (inner: 2, outer: 3) insertion on port 1:: 812 813 tx_vlan set 1 2 3 814 815 816tx_vlan set pvid 817~~~~~~~~~~~~~~~~ 818 819Set port based hardware insertion of VLAN ID in packets sent on a port:: 820 821 testpmd> tx_vlan set pvid (port_id) (vlan_id) (on|off) 822 823tx_vlan reset 824~~~~~~~~~~~~~ 825 826Disable hardware insertion of a VLAN header in packets sent on a port:: 827 828 testpmd> tx_vlan reset (port_id) 829 830csum set 831~~~~~~~~ 832 833Select hardware or software calculation of the checksum when 834transmitting a packet using the ``csum`` forwarding engine:: 835 836 testpmd> csum set (ip|udp|tcp|sctp|outer-ip) (hw|sw) (port_id) 837 838Where: 839 840* ``ip|udp|tcp|sctp`` always relate to the inner layer. 841 842* ``outer-ip`` relates to the outer IP layer (only for IPv4) in the case where the packet is recognized 843 as a tunnel packet by the forwarding engine (vxlan, gre and ipip are 844 supported). See also the ``csum parse-tunnel`` command. 845 846.. note:: 847 848 Check the NIC Datasheet for hardware limits. 849 850RSS queue region 851~~~~~~~~~~~~~~~~ 852 853Set RSS queue region span on a port:: 854 855 testpmd> set port (port_id) queue-region region_id (value) \ 856 queue_start_index (value) queue_num (value) 857 858Set flowtype mapping on a RSS queue region on a port:: 859 860 testpmd> set port (port_id) queue-region region_id (value) flowtype (value) 861 862where: 863 864* For the flowtype(pctype) of packet,the specific index for each type has 865 been defined in file i40e_type.h as enum i40e_filter_pctype. 866 867Set user priority mapping on a RSS queue region on a port:: 868 869 testpmd> set port (port_id) queue-region UP (value) region_id (value) 870 871Flush all queue region related configuration on a port:: 872 873 testpmd> set port (port_id) queue-region flush (on|off) 874 875where: 876 877* "on"is just an enable function which server for other configuration, 878 it is for all configuration about queue region from up layer, 879 at first will only keep in DPDK softwarestored in driver, 880 only after "flush on", it commit all configuration to HW. 881 "off" is just clean all configuration about queue region just now, 882 and restore all to DPDK i40e driver default config when start up. 883 884Show all queue region related configuration info on a port:: 885 886 testpmd> show port (port_id) queue-region 887 888.. note:: 889 890 Queue region only support on PF by now, so these command is 891 only for configuration of queue region on PF port. 892 893csum parse-tunnel 894~~~~~~~~~~~~~~~~~ 895 896Define how tunneled packets should be handled by the csum forward 897engine:: 898 899 testpmd> csum parse-tunnel (on|off) (tx_port_id) 900 901If enabled, the csum forward engine will try to recognize supported 902tunnel headers (vxlan, gre, ipip). 903 904If disabled, treat tunnel packets as non-tunneled packets (a inner 905header is handled as a packet payload). 906 907.. note:: 908 909 The port argument is the TX port like in the ``csum set`` command. 910 911Example: 912 913Consider a packet in packet like the following:: 914 915 eth_out/ipv4_out/udp_out/vxlan/eth_in/ipv4_in/tcp_in 916 917* If parse-tunnel is enabled, the ``ip|udp|tcp|sctp`` parameters of ``csum set`` 918 command relate to the inner headers (here ``ipv4_in`` and ``tcp_in``), and the 919 ``outer-ip parameter`` relates to the outer headers (here ``ipv4_out``). 920 921* If parse-tunnel is disabled, the ``ip|udp|tcp|sctp`` parameters of ``csum set`` 922 command relate to the outer headers, here ``ipv4_out`` and ``udp_out``. 923 924csum show 925~~~~~~~~~ 926 927Display tx checksum offload configuration:: 928 929 testpmd> csum show (port_id) 930 931tso set 932~~~~~~~ 933 934Enable TCP Segmentation Offload (TSO) in the ``csum`` forwarding engine:: 935 936 testpmd> tso set (segsize) (port_id) 937 938.. note:: 939 940 Check the NIC datasheet for hardware limits. 941 942tso show 943~~~~~~~~ 944 945Display the status of TCP Segmentation Offload:: 946 947 testpmd> tso show (port_id) 948 949set port - gro 950~~~~~~~~~~~~~~ 951 952Enable or disable GRO in ``csum`` forwarding engine:: 953 954 testpmd> set port <port_id> gro on|off 955 956If enabled, the csum forwarding engine will perform GRO on the TCP/IPv4 957packets received from the given port. 958 959If disabled, packets received from the given port won't be performed 960GRO. By default, GRO is disabled for all ports. 961 962.. note:: 963 964 When enable GRO for a port, TCP/IPv4 packets received from the port 965 will be performed GRO. After GRO, all merged packets have bad 966 checksums, since the GRO library doesn't re-calculate checksums for 967 the merged packets. Therefore, if users want the merged packets to 968 have correct checksums, please select HW IP checksum calculation and 969 HW TCP checksum calculation for the port which the merged packets are 970 transmitted to. 971 972show port - gro 973~~~~~~~~~~~~~~~ 974 975Display GRO configuration for a given port:: 976 977 testpmd> show port <port_id> gro 978 979set gro flush 980~~~~~~~~~~~~~ 981 982Set the cycle to flush the GROed packets from reassembly tables:: 983 984 testpmd> set gro flush <cycles> 985 986When enable GRO, the csum forwarding engine performs GRO on received 987packets, and the GROed packets are stored in reassembly tables. Users 988can use this command to determine when the GROed packets are flushed 989from the reassembly tables. 990 991The ``cycles`` is measured in GRO operation times. The csum forwarding 992engine flushes the GROed packets from the tables every ``cycles`` GRO 993operations. 994 995By default, the value of ``cycles`` is 1, which means flush GROed packets 996from the reassembly tables as soon as one GRO operation finishes. The value 997of ``cycles`` should be in the range of 1 to ``GRO_MAX_FLUSH_CYCLES``. 998 999Please note that the large value of ``cycles`` may cause the poor TCP/IP 1000stack performance. Because the GROed packets are delayed to arrive the 1001stack, thus causing more duplicated ACKs and TCP retransmissions. 1002 1003set port - gso 1004~~~~~~~~~~~~~~ 1005 1006Toggle per-port GSO support in ``csum`` forwarding engine:: 1007 1008 testpmd> set port <port_id> gso on|off 1009 1010If enabled, the csum forwarding engine will perform GSO on supported IPv4 1011packets, transmitted on the given port. 1012 1013If disabled, packets transmitted on the given port will not undergo GSO. 1014By default, GSO is disabled for all ports. 1015 1016.. note:: 1017 1018 When GSO is enabled on a port, supported IPv4 packets transmitted on that 1019 port undergo GSO. Afterwards, the segmented packets are represented by 1020 multi-segment mbufs; however, the csum forwarding engine doesn't calculation 1021 of checksums for GSO'd segments in SW. As a result, if users want correct 1022 checksums in GSO segments, they should enable HW checksum calculation for 1023 GSO-enabled ports. 1024 1025 For example, HW checksum calculation for VxLAN GSO'd packets may be enabled 1026 by setting the following options in the csum forwarding engine: 1027 1028 testpmd> csum set outer_ip hw <port_id> 1029 1030 testpmd> csum set ip hw <port_id> 1031 1032 testpmd> csum set tcp hw <port_id> 1033 1034set gso segsz 1035~~~~~~~~~~~~~ 1036 1037Set the maximum GSO segment size (measured in bytes), which includes the 1038packet header and the packet payload for GSO-enabled ports (global):: 1039 1040 testpmd> set gso segsz <length> 1041 1042show port - gso 1043~~~~~~~~~~~~~~~ 1044 1045Display the status of Generic Segmentation Offload for a given port:: 1046 1047 testpmd> show port <port_id> gso 1048 1049mac_addr add 1050~~~~~~~~~~~~ 1051 1052Add an alternative MAC address to a port:: 1053 1054 testpmd> mac_addr add (port_id) (XX:XX:XX:XX:XX:XX) 1055 1056mac_addr remove 1057~~~~~~~~~~~~~~~ 1058 1059Remove a MAC address from a port:: 1060 1061 testpmd> mac_addr remove (port_id) (XX:XX:XX:XX:XX:XX) 1062 1063mac_addr add (for VF) 1064~~~~~~~~~~~~~~~~~~~~~ 1065 1066Add an alternative MAC address for a VF to a port:: 1067 1068 testpmd> mac_add add port (port_id) vf (vf_id) (XX:XX:XX:XX:XX:XX) 1069 1070mac_addr set 1071~~~~~~~~~~~~ 1072 1073Set the default MAC address for a port:: 1074 1075 testpmd> mac_addr set (port_id) (XX:XX:XX:XX:XX:XX) 1076 1077mac_addr set (for VF) 1078~~~~~~~~~~~~~~~~~~~~~ 1079 1080Set the MAC address for a VF from the PF:: 1081 1082 testpmd> set vf mac addr (port_id) (vf_id) (XX:XX:XX:XX:XX:XX) 1083 1084set eth-peer 1085~~~~~~~~~~~~ 1086 1087Set the forwarding peer address for certain port:: 1088 1089 testpmd> set eth-peer (port_id) (perr_addr) 1090 1091This is equivalent to the ``--eth-peer`` command-line option. 1092 1093set port-uta 1094~~~~~~~~~~~~ 1095 1096Set the unicast hash filter(s) on/off for a port:: 1097 1098 testpmd> set port (port_id) uta (XX:XX:XX:XX:XX:XX|all) (on|off) 1099 1100set promisc 1101~~~~~~~~~~~ 1102 1103Set the promiscuous mode on for a port or for all ports. 1104In promiscuous mode packets are not dropped if they aren't for the specified MAC address:: 1105 1106 testpmd> set promisc (port_id|all) (on|off) 1107 1108set allmulti 1109~~~~~~~~~~~~ 1110 1111Set the allmulti mode for a port or for all ports:: 1112 1113 testpmd> set allmulti (port_id|all) (on|off) 1114 1115Same as the ifconfig (8) option. Controls how multicast packets are handled. 1116 1117set promisc (for VF) 1118~~~~~~~~~~~~~~~~~~~~ 1119 1120Set the unicast promiscuous mode for a VF from PF. 1121It's supported by Intel i40e NICs now. 1122In promiscuous mode packets are not dropped if they aren't for the specified MAC address:: 1123 1124 testpmd> set vf promisc (port_id) (vf_id) (on|off) 1125 1126set allmulticast (for VF) 1127~~~~~~~~~~~~~~~~~~~~~~~~~ 1128 1129Set the multicast promiscuous mode for a VF from PF. 1130It's supported by Intel i40e NICs now. 1131In promiscuous mode packets are not dropped if they aren't for the specified MAC address:: 1132 1133 testpmd> set vf allmulti (port_id) (vf_id) (on|off) 1134 1135set tx max bandwidth (for VF) 1136~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ 1137 1138Set TX max absolute bandwidth (Mbps) for a VF from PF:: 1139 1140 testpmd> set vf tx max-bandwidth (port_id) (vf_id) (max_bandwidth) 1141 1142set tc tx min bandwidth (for VF) 1143~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ 1144 1145Set all TCs' TX min relative bandwidth (%) for a VF from PF:: 1146 1147 testpmd> set vf tc tx min-bandwidth (port_id) (vf_id) (bw1, bw2, ...) 1148 1149set tc tx max bandwidth (for VF) 1150~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ 1151 1152Set a TC's TX max absolute bandwidth (Mbps) for a VF from PF:: 1153 1154 testpmd> set vf tc tx max-bandwidth (port_id) (vf_id) (tc_no) (max_bandwidth) 1155 1156set tc strict link priority mode 1157~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ 1158 1159Set some TCs' strict link priority mode on a physical port:: 1160 1161 testpmd> set tx strict-link-priority (port_id) (tc_bitmap) 1162 1163set tc tx min bandwidth 1164~~~~~~~~~~~~~~~~~~~~~~~ 1165 1166Set all TCs' TX min relative bandwidth (%) globally for all PF and VFs:: 1167 1168 testpmd> set tc tx min-bandwidth (port_id) (bw1, bw2, ...) 1169 1170set flow_ctrl rx 1171~~~~~~~~~~~~~~~~ 1172 1173Set the link flow control parameter on a port:: 1174 1175 testpmd> set flow_ctrl rx (on|off) tx (on|off) (high_water) (low_water) \ 1176 (pause_time) (send_xon) mac_ctrl_frame_fwd (on|off) \ 1177 autoneg (on|off) (port_id) 1178 1179Where: 1180 1181* ``high_water`` (integer): High threshold value to trigger XOFF. 1182 1183* ``low_water`` (integer): Low threshold value to trigger XON. 1184 1185* ``pause_time`` (integer): Pause quota in the Pause frame. 1186 1187* ``send_xon`` (0/1): Send XON frame. 1188 1189* ``mac_ctrl_frame_fwd``: Enable receiving MAC control frames. 1190 1191* ``autoneg``: Change the auto-negotiation parameter. 1192 1193set pfc_ctrl rx 1194~~~~~~~~~~~~~~~ 1195 1196Set the priority flow control parameter on a port:: 1197 1198 testpmd> set pfc_ctrl rx (on|off) tx (on|off) (high_water) (low_water) \ 1199 (pause_time) (priority) (port_id) 1200 1201Where: 1202 1203* ``high_water`` (integer): High threshold value. 1204 1205* ``low_water`` (integer): Low threshold value. 1206 1207* ``pause_time`` (integer): Pause quota in the Pause frame. 1208 1209* ``priority`` (0-7): VLAN User Priority. 1210 1211set stat_qmap 1212~~~~~~~~~~~~~ 1213 1214Set statistics mapping (qmapping 0..15) for RX/TX queue on port:: 1215 1216 testpmd> set stat_qmap (tx|rx) (port_id) (queue_id) (qmapping) 1217 1218For example, to set rx queue 2 on port 0 to mapping 5:: 1219 1220 testpmd>set stat_qmap rx 0 2 5 1221 1222set xstats-hide-zero 1223~~~~~~~~~~~~~~~~~~~~ 1224 1225Set the option to hide zero values for xstats display:: 1226 1227 testpmd> set xstats-hide-zero on|off 1228 1229.. note:: 1230 1231 By default, the zero values are displayed for xstats. 1232 1233set port - rx/tx (for VF) 1234~~~~~~~~~~~~~~~~~~~~~~~~~ 1235 1236Set VF receive/transmit from a port:: 1237 1238 testpmd> set port (port_id) vf (vf_id) (rx|tx) (on|off) 1239 1240set port - mac address filter (for VF) 1241~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ 1242 1243Add/Remove unicast or multicast MAC addr filter for a VF:: 1244 1245 testpmd> set port (port_id) vf (vf_id) (mac_addr) \ 1246 (exact-mac|exact-mac-vlan|hashmac|hashmac-vlan) (on|off) 1247 1248set port - rx mode(for VF) 1249~~~~~~~~~~~~~~~~~~~~~~~~~~ 1250 1251Set the VF receive mode of a port:: 1252 1253 testpmd> set port (port_id) vf (vf_id) \ 1254 rxmode (AUPE|ROPE|BAM|MPE) (on|off) 1255 1256The available receive modes are: 1257 1258* ``AUPE``: Accepts untagged VLAN. 1259 1260* ``ROPE``: Accepts unicast hash. 1261 1262* ``BAM``: Accepts broadcast packets. 1263 1264* ``MPE``: Accepts all multicast packets. 1265 1266set port - tx_rate (for Queue) 1267~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ 1268 1269Set TX rate limitation for a queue on a port:: 1270 1271 testpmd> set port (port_id) queue (queue_id) rate (rate_value) 1272 1273set port - tx_rate (for VF) 1274~~~~~~~~~~~~~~~~~~~~~~~~~~~ 1275 1276Set TX rate limitation for queues in VF on a port:: 1277 1278 testpmd> set port (port_id) vf (vf_id) rate (rate_value) queue_mask (queue_mask) 1279 1280set port - mirror rule 1281~~~~~~~~~~~~~~~~~~~~~~ 1282 1283Set pool or vlan type mirror rule for a port:: 1284 1285 testpmd> set port (port_id) mirror-rule (rule_id) \ 1286 (pool-mirror-up|pool-mirror-down|vlan-mirror) \ 1287 (poolmask|vlanid[,vlanid]*) dst-pool (pool_id) (on|off) 1288 1289Set link mirror rule for a port:: 1290 1291 testpmd> set port (port_id) mirror-rule (rule_id) \ 1292 (uplink-mirror|downlink-mirror) dst-pool (pool_id) (on|off) 1293 1294For example to enable mirror traffic with vlan 0,1 to pool 0:: 1295 1296 set port 0 mirror-rule 0 vlan-mirror 0,1 dst-pool 0 on 1297 1298reset port - mirror rule 1299~~~~~~~~~~~~~~~~~~~~~~~~ 1300 1301Reset a mirror rule for a port:: 1302 1303 testpmd> reset port (port_id) mirror-rule (rule_id) 1304 1305set flush_rx 1306~~~~~~~~~~~~ 1307 1308Set the flush on RX streams before forwarding. 1309The default is flush ``on``. 1310Mainly used with PCAP drivers to turn off the default behavior of flushing the first 512 packets on RX streams:: 1311 1312 testpmd> set flush_rx off 1313 1314set bypass mode 1315~~~~~~~~~~~~~~~ 1316 1317Set the bypass mode for the lowest port on bypass enabled NIC:: 1318 1319 testpmd> set bypass mode (normal|bypass|isolate) (port_id) 1320 1321set bypass event 1322~~~~~~~~~~~~~~~~ 1323 1324Set the event required to initiate specified bypass mode for the lowest port on a bypass enabled:: 1325 1326 testpmd> set bypass event (timeout|os_on|os_off|power_on|power_off) \ 1327 mode (normal|bypass|isolate) (port_id) 1328 1329Where: 1330 1331* ``timeout``: Enable bypass after watchdog timeout. 1332 1333* ``os_on``: Enable bypass when OS/board is powered on. 1334 1335* ``os_off``: Enable bypass when OS/board is powered off. 1336 1337* ``power_on``: Enable bypass when power supply is turned on. 1338 1339* ``power_off``: Enable bypass when power supply is turned off. 1340 1341 1342set bypass timeout 1343~~~~~~~~~~~~~~~~~~ 1344 1345Set the bypass watchdog timeout to ``n`` seconds where 0 = instant:: 1346 1347 testpmd> set bypass timeout (0|1.5|2|3|4|8|16|32) 1348 1349show bypass config 1350~~~~~~~~~~~~~~~~~~ 1351 1352Show the bypass configuration for a bypass enabled NIC using the lowest port on the NIC:: 1353 1354 testpmd> show bypass config (port_id) 1355 1356set link up 1357~~~~~~~~~~~ 1358 1359Set link up for a port:: 1360 1361 testpmd> set link-up port (port id) 1362 1363set link down 1364~~~~~~~~~~~~~ 1365 1366Set link down for a port:: 1367 1368 testpmd> set link-down port (port id) 1369 1370E-tag set 1371~~~~~~~~~ 1372 1373Enable E-tag insertion for a VF on a port:: 1374 1375 testpmd> E-tag set insertion on port-tag-id (value) port (port_id) vf (vf_id) 1376 1377Disable E-tag insertion for a VF on a port:: 1378 1379 testpmd> E-tag set insertion off port (port_id) vf (vf_id) 1380 1381Enable/disable E-tag stripping on a port:: 1382 1383 testpmd> E-tag set stripping (on|off) port (port_id) 1384 1385Enable/disable E-tag based forwarding on a port:: 1386 1387 testpmd> E-tag set forwarding (on|off) port (port_id) 1388 1389Add an E-tag forwarding filter on a port:: 1390 1391 testpmd> E-tag set filter add e-tag-id (value) dst-pool (pool_id) port (port_id) 1392 1393Delete an E-tag forwarding filter on a port:: 1394 testpmd> E-tag set filter del e-tag-id (value) port (port_id) 1395 1396ddp add 1397~~~~~~~ 1398 1399Load a dynamic device personalization (DDP) profile and store backup profile:: 1400 1401 testpmd> ddp add (port_id) (profile_path[,backup_profile_path]) 1402 1403ddp del 1404~~~~~~~ 1405 1406Delete a dynamic device personalization profile and restore backup profile:: 1407 1408 testpmd> ddp del (port_id) (backup_profile_path) 1409 1410ptype mapping 1411~~~~~~~~~~~~~ 1412 1413List all items from the ptype mapping table:: 1414 1415 testpmd> ptype mapping get (port_id) (valid_only) 1416 1417Where: 1418 1419* ``valid_only``: A flag indicates if only list valid items(=1) or all itemss(=0). 1420 1421Replace a specific or a group of software defined ptype with a new one:: 1422 1423 testpmd> ptype mapping replace (port_id) (target) (mask) (pkt_type) 1424 1425where: 1426 1427* ``target``: A specific software ptype or a mask to represent a group of software ptypes. 1428 1429* ``mask``: A flag indicate if "target" is a specific software ptype(=0) or a ptype mask(=1). 1430 1431* ``pkt_type``: The new software ptype to replace the old ones. 1432 1433Update hardware defined ptype to software defined packet type mapping table:: 1434 1435 testpmd> ptype mapping update (port_id) (hw_ptype) (sw_ptype) 1436 1437where: 1438 1439* ``hw_ptype``: hardware ptype as the index of the ptype mapping table. 1440 1441* ``sw_ptype``: software ptype as the value of the ptype mapping table. 1442 1443Reset ptype mapping table:: 1444 1445 testpmd> ptype mapping reset (port_id) 1446 1447Port Functions 1448-------------- 1449 1450The following sections show functions for configuring ports. 1451 1452.. note:: 1453 1454 Port configuration changes only become active when forwarding is started/restarted. 1455 1456port attach 1457~~~~~~~~~~~ 1458 1459Attach a port specified by pci address or virtual device args:: 1460 1461 testpmd> port attach (identifier) 1462 1463To attach a new pci device, the device should be recognized by kernel first. 1464Then it should be moved under DPDK management. 1465Finally the port can be attached to testpmd. 1466 1467For example, to move a pci device using ixgbe under DPDK management: 1468 1469.. code-block:: console 1470 1471 # Check the status of the available devices. 1472 ./usertools/dpdk-devbind.py --status 1473 1474 Network devices using DPDK-compatible driver 1475 ============================================ 1476 <none> 1477 1478 Network devices using kernel driver 1479 =================================== 1480 0000:0a:00.0 '82599ES 10-Gigabit' if=eth2 drv=ixgbe unused= 1481 1482 1483 # Bind the device to igb_uio. 1484 sudo ./usertools/dpdk-devbind.py -b igb_uio 0000:0a:00.0 1485 1486 1487 # Recheck the status of the devices. 1488 ./usertools/dpdk-devbind.py --status 1489 Network devices using DPDK-compatible driver 1490 ============================================ 1491 0000:0a:00.0 '82599ES 10-Gigabit' drv=igb_uio unused= 1492 1493To attach a port created by virtual device, above steps are not needed. 1494 1495For example, to attach a port whose pci address is 0000:0a:00.0. 1496 1497.. code-block:: console 1498 1499 testpmd> port attach 0000:0a:00.0 1500 Attaching a new port... 1501 EAL: PCI device 0000:0a:00.0 on NUMA socket -1 1502 EAL: probe driver: 8086:10fb rte_ixgbe_pmd 1503 EAL: PCI memory mapped at 0x7f83bfa00000 1504 EAL: PCI memory mapped at 0x7f83bfa80000 1505 PMD: eth_ixgbe_dev_init(): MAC: 2, PHY: 18, SFP+: 5 1506 PMD: eth_ixgbe_dev_init(): port 0 vendorID=0x8086 deviceID=0x10fb 1507 Port 0 is attached. Now total ports is 1 1508 Done 1509 1510For example, to attach a port created by pcap PMD. 1511 1512.. code-block:: console 1513 1514 testpmd> port attach net_pcap0 1515 Attaching a new port... 1516 PMD: Initializing pmd_pcap for net_pcap0 1517 PMD: Creating pcap-backed ethdev on numa socket 0 1518 Port 0 is attached. Now total ports is 1 1519 Done 1520 1521In this case, identifier is ``net_pcap0``. 1522This identifier format is the same as ``--vdev`` format of DPDK applications. 1523 1524For example, to re-attach a bonded port which has been previously detached, 1525the mode and slave parameters must be given. 1526 1527.. code-block:: console 1528 1529 testpmd> port attach net_bond_0,mode=0,slave=1 1530 Attaching a new port... 1531 EAL: Initializing pmd_bond for net_bond_0 1532 EAL: Create bonded device net_bond_0 on port 0 in mode 0 on socket 0. 1533 Port 0 is attached. Now total ports is 1 1534 Done 1535 1536 1537port detach 1538~~~~~~~~~~~ 1539 1540Detach a specific port:: 1541 1542 testpmd> port detach (port_id) 1543 1544Before detaching a port, the port should be stopped and closed. 1545 1546For example, to detach a pci device port 0. 1547 1548.. code-block:: console 1549 1550 testpmd> port stop 0 1551 Stopping ports... 1552 Done 1553 testpmd> port close 0 1554 Closing ports... 1555 Done 1556 1557 testpmd> port detach 0 1558 Detaching a port... 1559 EAL: PCI device 0000:0a:00.0 on NUMA socket -1 1560 EAL: remove driver: 8086:10fb rte_ixgbe_pmd 1561 EAL: PCI memory unmapped at 0x7f83bfa00000 1562 EAL: PCI memory unmapped at 0x7f83bfa80000 1563 Done 1564 1565 1566For example, to detach a virtual device port 0. 1567 1568.. code-block:: console 1569 1570 testpmd> port stop 0 1571 Stopping ports... 1572 Done 1573 testpmd> port close 0 1574 Closing ports... 1575 Done 1576 1577 testpmd> port detach 0 1578 Detaching a port... 1579 PMD: Closing pcap ethdev on numa socket 0 1580 Port 'net_pcap0' is detached. Now total ports is 0 1581 Done 1582 1583To remove a pci device completely from the system, first detach the port from testpmd. 1584Then the device should be moved under kernel management. 1585Finally the device can be removed using kernel pci hotplug functionality. 1586 1587For example, to move a pci device under kernel management: 1588 1589.. code-block:: console 1590 1591 sudo ./usertools/dpdk-devbind.py -b ixgbe 0000:0a:00.0 1592 1593 ./usertools/dpdk-devbind.py --status 1594 1595 Network devices using DPDK-compatible driver 1596 ============================================ 1597 <none> 1598 1599 Network devices using kernel driver 1600 =================================== 1601 0000:0a:00.0 '82599ES 10-Gigabit' if=eth2 drv=ixgbe unused=igb_uio 1602 1603To remove a port created by a virtual device, above steps are not needed. 1604 1605port start 1606~~~~~~~~~~ 1607 1608Start all ports or a specific port:: 1609 1610 testpmd> port start (port_id|all) 1611 1612port stop 1613~~~~~~~~~ 1614 1615Stop all ports or a specific port:: 1616 1617 testpmd> port stop (port_id|all) 1618 1619port close 1620~~~~~~~~~~ 1621 1622Close all ports or a specific port:: 1623 1624 testpmd> port close (port_id|all) 1625 1626port start/stop queue 1627~~~~~~~~~~~~~~~~~~~~~ 1628 1629Start/stop a rx/tx queue on a specific port:: 1630 1631 testpmd> port (port_id) (rxq|txq) (queue_id) (start|stop) 1632 1633Only take effect when port is started. 1634 1635port config - speed 1636~~~~~~~~~~~~~~~~~~~ 1637 1638Set the speed and duplex mode for all ports or a specific port:: 1639 1640 testpmd> port config (port_id|all) speed (10|100|1000|10000|25000|40000|50000|100000|auto) \ 1641 duplex (half|full|auto) 1642 1643port config - queues/descriptors 1644~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ 1645 1646Set number of queues/descriptors for rxq, txq, rxd and txd:: 1647 1648 testpmd> port config all (rxq|txq|rxd|txd) (value) 1649 1650This is equivalent to the ``--rxq``, ``--txq``, ``--rxd`` and ``--txd`` command-line options. 1651 1652port config - max-pkt-len 1653~~~~~~~~~~~~~~~~~~~~~~~~~ 1654 1655Set the maximum packet length:: 1656 1657 testpmd> port config all max-pkt-len (value) 1658 1659This is equivalent to the ``--max-pkt-len`` command-line option. 1660 1661port config - CRC Strip 1662~~~~~~~~~~~~~~~~~~~~~~~ 1663 1664Set hardware CRC stripping on or off for all ports:: 1665 1666 testpmd> port config all crc-strip (on|off) 1667 1668CRC stripping is on by default. 1669 1670The ``off`` option is equivalent to the ``--disable-crc-strip`` command-line option. 1671 1672port config - scatter 1673~~~~~~~~~~~~~~~~~~~~~~~ 1674 1675Set RX scatter mode on or off for all ports:: 1676 1677 testpmd> port config all scatter (on|off) 1678 1679RX scatter mode is off by default. 1680 1681The ``on`` option is equivalent to the ``--enable-scatter`` command-line option. 1682 1683port config - RX Checksum 1684~~~~~~~~~~~~~~~~~~~~~~~~~ 1685 1686Set hardware RX checksum offload to on or off for all ports:: 1687 1688 testpmd> port config all rx-cksum (on|off) 1689 1690Checksum offload is off by default. 1691 1692The ``on`` option is equivalent to the ``--enable-rx-cksum`` command-line option. 1693 1694port config - VLAN 1695~~~~~~~~~~~~~~~~~~ 1696 1697Set hardware VLAN on or off for all ports:: 1698 1699 testpmd> port config all hw-vlan (on|off) 1700 1701Hardware VLAN is off by default. 1702 1703The ``on`` option is equivalent to the ``--enable-hw-vlan`` command-line option. 1704 1705port config - VLAN filter 1706~~~~~~~~~~~~~~~~~~~~~~~~~ 1707 1708Set hardware VLAN filter on or off for all ports:: 1709 1710 testpmd> port config all hw-vlan-filter (on|off) 1711 1712Hardware VLAN filter is off by default. 1713 1714The ``on`` option is equivalent to the ``--enable-hw-vlan-filter`` command-line option. 1715 1716port config - VLAN strip 1717~~~~~~~~~~~~~~~~~~~~~~~~ 1718 1719Set hardware VLAN strip on or off for all ports:: 1720 1721 testpmd> port config all hw-vlan-strip (on|off) 1722 1723Hardware VLAN strip is off by default. 1724 1725The ``on`` option is equivalent to the ``--enable-hw-vlan-strip`` command-line option. 1726 1727port config - VLAN extend 1728~~~~~~~~~~~~~~~~~~~~~~~~~ 1729 1730Set hardware VLAN extend on or off for all ports:: 1731 1732 testpmd> port config all hw-vlan-extend (on|off) 1733 1734Hardware VLAN extend is off by default. 1735 1736The ``on`` option is equivalent to the ``--enable-hw-vlan-extend`` command-line option. 1737 1738port config - Drop Packets 1739~~~~~~~~~~~~~~~~~~~~~~~~~~ 1740 1741Set packet drop for packets with no descriptors on or off for all ports:: 1742 1743 testpmd> port config all drop-en (on|off) 1744 1745Packet dropping for packets with no descriptors is off by default. 1746 1747The ``on`` option is equivalent to the ``--enable-drop-en`` command-line option. 1748 1749port config - RSS 1750~~~~~~~~~~~~~~~~~ 1751 1752Set the RSS (Receive Side Scaling) mode on or off:: 1753 1754 testpmd> port config all rss (all|ip|tcp|udp|sctp|ether|port|vxlan|geneve|nvgre|none) 1755 1756RSS is on by default. 1757 1758The ``none`` option is equivalent to the ``--disable-rss`` command-line option. 1759 1760port config - RSS Reta 1761~~~~~~~~~~~~~~~~~~~~~~ 1762 1763Set the RSS (Receive Side Scaling) redirection table:: 1764 1765 testpmd> port config all rss reta (hash,queue)[,(hash,queue)] 1766 1767port config - DCB 1768~~~~~~~~~~~~~~~~~ 1769 1770Set the DCB mode for an individual port:: 1771 1772 testpmd> port config (port_id) dcb vt (on|off) (traffic_class) pfc (on|off) 1773 1774The traffic class should be 4 or 8. 1775 1776port config - Burst 1777~~~~~~~~~~~~~~~~~~~ 1778 1779Set the number of packets per burst:: 1780 1781 testpmd> port config all burst (value) 1782 1783This is equivalent to the ``--burst`` command-line option. 1784 1785port config - Threshold 1786~~~~~~~~~~~~~~~~~~~~~~~ 1787 1788Set thresholds for TX/RX queues:: 1789 1790 testpmd> port config all (threshold) (value) 1791 1792Where the threshold type can be: 1793 1794* ``txpt:`` Set the prefetch threshold register of the TX rings, 0 <= value <= 255. 1795 1796* ``txht:`` Set the host threshold register of the TX rings, 0 <= value <= 255. 1797 1798* ``txwt:`` Set the write-back threshold register of the TX rings, 0 <= value <= 255. 1799 1800* ``rxpt:`` Set the prefetch threshold register of the RX rings, 0 <= value <= 255. 1801 1802* ``rxht:`` Set the host threshold register of the RX rings, 0 <= value <= 255. 1803 1804* ``rxwt:`` Set the write-back threshold register of the RX rings, 0 <= value <= 255. 1805 1806* ``txfreet:`` Set the transmit free threshold of the TX rings, 0 <= value <= txd. 1807 1808* ``rxfreet:`` Set the transmit free threshold of the RX rings, 0 <= value <= rxd. 1809 1810* ``txrst:`` Set the transmit RS bit threshold of TX rings, 0 <= value <= txd. 1811 1812These threshold options are also available from the command-line. 1813 1814port config - E-tag 1815~~~~~~~~~~~~~~~~~~~ 1816 1817Set the value of ether-type for E-tag:: 1818 1819 testpmd> port config (port_id|all) l2-tunnel E-tag ether-type (value) 1820 1821Enable/disable the E-tag support:: 1822 1823 testpmd> port config (port_id|all) l2-tunnel E-tag (enable|disable) 1824 1825port config pctype mapping 1826~~~~~~~~~~~~~~~~~~~~~~~~~~ 1827 1828Reset pctype mapping table:: 1829 1830 testpmd> port config (port_id) pctype mapping reset 1831 1832Update hardware defined pctype to software defined flow type mapping table:: 1833 1834 testpmd> port config (port_id) pctype mapping update (pctype_id_0[,pctype_id_1]*) (flow_type_id) 1835 1836where: 1837 1838* ``pctype_id_x``: hardware pctype id as index of bit in bitmask value of the pctype mapping table. 1839 1840* ``flow_type_id``: software flow type id as the index of the pctype mapping table. 1841 1842port config input set 1843~~~~~~~~~~~~~~~~~~~~~ 1844 1845Config RSS/FDIR/FDIR flexible payload input set for some pctype:: 1846 testpmd> port config (port_id) pctype (pctype_id) \ 1847 (hash_inset|fdir_inset|fdir_flx_inset) \ 1848 (get|set|clear) field (field_idx) 1849 1850Clear RSS/FDIR/FDIR flexible payload input set for some pctype:: 1851 testpmd> port config (port_id) pctype (pctype_id) \ 1852 (hash_inset|fdir_inset|fdir_flx_inset) clear all 1853 1854where: 1855 1856* ``pctype_id``: hardware packet classification types. 1857* ``field_idx``: hardware field index. 1858 1859Link Bonding Functions 1860---------------------- 1861 1862The Link Bonding functions make it possible to dynamically create and 1863manage link bonding devices from within testpmd interactive prompt. 1864 1865create bonded device 1866~~~~~~~~~~~~~~~~~~~~ 1867 1868Create a new bonding device:: 1869 1870 testpmd> create bonded device (mode) (socket) 1871 1872For example, to create a bonded device in mode 1 on socket 0:: 1873 1874 testpmd> create bonded 1 0 1875 created new bonded device (port X) 1876 1877add bonding slave 1878~~~~~~~~~~~~~~~~~ 1879 1880Adds Ethernet device to a Link Bonding device:: 1881 1882 testpmd> add bonding slave (slave id) (port id) 1883 1884For example, to add Ethernet device (port 6) to a Link Bonding device (port 10):: 1885 1886 testpmd> add bonding slave 6 10 1887 1888 1889remove bonding slave 1890~~~~~~~~~~~~~~~~~~~~ 1891 1892Removes an Ethernet slave device from a Link Bonding device:: 1893 1894 testpmd> remove bonding slave (slave id) (port id) 1895 1896For example, to remove Ethernet slave device (port 6) to a Link Bonding device (port 10):: 1897 1898 testpmd> remove bonding slave 6 10 1899 1900set bonding mode 1901~~~~~~~~~~~~~~~~ 1902 1903Set the Link Bonding mode of a Link Bonding device:: 1904 1905 testpmd> set bonding mode (value) (port id) 1906 1907For example, to set the bonding mode of a Link Bonding device (port 10) to broadcast (mode 3):: 1908 1909 testpmd> set bonding mode 3 10 1910 1911set bonding primary 1912~~~~~~~~~~~~~~~~~~~ 1913 1914Set an Ethernet slave device as the primary device on a Link Bonding device:: 1915 1916 testpmd> set bonding primary (slave id) (port id) 1917 1918For example, to set the Ethernet slave device (port 6) as the primary port of a Link Bonding device (port 10):: 1919 1920 testpmd> set bonding primary 6 10 1921 1922set bonding mac 1923~~~~~~~~~~~~~~~ 1924 1925Set the MAC address of a Link Bonding device:: 1926 1927 testpmd> set bonding mac (port id) (mac) 1928 1929For example, to set the MAC address of a Link Bonding device (port 10) to 00:00:00:00:00:01:: 1930 1931 testpmd> set bonding mac 10 00:00:00:00:00:01 1932 1933set bonding xmit_balance_policy 1934~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ 1935 1936Set the transmission policy for a Link Bonding device when it is in Balance XOR mode:: 1937 1938 testpmd> set bonding xmit_balance_policy (port_id) (l2|l23|l34) 1939 1940For example, set a Link Bonding device (port 10) to use a balance policy of layer 3+4 (IP addresses & UDP ports):: 1941 1942 testpmd> set bonding xmit_balance_policy 10 l34 1943 1944 1945set bonding mon_period 1946~~~~~~~~~~~~~~~~~~~~~~ 1947 1948Set the link status monitoring polling period in milliseconds for a bonding device. 1949 1950This adds support for PMD slave devices which do not support link status interrupts. 1951When the mon_period is set to a value greater than 0 then all PMD's which do not support 1952link status ISR will be queried every polling interval to check if their link status has changed:: 1953 1954 testpmd> set bonding mon_period (port_id) (value) 1955 1956For example, to set the link status monitoring polling period of bonded device (port 5) to 150ms:: 1957 1958 testpmd> set bonding mon_period 5 150 1959 1960 1961set bonding lacp dedicated_queue 1962~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ 1963 1964Enable dedicated tx/rx queues on bonding devices slaves to handle LACP control plane traffic 1965when in mode 4 (link-aggregration-802.3ad):: 1966 1967 testpmd> set bonding lacp dedicated_queues (port_id) (enable|disable) 1968 1969 1970set bonding agg_mode 1971~~~~~~~~~~~~~~~~~~~~ 1972 1973Enable one of the specific aggregators mode when in mode 4 (link-aggregration-802.3ad):: 1974 1975 testpmd> set bonding agg_mode (port_id) (bandwidth|count|stable) 1976 1977 1978show bonding config 1979~~~~~~~~~~~~~~~~~~~ 1980 1981Show the current configuration of a Link Bonding device:: 1982 1983 testpmd> show bonding config (port id) 1984 1985For example, 1986to show the configuration a Link Bonding device (port 9) with 3 slave devices (1, 3, 4) 1987in balance mode with a transmission policy of layer 2+3:: 1988 1989 testpmd> show bonding config 9 1990 Bonding mode: 2 1991 Balance Xmit Policy: BALANCE_XMIT_POLICY_LAYER23 1992 Slaves (3): [1 3 4] 1993 Active Slaves (3): [1 3 4] 1994 Primary: [3] 1995 1996 1997Register Functions 1998------------------ 1999 2000The Register Functions can be used to read from and write to registers on the network card referenced by a port number. 2001This is mainly useful for debugging purposes. 2002Reference should be made to the appropriate datasheet for the network card for details on the register addresses 2003and fields that can be accessed. 2004 2005read reg 2006~~~~~~~~ 2007 2008Display the value of a port register:: 2009 2010 testpmd> read reg (port_id) (address) 2011 2012For example, to examine the Flow Director control register (FDIRCTL, 0x0000EE000) on an Intel 82599 10 GbE Controller:: 2013 2014 testpmd> read reg 0 0xEE00 2015 port 0 PCI register at offset 0xEE00: 0x4A060029 (1241907241) 2016 2017read regfield 2018~~~~~~~~~~~~~ 2019 2020Display a port register bit field:: 2021 2022 testpmd> read regfield (port_id) (address) (bit_x) (bit_y) 2023 2024For example, reading the lowest two bits from the register in the example above:: 2025 2026 testpmd> read regfield 0 0xEE00 0 1 2027 port 0 PCI register at offset 0xEE00: bits[0, 1]=0x1 (1) 2028 2029read regbit 2030~~~~~~~~~~~ 2031 2032Display a single port register bit:: 2033 2034 testpmd> read regbit (port_id) (address) (bit_x) 2035 2036For example, reading the lowest bit from the register in the example above:: 2037 2038 testpmd> read regbit 0 0xEE00 0 2039 port 0 PCI register at offset 0xEE00: bit 0=1 2040 2041write reg 2042~~~~~~~~~ 2043 2044Set the value of a port register:: 2045 2046 testpmd> write reg (port_id) (address) (value) 2047 2048For example, to clear a register:: 2049 2050 testpmd> write reg 0 0xEE00 0x0 2051 port 0 PCI register at offset 0xEE00: 0x00000000 (0) 2052 2053write regfield 2054~~~~~~~~~~~~~~ 2055 2056Set bit field of a port register:: 2057 2058 testpmd> write regfield (port_id) (address) (bit_x) (bit_y) (value) 2059 2060For example, writing to the register cleared in the example above:: 2061 2062 testpmd> write regfield 0 0xEE00 0 1 2 2063 port 0 PCI register at offset 0xEE00: 0x00000002 (2) 2064 2065write regbit 2066~~~~~~~~~~~~ 2067 2068Set single bit value of a port register:: 2069 2070 testpmd> write regbit (port_id) (address) (bit_x) (value) 2071 2072For example, to set the high bit in the register from the example above:: 2073 2074 testpmd> write regbit 0 0xEE00 31 1 2075 port 0 PCI register at offset 0xEE00: 0x8000000A (2147483658) 2076 2077Traffic Metering and Policing 2078----------------------------- 2079 2080The following section shows functions for configuring traffic metering and 2081policing on the ethernet device through the use of generic ethdev API. 2082 2083show port traffic management capability 2084~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ 2085 2086Show traffic metering and policing capability of the port:: 2087 2088 testpmd> show port meter cap (port_id) 2089 2090add port meter profile (srTCM rfc2967) 2091~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ 2092 2093Add meter profile (srTCM rfc2697) to the ethernet device:: 2094 2095 testpmd> add port meter profile srtcm_rfc2697 (port_id) (profile_id) \ 2096 (cir) (cbs) (ebs) 2097 2098where: 2099 2100* ``profile_id``: ID for the meter profile. 2101* ``cir``: Committed Information Rate (CIR) (bytes/second). 2102* ``cbs``: Committed Burst Size (CBS) (bytes). 2103* ``ebs``: Excess Burst Size (EBS) (bytes). 2104 2105add port meter profile (trTCM rfc2968) 2106~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ 2107 2108Add meter profile (srTCM rfc2698) to the ethernet device:: 2109 2110 testpmd> add port meter profile trtcm_rfc2698 (port_id) (profile_id) \ 2111 (cir) (pir) (cbs) (pbs) 2112 2113where: 2114 2115* ``profile_id``: ID for the meter profile. 2116* ``cir``: Committed information rate (bytes/second). 2117* ``pir``: Peak information rate (bytes/second). 2118* ``cbs``: Committed burst size (bytes). 2119* ``pbs``: Peak burst size (bytes). 2120 2121add port meter profile (trTCM rfc4115) 2122~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ 2123 2124Add meter profile (trTCM rfc4115) to the ethernet device:: 2125 2126 testpmd> add port meter profile trtcm_rfc4115 (port_id) (profile_id) \ 2127 (cir) (eir) (cbs) (ebs) 2128 2129where: 2130 2131* ``profile_id``: ID for the meter profile. 2132* ``cir``: Committed information rate (bytes/second). 2133* ``eir``: Excess information rate (bytes/second). 2134* ``cbs``: Committed burst size (bytes). 2135* ``ebs``: Excess burst size (bytes). 2136 2137delete port meter profile 2138~~~~~~~~~~~~~~~~~~~~~~~~~ 2139 2140Delete meter profile from the ethernet device:: 2141 2142 testpmd> del port meter profile (port_id) (profile_id) 2143 2144create port meter 2145~~~~~~~~~~~~~~~~~ 2146 2147Create new meter object for the ethernet device:: 2148 2149 testpmd> create port meter (port_id) (mtr_id) (profile_id) \ 2150 (meter_enable) (g_action) (y_action) (r_action) (stats_mask) (shared) \ 2151 (use_pre_meter_color) [(dscp_tbl_entry0) (dscp_tbl_entry1)...\ 2152 (dscp_tbl_entry63)] 2153 2154where: 2155 2156* ``mtr_id``: meter object ID. 2157* ``profile_id``: ID for the meter profile. 2158* ``meter_enable``: When this parameter has a non-zero value, the meter object 2159 gets enabled at the time of creation, otherwise remains disabled. 2160* ``g_action``: Policer action for the packet with green color. 2161* ``y_action``: Policer action for the packet with yellow color. 2162* ``r_action``: Policer action for the packet with red color. 2163* ``stats_mask``: Mask of statistics counter types to be enabled for the 2164 meter object. 2165* ``shared``: When this parameter has a non-zero value, the meter object is 2166 shared by multiple flows. Otherwise, meter object is used by single flow. 2167* ``use_pre_meter_color``: When this parameter has a non-zero value, the 2168 input color for the current meter object is determined by the latest meter 2169 object in the same flow. Otherwise, the current meter object uses the 2170 *dscp_table* to determine the input color. 2171* ``dscp_tbl_entryx``: DSCP table entry x providing meter providing input 2172 color, 0 <= x <= 63. 2173 2174enable port meter 2175~~~~~~~~~~~~~~~~~ 2176 2177Enable meter for the ethernet device:: 2178 2179 testpmd> enable port meter (port_id) (mtr_id) 2180 2181disable port meter 2182~~~~~~~~~~~~~~~~~~ 2183 2184Disable meter for the ethernet device:: 2185 2186 testpmd> disable port meter (port_id) (mtr_id) 2187 2188delete port meter 2189~~~~~~~~~~~~~~~~~ 2190 2191Delete meter for the ethernet device:: 2192 2193 testpmd> del port meter (port_id) (mtr_id) 2194 2195Set port meter profile 2196~~~~~~~~~~~~~~~~~~~~~~ 2197 2198Set meter profile for the ethernet device:: 2199 2200 testpmd> set port meter profile (port_id) (mtr_id) (profile_id) 2201 2202set port meter dscp table 2203~~~~~~~~~~~~~~~~~~~~~~~~~ 2204 2205Set meter dscp table for the ethernet device:: 2206 2207 testpmd> set port meter dscp table (port_id) (mtr_id) [(dscp_tbl_entry0) \ 2208 (dscp_tbl_entry1)...(dscp_tbl_entry63)] 2209 2210set port meter policer action 2211~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ 2212 2213Set meter policer action for the ethernet device:: 2214 2215 testpmd> set port meter policer action (port_id) (mtr_id) (action_mask) \ 2216 (action0) [(action1) (action1)] 2217 2218where: 2219 2220* ``action_mask``: Bit mask indicating which policer actions need to be 2221 updated. One or more policer actions can be updated in a single function 2222 invocation. To update the policer action associated with color C, bit 2223 (1 << C) needs to be set in *action_mask* and element at position C 2224 in the *actions* array needs to be valid. 2225* ``actionx``: Policer action for the color x, 2226 RTE_MTR_GREEN <= x < RTE_MTR_COLORS 2227 2228set port meter stats mask 2229~~~~~~~~~~~~~~~~~~~~~~~~~ 2230 2231Set meter stats mask for the ethernet device:: 2232 2233 testpmd> set port meter stats mask (port_id) (mtr_id) (stats_mask) 2234 2235where: 2236 2237* ``stats_mask``: Bit mask indicating statistics counter types to be enabled. 2238 2239show port meter stats 2240~~~~~~~~~~~~~~~~~~~~~ 2241 2242Show meter stats of the ethernet device:: 2243 2244 testpmd> show port meter stats (port_id) (mtr_id) (clear) 2245 2246where: 2247 2248* ``clear``: Flag that indicates whether the statistics counters should 2249 be cleared (i.e. set to zero) immediately after they have been read or not. 2250 2251Traffic Management 2252------------------ 2253 2254The following section shows functions for configuring traffic management on 2255on the ethernet device through the use of generic TM API. 2256 2257show port traffic management capability 2258~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ 2259 2260Show traffic management capability of the port:: 2261 2262 testpmd> show port tm cap (port_id) 2263 2264show port traffic management capability (hierarchy level) 2265~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ 2266 2267Show traffic management hierarchy level capability of the port:: 2268 2269 testpmd> show port tm level cap (port_id) (level_id) 2270 2271show port traffic management capability (hierarchy node level) 2272~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ 2273 2274Show the traffic management hierarchy node capability of the port:: 2275 2276 testpmd> show port tm node cap (port_id) (node_id) 2277 2278show port traffic management hierarchy node type 2279~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ 2280 2281Show the port traffic management hierarchy node type:: 2282 2283 testpmd> show port tm node type (port_id) (node_id) 2284 2285show port traffic management hierarchy node stats 2286~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ 2287 2288Show the port traffic management hierarchy node statistics:: 2289 2290 testpmd> show port tm node stats (port_id) (node_id) (clear) 2291 2292where: 2293 2294* ``clear``: When this parameter has a non-zero value, the statistics counters 2295 are cleared (i.e. set to zero) immediately after they have been read, 2296 otherwise the statistics counters are left untouched. 2297 2298Add port traffic management private shaper profile 2299~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ 2300 2301Add the port traffic management private shaper profile:: 2302 2303 testpmd> add port tm node shaper profile (port_id) (shaper_profile_id) \ 2304 (tb_rate) (tb_size) (packet_length_adjust) 2305 2306where: 2307 2308* ``shaper_profile id``: Shaper profile ID for the new profile. 2309* ``tb_rate``: Token bucket rate (bytes per second). 2310* ``tb_size``: Token bucket size (bytes). 2311* ``packet_length_adjust``: The value (bytes) to be added to the length of 2312 each packet for the purpose of shaping. This parameter value can be used to 2313 correct the packet length with the framing overhead bytes that are consumed 2314 on the wire. 2315 2316Delete port traffic management private shaper profile 2317~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ 2318 2319Delete the port traffic management private shaper:: 2320 2321 testpmd> del port tm node shaper profile (port_id) (shaper_profile_id) 2322 2323where: 2324 2325* ``shaper_profile id``: Shaper profile ID that needs to be deleted. 2326 2327Add port traffic management shared shaper 2328~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ 2329 2330Create the port traffic management shared shaper:: 2331 2332 testpmd> add port tm node shared shaper (port_id) (shared_shaper_id) \ 2333 (shaper_profile_id) 2334 2335where: 2336 2337* ``shared_shaper_id``: Shared shaper ID to be created. 2338* ``shaper_profile id``: Shaper profile ID for shared shaper. 2339 2340Set port traffic management shared shaper 2341~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ 2342 2343Update the port traffic management shared shaper:: 2344 2345 testpmd> set port tm node shared shaper (port_id) (shared_shaper_id) \ 2346 (shaper_profile_id) 2347 2348where: 2349 2350* ``shared_shaper_id``: Shared shaper ID to be update. 2351* ``shaper_profile id``: Shaper profile ID for shared shaper. 2352 2353Delete port traffic management shared shaper 2354~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ 2355 2356Delete the port traffic management shared shaper:: 2357 2358 testpmd> del port tm node shared shaper (port_id) (shared_shaper_id) 2359 2360where: 2361 2362* ``shared_shaper_id``: Shared shaper ID to be deleted. 2363 2364Set port traffic management hiearchy node private shaper 2365~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ 2366 2367set the port traffic management hierarchy node private shaper:: 2368 2369 testpmd> set port tm node shaper profile (port_id) (node_id) \ 2370 (shaper_profile_id) 2371 2372where: 2373 2374* ``shaper_profile id``: Private shaper profile ID to be enabled on the 2375 hierarchy node. 2376 2377Add port traffic management WRED profile 2378~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ 2379 2380Create a new WRED profile:: 2381 2382 testpmd> add port tm node wred profile (port_id) (wred_profile_id) \ 2383 (color_g) (min_th_g) (max_th_g) (maxp_inv_g) (wq_log2_g) \ 2384 (color_y) (min_th_y) (max_th_y) (maxp_inv_y) (wq_log2_y) \ 2385 (color_r) (min_th_r) (max_th_r) (maxp_inv_r) (wq_log2_r) 2386 2387where: 2388 2389* ``wred_profile id``: Identifier for the newly create WRED profile 2390* ``color_g``: Packet color (green) 2391* ``min_th_g``: Minimum queue threshold for packet with green color 2392* ``max_th_g``: Minimum queue threshold for packet with green color 2393* ``maxp_inv_g``: Inverse of packet marking probability maximum value (maxp) 2394* ``wq_log2_g``: Negated log2 of queue weight (wq) 2395* ``color_y``: Packet color (yellow) 2396* ``min_th_y``: Minimum queue threshold for packet with yellow color 2397* ``max_th_y``: Minimum queue threshold for packet with yellow color 2398* ``maxp_inv_y``: Inverse of packet marking probability maximum value (maxp) 2399* ``wq_log2_y``: Negated log2 of queue weight (wq) 2400* ``color_r``: Packet color (red) 2401* ``min_th_r``: Minimum queue threshold for packet with yellow color 2402* ``max_th_r``: Minimum queue threshold for packet with yellow color 2403* ``maxp_inv_r``: Inverse of packet marking probability maximum value (maxp) 2404* ``wq_log2_r``: Negated log2 of queue weight (wq) 2405 2406Delete port traffic management WRED profile 2407~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ 2408 2409Delete the WRED profile:: 2410 2411 testpmd> del port tm node wred profile (port_id) (wred_profile_id) 2412 2413Add port traffic management hierarchy nonleaf node 2414~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ 2415 2416Add nonleaf node to port traffic management hiearchy:: 2417 2418 testpmd> add port tm nonleaf node (port_id) (node_id) (parent_node_id) \ 2419 (priority) (weight) (level_id) (shaper_profile_id) \ 2420 (n_sp_priorities) (stats_mask) (n_shared_shapers) \ 2421 [(shared_shaper_0) (shared_shaper_1) ...] \ 2422 2423where: 2424 2425* ``parent_node_id``: Node ID of the parent. 2426* ``priority``: Node priority (highest node priority is zero). This is used by 2427 the SP algorithm running on the parent node for scheduling this node. 2428* ``weight``: Node weight (lowest weight is one). The node weight is relative 2429 to the weight sum of all siblings that have the same priority. It is used by 2430 the WFQ algorithm running on the parent node for scheduling this node. 2431* ``level_id``: Hiearchy level of the node. 2432* ``shaper_profile_id``: Shaper profile ID of the private shaper to be used by 2433 the node. 2434* ``n_sp_priorities``: Number of strict priorities. 2435* ``stats_mask``: Mask of statistics counter types to be enabled for this node. 2436* ``n_shared_shapers``: Number of shared shapers. 2437* ``shared_shaper_id``: Shared shaper id. 2438 2439Add port traffic management hierarchy leaf node 2440~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ 2441 2442Add leaf node to port traffic management hiearchy:: 2443 2444 testpmd> add port tm leaf node (port_id) (node_id) (parent_node_id) \ 2445 (priority) (weight) (level_id) (shaper_profile_id) \ 2446 (cman_mode) (wred_profile_id) (stats_mask) (n_shared_shapers) \ 2447 [(shared_shaper_id) (shared_shaper_id) ...] \ 2448 2449where: 2450 2451* ``parent_node_id``: Node ID of the parent. 2452* ``priority``: Node priority (highest node priority is zero). This is used by 2453 the SP algorithm running on the parent node for scheduling this node. 2454* ``weight``: Node weight (lowest weight is one). The node weight is relative 2455 to the weight sum of all siblings that have the same priority. It is used by 2456 the WFQ algorithm running on the parent node for scheduling this node. 2457* ``level_id``: Hiearchy level of the node. 2458* ``shaper_profile_id``: Shaper profile ID of the private shaper to be used by 2459 the node. 2460* ``cman_mode``: Congestion management mode to be enabled for this node. 2461* ``wred_profile_id``: WRED profile id to be enabled for this node. 2462* ``stats_mask``: Mask of statistics counter types to be enabled for this node. 2463* ``n_shared_shapers``: Number of shared shapers. 2464* ``shared_shaper_id``: Shared shaper id. 2465 2466Delete port traffic management hierarchy node 2467~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ 2468 2469Delete node from port traffic management hiearchy:: 2470 2471 testpmd> del port tm node (port_id) (node_id) 2472 2473Update port traffic management hierarchy parent node 2474~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ 2475 2476Update port traffic management hierarchy parent node:: 2477 2478 testpmd> set port tm node parent (port_id) (node_id) (parent_node_id) \ 2479 (priority) (weight) 2480 2481This function can only be called after the hierarchy commit invocation. Its 2482success depends on the port support for this operation, as advertised through 2483the port capability set. This function is valid for all nodes of the traffic 2484management hierarchy except root node. 2485 2486Commit port traffic management hierarchy 2487~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ 2488 2489Commit the traffic management hierarchy on the port:: 2490 2491 testpmd> port tm hierarchy commit (port_id) (clean_on_fail) 2492 2493where: 2494 2495* ``clean_on_fail``: When set to non-zero, hierarchy is cleared on function 2496 call failure. On the other hand, hierarchy is preserved when this parameter 2497 is equal to zero. 2498 2499Set port traffic management default hierarchy (tm forwarding mode) 2500~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ 2501 2502set the traffic management default hierarchy on the port:: 2503 2504 testpmd> set port tm hierarchy default (port_id) 2505 2506Filter Functions 2507---------------- 2508 2509This section details the available filter functions that are available. 2510 2511Note these functions interface the deprecated legacy filtering framework, 2512superseded by *rte_flow*. See `Flow rules management`_. 2513 2514ethertype_filter 2515~~~~~~~~~~~~~~~~~~~~ 2516 2517Add or delete a L2 Ethertype filter, which identify packets by their L2 Ethertype mainly assign them to a receive queue:: 2518 2519 ethertype_filter (port_id) (add|del) (mac_addr|mac_ignr) (mac_address) \ 2520 ethertype (ether_type) (drop|fwd) queue (queue_id) 2521 2522The available information parameters are: 2523 2524* ``port_id``: The port which the Ethertype filter assigned on. 2525 2526* ``mac_addr``: Compare destination mac address. 2527 2528* ``mac_ignr``: Ignore destination mac address match. 2529 2530* ``mac_address``: Destination mac address to match. 2531 2532* ``ether_type``: The EtherType value want to match, 2533 for example 0x0806 for ARP packet. 0x0800 (IPv4) and 0x86DD (IPv6) are invalid. 2534 2535* ``queue_id``: The receive queue associated with this EtherType filter. 2536 It is meaningless when deleting or dropping. 2537 2538Example, to add/remove an ethertype filter rule:: 2539 2540 testpmd> ethertype_filter 0 add mac_ignr 00:11:22:33:44:55 \ 2541 ethertype 0x0806 fwd queue 3 2542 2543 testpmd> ethertype_filter 0 del mac_ignr 00:11:22:33:44:55 \ 2544 ethertype 0x0806 fwd queue 3 2545 25462tuple_filter 2547~~~~~~~~~~~~~~~~~ 2548 2549Add or delete a 2-tuple filter, 2550which identifies packets by specific protocol and destination TCP/UDP port 2551and forwards packets into one of the receive queues:: 2552 2553 2tuple_filter (port_id) (add|del) dst_port (dst_port_value) \ 2554 protocol (protocol_value) mask (mask_value) \ 2555 tcp_flags (tcp_flags_value) priority (prio_value) \ 2556 queue (queue_id) 2557 2558The available information parameters are: 2559 2560* ``port_id``: The port which the 2-tuple filter assigned on. 2561 2562* ``dst_port_value``: Destination port in L4. 2563 2564* ``protocol_value``: IP L4 protocol. 2565 2566* ``mask_value``: Participates in the match or not by bit for field above, 1b means participate. 2567 2568* ``tcp_flags_value``: TCP control bits. The non-zero value is invalid, when the pro_value is not set to 0x06 (TCP). 2569 2570* ``prio_value``: Priority of this filter. 2571 2572* ``queue_id``: The receive queue associated with this 2-tuple filter. 2573 2574Example, to add/remove an 2tuple filter rule:: 2575 2576 testpmd> 2tuple_filter 0 add dst_port 32 protocol 0x06 mask 0x03 \ 2577 tcp_flags 0x02 priority 3 queue 3 2578 2579 testpmd> 2tuple_filter 0 del dst_port 32 protocol 0x06 mask 0x03 \ 2580 tcp_flags 0x02 priority 3 queue 3 2581 25825tuple_filter 2583~~~~~~~~~~~~~~~~~ 2584 2585Add or delete a 5-tuple filter, 2586which consists of a 5-tuple (protocol, source and destination IP addresses, source and destination TCP/UDP/SCTP port) 2587and routes packets into one of the receive queues:: 2588 2589 5tuple_filter (port_id) (add|del) dst_ip (dst_address) src_ip \ 2590 (src_address) dst_port (dst_port_value) \ 2591 src_port (src_port_value) protocol (protocol_value) \ 2592 mask (mask_value) tcp_flags (tcp_flags_value) \ 2593 priority (prio_value) queue (queue_id) 2594 2595The available information parameters are: 2596 2597* ``port_id``: The port which the 5-tuple filter assigned on. 2598 2599* ``dst_address``: Destination IP address. 2600 2601* ``src_address``: Source IP address. 2602 2603* ``dst_port_value``: TCP/UDP destination port. 2604 2605* ``src_port_value``: TCP/UDP source port. 2606 2607* ``protocol_value``: L4 protocol. 2608 2609* ``mask_value``: Participates in the match or not by bit for field above, 1b means participate 2610 2611* ``tcp_flags_value``: TCP control bits. The non-zero value is invalid, when the protocol_value is not set to 0x06 (TCP). 2612 2613* ``prio_value``: The priority of this filter. 2614 2615* ``queue_id``: The receive queue associated with this 5-tuple filter. 2616 2617Example, to add/remove an 5tuple filter rule:: 2618 2619 testpmd> 5tuple_filter 0 add dst_ip 2.2.2.5 src_ip 2.2.2.4 \ 2620 dst_port 64 src_port 32 protocol 0x06 mask 0x1F \ 2621 flags 0x0 priority 3 queue 3 2622 2623 testpmd> 5tuple_filter 0 del dst_ip 2.2.2.5 src_ip 2.2.2.4 \ 2624 dst_port 64 src_port 32 protocol 0x06 mask 0x1F \ 2625 flags 0x0 priority 3 queue 3 2626 2627syn_filter 2628~~~~~~~~~~ 2629 2630Using the SYN filter, TCP packets whose *SYN* flag is set can be forwarded to a separate queue:: 2631 2632 syn_filter (port_id) (add|del) priority (high|low) queue (queue_id) 2633 2634The available information parameters are: 2635 2636* ``port_id``: The port which the SYN filter assigned on. 2637 2638* ``high``: This SYN filter has higher priority than other filters. 2639 2640* ``low``: This SYN filter has lower priority than other filters. 2641 2642* ``queue_id``: The receive queue associated with this SYN filter 2643 2644Example:: 2645 2646 testpmd> syn_filter 0 add priority high queue 3 2647 2648flex_filter 2649~~~~~~~~~~~ 2650 2651With flex filter, packets can be recognized by any arbitrary pattern within the first 128 bytes of the packet 2652and routed into one of the receive queues:: 2653 2654 flex_filter (port_id) (add|del) len (len_value) bytes (bytes_value) \ 2655 mask (mask_value) priority (prio_value) queue (queue_id) 2656 2657The available information parameters are: 2658 2659* ``port_id``: The port which the Flex filter is assigned on. 2660 2661* ``len_value``: Filter length in bytes, no greater than 128. 2662 2663* ``bytes_value``: A string in hexadecimal, means the value the flex filter needs to match. 2664 2665* ``mask_value``: A string in hexadecimal, bit 1 means corresponding byte participates in the match. 2666 2667* ``prio_value``: The priority of this filter. 2668 2669* ``queue_id``: The receive queue associated with this Flex filter. 2670 2671Example:: 2672 2673 testpmd> flex_filter 0 add len 16 bytes 0x00000000000000000000000008060000 \ 2674 mask 000C priority 3 queue 3 2675 2676 testpmd> flex_filter 0 del len 16 bytes 0x00000000000000000000000008060000 \ 2677 mask 000C priority 3 queue 3 2678 2679 2680.. _testpmd_flow_director: 2681 2682flow_director_filter 2683~~~~~~~~~~~~~~~~~~~~ 2684 2685The Flow Director works in receive mode to identify specific flows or sets of flows and route them to specific queues. 2686 2687Four types of filtering are supported which are referred to as Perfect Match, Signature, Perfect-mac-vlan and 2688Perfect-tunnel filters, the match mode is set by the ``--pkt-filter-mode`` command-line parameter: 2689 2690* Perfect match filters. 2691 The hardware checks a match between the masked fields of the received packets and the programmed filters. 2692 The masked fields are for IP flow. 2693 2694* Signature filters. 2695 The hardware checks a match between a hash-based signature of the masked fields of the received packet. 2696 2697* Perfect-mac-vlan match filters. 2698 The hardware checks a match between the masked fields of the received packets and the programmed filters. 2699 The masked fields are for MAC VLAN flow. 2700 2701* Perfect-tunnel match filters. 2702 The hardware checks a match between the masked fields of the received packets and the programmed filters. 2703 The masked fields are for tunnel flow. 2704 2705* Perfect-raw-flow-type match filters. 2706 The hardware checks a match between the masked fields of the received packets and pre-loaded raw (template) packet. 2707 The masked fields are specified by input sets. 2708 2709The Flow Director filters can match the different fields for different type of packet: flow type, specific input set 2710per flow type and the flexible payload. 2711 2712The Flow Director can also mask out parts of all of these fields so that filters 2713are only applied to certain fields or parts of the fields. 2714 2715Note that for raw flow type mode the source and destination fields in the 2716raw packet buffer need to be presented in a reversed order with respect 2717to the expected received packets. 2718For example: IP source and destination addresses or TCP/UDP/SCTP 2719source and destination ports 2720 2721Different NICs may have different capabilities, command show port fdir (port_id) can be used to acquire the information. 2722 2723# Commands to add flow director filters of different flow types:: 2724 2725 flow_director_filter (port_id) mode IP (add|del|update) \ 2726 flow (ipv4-other|ipv4-frag|ipv6-other|ipv6-frag) \ 2727 src (src_ip_address) dst (dst_ip_address) \ 2728 tos (tos_value) proto (proto_value) ttl (ttl_value) \ 2729 vlan (vlan_value) flexbytes (flexbytes_value) \ 2730 (drop|fwd) pf|vf(vf_id) queue (queue_id) \ 2731 fd_id (fd_id_value) 2732 2733 flow_director_filter (port_id) mode IP (add|del|update) \ 2734 flow (ipv4-tcp|ipv4-udp|ipv6-tcp|ipv6-udp) \ 2735 src (src_ip_address) (src_port) \ 2736 dst (dst_ip_address) (dst_port) \ 2737 tos (tos_value) ttl (ttl_value) \ 2738 vlan (vlan_value) flexbytes (flexbytes_value) \ 2739 (drop|fwd) queue pf|vf(vf_id) (queue_id) \ 2740 fd_id (fd_id_value) 2741 2742 flow_director_filter (port_id) mode IP (add|del|update) \ 2743 flow (ipv4-sctp|ipv6-sctp) \ 2744 src (src_ip_address) (src_port) \ 2745 dst (dst_ip_address) (dst_port) \ 2746 tos (tos_value) ttl (ttl_value) \ 2747 tag (verification_tag) vlan (vlan_value) \ 2748 flexbytes (flexbytes_value) (drop|fwd) \ 2749 pf|vf(vf_id) queue (queue_id) fd_id (fd_id_value) 2750 2751 flow_director_filter (port_id) mode IP (add|del|update) flow l2_payload \ 2752 ether (ethertype) flexbytes (flexbytes_value) \ 2753 (drop|fwd) pf|vf(vf_id) queue (queue_id) 2754 fd_id (fd_id_value) 2755 2756 flow_director_filter (port_id) mode MAC-VLAN (add|del|update) \ 2757 mac (mac_address) vlan (vlan_value) \ 2758 flexbytes (flexbytes_value) (drop|fwd) \ 2759 queue (queue_id) fd_id (fd_id_value) 2760 2761 flow_director_filter (port_id) mode Tunnel (add|del|update) \ 2762 mac (mac_address) vlan (vlan_value) \ 2763 tunnel (NVGRE|VxLAN) tunnel-id (tunnel_id_value) \ 2764 flexbytes (flexbytes_value) (drop|fwd) \ 2765 queue (queue_id) fd_id (fd_id_value) 2766 2767 flow_director_filter (port_id) mode raw (add|del|update) flow (flow_id) \ 2768 (drop|fwd) queue (queue_id) fd_id (fd_id_value) \ 2769 packet (packet file name) 2770 2771For example, to add an ipv4-udp flow type filter:: 2772 2773 testpmd> flow_director_filter 0 mode IP add flow ipv4-udp src 2.2.2.3 32 \ 2774 dst 2.2.2.5 33 tos 2 ttl 40 vlan 0x1 flexbytes (0x88,0x48) \ 2775 fwd pf queue 1 fd_id 1 2776 2777For example, add an ipv4-other flow type filter:: 2778 2779 testpmd> flow_director_filter 0 mode IP add flow ipv4-other src 2.2.2.3 \ 2780 dst 2.2.2.5 tos 2 proto 20 ttl 40 vlan 0x1 \ 2781 flexbytes (0x88,0x48) fwd pf queue 1 fd_id 1 2782 2783flush_flow_director 2784~~~~~~~~~~~~~~~~~~~ 2785 2786Flush all flow director filters on a device:: 2787 2788 testpmd> flush_flow_director (port_id) 2789 2790Example, to flush all flow director filter on port 0:: 2791 2792 testpmd> flush_flow_director 0 2793 2794flow_director_mask 2795~~~~~~~~~~~~~~~~~~ 2796 2797Set flow director's input masks:: 2798 2799 flow_director_mask (port_id) mode IP vlan (vlan_value) \ 2800 src_mask (ipv4_src) (ipv6_src) (src_port) \ 2801 dst_mask (ipv4_dst) (ipv6_dst) (dst_port) 2802 2803 flow_director_mask (port_id) mode MAC-VLAN vlan (vlan_value) 2804 2805 flow_director_mask (port_id) mode Tunnel vlan (vlan_value) \ 2806 mac (mac_value) tunnel-type (tunnel_type_value) \ 2807 tunnel-id (tunnel_id_value) 2808 2809Example, to set flow director mask on port 0:: 2810 2811 testpmd> flow_director_mask 0 mode IP vlan 0xefff \ 2812 src_mask 255.255.255.255 \ 2813 FFFF:FFFF:FFFF:FFFF:FFFF:FFFF:FFFF:FFFF 0xFFFF \ 2814 dst_mask 255.255.255.255 \ 2815 FFFF:FFFF:FFFF:FFFF:FFFF:FFFF:FFFF:FFFF 0xFFFF 2816 2817flow_director_flex_mask 2818~~~~~~~~~~~~~~~~~~~~~~~ 2819 2820set masks of flow director's flexible payload based on certain flow type:: 2821 2822 testpmd> flow_director_flex_mask (port_id) \ 2823 flow (none|ipv4-other|ipv4-frag|ipv4-tcp|ipv4-udp|ipv4-sctp| \ 2824 ipv6-other|ipv6-frag|ipv6-tcp|ipv6-udp|ipv6-sctp| \ 2825 l2_payload|all) (mask) 2826 2827Example, to set flow director's flex mask for all flow type on port 0:: 2828 2829 testpmd> flow_director_flex_mask 0 flow all \ 2830 (0xff,0xff,0,0,0,0,0,0,0,0,0,0,0,0,0,0) 2831 2832 2833flow_director_flex_payload 2834~~~~~~~~~~~~~~~~~~~~~~~~~~ 2835 2836Configure flexible payload selection:: 2837 2838 flow_director_flex_payload (port_id) (raw|l2|l3|l4) (config) 2839 2840For example, to select the first 16 bytes from the offset 4 (bytes) of packet's payload as flexible payload:: 2841 2842 testpmd> flow_director_flex_payload 0 l4 \ 2843 (4,5,6,7,8,9,10,11,12,13,14,15,16,17,18,19) 2844 2845get_sym_hash_ena_per_port 2846~~~~~~~~~~~~~~~~~~~~~~~~~ 2847 2848Get symmetric hash enable configuration per port:: 2849 2850 get_sym_hash_ena_per_port (port_id) 2851 2852For example, to get symmetric hash enable configuration of port 1:: 2853 2854 testpmd> get_sym_hash_ena_per_port 1 2855 2856set_sym_hash_ena_per_port 2857~~~~~~~~~~~~~~~~~~~~~~~~~ 2858 2859Set symmetric hash enable configuration per port to enable or disable:: 2860 2861 set_sym_hash_ena_per_port (port_id) (enable|disable) 2862 2863For example, to set symmetric hash enable configuration of port 1 to enable:: 2864 2865 testpmd> set_sym_hash_ena_per_port 1 enable 2866 2867get_hash_global_config 2868~~~~~~~~~~~~~~~~~~~~~~ 2869 2870Get the global configurations of hash filters:: 2871 2872 get_hash_global_config (port_id) 2873 2874For example, to get the global configurations of hash filters of port 1:: 2875 2876 testpmd> get_hash_global_config 1 2877 2878set_hash_global_config 2879~~~~~~~~~~~~~~~~~~~~~~ 2880 2881Set the global configurations of hash filters:: 2882 2883 set_hash_global_config (port_id) (toeplitz|simple_xor|default) \ 2884 (ipv4|ipv4-frag|ipv4-tcp|ipv4-udp|ipv4-sctp|ipv4-other|ipv6|ipv6-frag| \ 2885 ipv6-tcp|ipv6-udp|ipv6-sctp|ipv6-other|l2_payload|<flow_id>) \ 2886 (enable|disable) 2887 2888For example, to enable simple_xor for flow type of ipv6 on port 2:: 2889 2890 testpmd> set_hash_global_config 2 simple_xor ipv6 enable 2891 2892set_hash_input_set 2893~~~~~~~~~~~~~~~~~~ 2894 2895Set the input set for hash:: 2896 2897 set_hash_input_set (port_id) (ipv4-frag|ipv4-tcp|ipv4-udp|ipv4-sctp| \ 2898 ipv4-other|ipv6-frag|ipv6-tcp|ipv6-udp|ipv6-sctp|ipv6-other| \ 2899 l2_payload|<flow_id>) (ovlan|ivlan|src-ipv4|dst-ipv4|src-ipv6|dst-ipv6| \ 2900 ipv4-tos|ipv4-proto|ipv6-tc|ipv6-next-header|udp-src-port|udp-dst-port| \ 2901 tcp-src-port|tcp-dst-port|sctp-src-port|sctp-dst-port|sctp-veri-tag| \ 2902 udp-key|gre-key|fld-1st|fld-2nd|fld-3rd|fld-4th|fld-5th|fld-6th|fld-7th| \ 2903 fld-8th|none) (select|add) 2904 2905For example, to add source IP to hash input set for flow type of ipv4-udp on port 0:: 2906 2907 testpmd> set_hash_input_set 0 ipv4-udp src-ipv4 add 2908 2909set_fdir_input_set 2910~~~~~~~~~~~~~~~~~~ 2911 2912The Flow Director filters can match the different fields for different type of packet, i.e. specific input set 2913on per flow type and the flexible payload. This command can be used to change input set for each flow type. 2914 2915Set the input set for flow director:: 2916 2917 set_fdir_input_set (port_id) (ipv4-frag|ipv4-tcp|ipv4-udp|ipv4-sctp| \ 2918 ipv4-other|ipv6|ipv6-frag|ipv6-tcp|ipv6-udp|ipv6-sctp|ipv6-other| \ 2919 l2_payload|<flow_id>) (ivlan|ethertype|src-ipv4|dst-ipv4|src-ipv6|dst-ipv6| \ 2920 ipv4-tos|ipv4-proto|ipv4-ttl|ipv6-tc|ipv6-next-header|ipv6-hop-limits| \ 2921 tudp-src-port|udp-dst-port|cp-src-port|tcp-dst-port|sctp-src-port| \ 2922 sctp-dst-port|sctp-veri-tag|none) (select|add) 2923 2924For example to add source IP to FD input set for flow type of ipv4-udp on port 0:: 2925 2926 testpmd> set_fdir_input_set 0 ipv4-udp src-ipv4 add 2927 2928global_config 2929~~~~~~~~~~~~~ 2930 2931Set different GRE key length for input set:: 2932 2933 global_config (port_id) gre-key-len (number in bytes) 2934 2935For example to set GRE key length for input set to 4 bytes on port 0:: 2936 2937 testpmd> global_config 0 gre-key-len 4 2938 2939 2940.. _testpmd_rte_flow: 2941 2942Flow rules management 2943--------------------- 2944 2945Control of the generic flow API (*rte_flow*) is fully exposed through the 2946``flow`` command (validation, creation, destruction, queries and operation 2947modes). 2948 2949Considering *rte_flow* overlaps with all `Filter Functions`_, using both 2950features simultaneously may cause undefined side-effects and is therefore 2951not recommended. 2952 2953``flow`` syntax 2954~~~~~~~~~~~~~~~ 2955 2956Because the ``flow`` command uses dynamic tokens to handle the large number 2957of possible flow rules combinations, its behavior differs slightly from 2958other commands, in particular: 2959 2960- Pressing *?* or the *<tab>* key displays contextual help for the current 2961 token, not that of the entire command. 2962 2963- Optional and repeated parameters are supported (provided they are listed 2964 in the contextual help). 2965 2966The first parameter stands for the operation mode. Possible operations and 2967their general syntax are described below. They are covered in detail in the 2968following sections. 2969 2970- Check whether a flow rule can be created:: 2971 2972 flow validate {port_id} 2973 [group {group_id}] [priority {level}] [ingress] [egress] 2974 pattern {item} [/ {item} [...]] / end 2975 actions {action} [/ {action} [...]] / end 2976 2977- Create a flow rule:: 2978 2979 flow create {port_id} 2980 [group {group_id}] [priority {level}] [ingress] [egress] 2981 pattern {item} [/ {item} [...]] / end 2982 actions {action} [/ {action} [...]] / end 2983 2984- Destroy specific flow rules:: 2985 2986 flow destroy {port_id} rule {rule_id} [...] 2987 2988- Destroy all flow rules:: 2989 2990 flow flush {port_id} 2991 2992- Query an existing flow rule:: 2993 2994 flow query {port_id} {rule_id} {action} 2995 2996- List existing flow rules sorted by priority, filtered by group 2997 identifiers:: 2998 2999 flow list {port_id} [group {group_id}] [...] 3000 3001- Restrict ingress traffic to the defined flow rules:: 3002 3003 flow isolate {port_id} {boolean} 3004 3005Validating flow rules 3006~~~~~~~~~~~~~~~~~~~~~ 3007 3008``flow validate`` reports whether a flow rule would be accepted by the 3009underlying device in its current state but stops short of creating it. It is 3010bound to ``rte_flow_validate()``:: 3011 3012 flow validate {port_id} 3013 [group {group_id}] [priority {level}] [ingress] [egress] 3014 pattern {item} [/ {item} [...]] / end 3015 actions {action} [/ {action} [...]] / end 3016 3017If successful, it will show:: 3018 3019 Flow rule validated 3020 3021Otherwise it will show an error message of the form:: 3022 3023 Caught error type [...] ([...]): [...] 3024 3025This command uses the same parameters as ``flow create``, their format is 3026described in `Creating flow rules`_. 3027 3028Check whether redirecting any Ethernet packet received on port 0 to RX queue 3029index 6 is supported:: 3030 3031 testpmd> flow validate 0 ingress pattern eth / end 3032 actions queue index 6 / end 3033 Flow rule validated 3034 testpmd> 3035 3036Port 0 does not support TCPv6 rules:: 3037 3038 testpmd> flow validate 0 ingress pattern eth / ipv6 / tcp / end 3039 actions drop / end 3040 Caught error type 9 (specific pattern item): Invalid argument 3041 testpmd> 3042 3043Creating flow rules 3044~~~~~~~~~~~~~~~~~~~ 3045 3046``flow create`` validates and creates the specified flow rule. It is bound 3047to ``rte_flow_create()``:: 3048 3049 flow create {port_id} 3050 [group {group_id}] [priority {level}] [ingress] [egress] 3051 pattern {item} [/ {item} [...]] / end 3052 actions {action} [/ {action} [...]] / end 3053 3054If successful, it will return a flow rule ID usable with other commands:: 3055 3056 Flow rule #[...] created 3057 3058Otherwise it will show an error message of the form:: 3059 3060 Caught error type [...] ([...]): [...] 3061 3062Parameters describe in the following order: 3063 3064- Attributes (*group*, *priority*, *ingress*, *egress* tokens). 3065- A matching pattern, starting with the *pattern* token and terminated by an 3066 *end* pattern item. 3067- Actions, starting with the *actions* token and terminated by an *end* 3068 action. 3069 3070These translate directly to *rte_flow* objects provided as-is to the 3071underlying functions. 3072 3073The shortest valid definition only comprises mandatory tokens:: 3074 3075 testpmd> flow create 0 pattern end actions end 3076 3077Note that PMDs may refuse rules that essentially do nothing such as this 3078one. 3079 3080**All unspecified object values are automatically initialized to 0.** 3081 3082Attributes 3083^^^^^^^^^^ 3084 3085These tokens affect flow rule attributes (``struct rte_flow_attr``) and are 3086specified before the ``pattern`` token. 3087 3088- ``group {group id}``: priority group. 3089- ``priority {level}``: priority level within group. 3090- ``ingress``: rule applies to ingress traffic. 3091- ``egress``: rule applies to egress traffic. 3092 3093Each instance of an attribute specified several times overrides the previous 3094value as shown below (group 4 is used):: 3095 3096 testpmd> flow create 0 group 42 group 24 group 4 [...] 3097 3098Note that once enabled, ``ingress`` and ``egress`` cannot be disabled. 3099 3100While not specifying a direction is an error, some rules may allow both 3101simultaneously. 3102 3103Most rules affect RX therefore contain the ``ingress`` token:: 3104 3105 testpmd> flow create 0 ingress pattern [...] 3106 3107Matching pattern 3108^^^^^^^^^^^^^^^^ 3109 3110A matching pattern starts after the ``pattern`` token. It is made of pattern 3111items and is terminated by a mandatory ``end`` item. 3112 3113Items are named after their type (*RTE_FLOW_ITEM_TYPE_* from ``enum 3114rte_flow_item_type``). 3115 3116The ``/`` token is used as a separator between pattern items as shown 3117below:: 3118 3119 testpmd> flow create 0 ingress pattern eth / ipv4 / udp / end [...] 3120 3121Note that protocol items like these must be stacked from lowest to highest 3122layer to make sense. For instance, the following rule is either invalid or 3123unlikely to match any packet:: 3124 3125 testpmd> flow create 0 ingress pattern eth / udp / ipv4 / end [...] 3126 3127More information on these restrictions can be found in the *rte_flow* 3128documentation. 3129 3130Several items support additional specification structures, for example 3131``ipv4`` allows specifying source and destination addresses as follows:: 3132 3133 testpmd> flow create 0 ingress pattern eth / ipv4 src is 10.1.1.1 3134 dst is 10.2.0.0 / end [...] 3135 3136This rule matches all IPv4 traffic with the specified properties. 3137 3138In this example, ``src`` and ``dst`` are field names of the underlying 3139``struct rte_flow_item_ipv4`` object. All item properties can be specified 3140in a similar fashion. 3141 3142The ``is`` token means that the subsequent value must be matched exactly, 3143and assigns ``spec`` and ``mask`` fields in ``struct rte_flow_item`` 3144accordingly. Possible assignment tokens are: 3145 3146- ``is``: match value perfectly (with full bit-mask). 3147- ``spec``: match value according to configured bit-mask. 3148- ``last``: specify upper bound to establish a range. 3149- ``mask``: specify bit-mask with relevant bits set to one. 3150- ``prefix``: generate bit-mask from a prefix length. 3151 3152These yield identical results:: 3153 3154 ipv4 src is 10.1.1.1 3155 3156:: 3157 3158 ipv4 src spec 10.1.1.1 src mask 255.255.255.255 3159 3160:: 3161 3162 ipv4 src spec 10.1.1.1 src prefix 32 3163 3164:: 3165 3166 ipv4 src is 10.1.1.1 src last 10.1.1.1 # range with a single value 3167 3168:: 3169 3170 ipv4 src is 10.1.1.1 src last 0 # 0 disables range 3171 3172Inclusive ranges can be defined with ``last``:: 3173 3174 ipv4 src is 10.1.1.1 src last 10.2.3.4 # 10.1.1.1 to 10.2.3.4 3175 3176Note that ``mask`` affects both ``spec`` and ``last``:: 3177 3178 ipv4 src is 10.1.1.1 src last 10.2.3.4 src mask 255.255.0.0 3179 # matches 10.1.0.0 to 10.2.255.255 3180 3181Properties can be modified multiple times:: 3182 3183 ipv4 src is 10.1.1.1 src is 10.1.2.3 src is 10.2.3.4 # matches 10.2.3.4 3184 3185:: 3186 3187 ipv4 src is 10.1.1.1 src prefix 24 src prefix 16 # matches 10.1.0.0/16 3188 3189Pattern items 3190^^^^^^^^^^^^^ 3191 3192This section lists supported pattern items and their attributes, if any. 3193 3194- ``end``: end list of pattern items. 3195 3196- ``void``: no-op pattern item. 3197 3198- ``invert``: perform actions when pattern does not match. 3199 3200- ``any``: match any protocol for the current layer. 3201 3202 - ``num {unsigned}``: number of layers covered. 3203 3204- ``pf``: match packets addressed to the physical function. 3205 3206- ``vf``: match packets addressed to a virtual function ID. 3207 3208 - ``id {unsigned}``: destination VF ID. 3209 3210- ``port``: device-specific physical port index to use. 3211 3212 - ``index {unsigned}``: physical port index. 3213 3214- ``raw``: match an arbitrary byte string. 3215 3216 - ``relative {boolean}``: look for pattern after the previous item. 3217 - ``search {boolean}``: search pattern from offset (see also limit). 3218 - ``offset {integer}``: absolute or relative offset for pattern. 3219 - ``limit {unsigned}``: search area limit for start of pattern. 3220 - ``pattern {string}``: byte string to look for. 3221 3222- ``eth``: match Ethernet header. 3223 3224 - ``dst {MAC-48}``: destination MAC. 3225 - ``src {MAC-48}``: source MAC. 3226 - ``type {unsigned}``: EtherType. 3227 3228- ``vlan``: match 802.1Q/ad VLAN tag. 3229 3230 - ``tpid {unsigned}``: tag protocol identifier. 3231 - ``tci {unsigned}``: tag control information. 3232 - ``pcp {unsigned}``: priority code point. 3233 - ``dei {unsigned}``: drop eligible indicator. 3234 - ``vid {unsigned}``: VLAN identifier. 3235 3236- ``ipv4``: match IPv4 header. 3237 3238 - ``tos {unsigned}``: type of service. 3239 - ``ttl {unsigned}``: time to live. 3240 - ``proto {unsigned}``: next protocol ID. 3241 - ``src {ipv4 address}``: source address. 3242 - ``dst {ipv4 address}``: destination address. 3243 3244- ``ipv6``: match IPv6 header. 3245 3246 - ``tc {unsigned}``: traffic class. 3247 - ``flow {unsigned}``: flow label. 3248 - ``proto {unsigned}``: protocol (next header). 3249 - ``hop {unsigned}``: hop limit. 3250 - ``src {ipv6 address}``: source address. 3251 - ``dst {ipv6 address}``: destination address. 3252 3253- ``icmp``: match ICMP header. 3254 3255 - ``type {unsigned}``: ICMP packet type. 3256 - ``code {unsigned}``: ICMP packet code. 3257 3258- ``udp``: match UDP header. 3259 3260 - ``src {unsigned}``: UDP source port. 3261 - ``dst {unsigned}``: UDP destination port. 3262 3263- ``tcp``: match TCP header. 3264 3265 - ``src {unsigned}``: TCP source port. 3266 - ``dst {unsigned}``: TCP destination port. 3267 3268- ``sctp``: match SCTP header. 3269 3270 - ``src {unsigned}``: SCTP source port. 3271 - ``dst {unsigned}``: SCTP destination port. 3272 - ``tag {unsigned}``: validation tag. 3273 - ``cksum {unsigned}``: checksum. 3274 3275- ``vxlan``: match VXLAN header. 3276 3277 - ``vni {unsigned}``: VXLAN identifier. 3278 3279- ``e_tag``: match IEEE 802.1BR E-Tag header. 3280 3281 - ``grp_ecid_b {unsigned}``: GRP and E-CID base. 3282 3283- ``nvgre``: match NVGRE header. 3284 3285 - ``tni {unsigned}``: virtual subnet ID. 3286 3287- ``mpls``: match MPLS header. 3288 3289 - ``label {unsigned}``: MPLS label. 3290 3291- ``gre``: match GRE header. 3292 3293 - ``protocol {unsigned}``: protocol type. 3294 3295- ``fuzzy``: fuzzy pattern match, expect faster than default. 3296 3297 - ``thresh {unsigned}``: accuracy threshold. 3298 3299- ``gtp``, ``gtpc``, ``gtpu``: match GTPv1 header. 3300 3301 - ``teid {unsigned}``: tunnel endpoint identifier. 3302 3303- ``geneve``: match GENEVE header. 3304 3305 - ``vni {unsigned}``: virtual network identifier. 3306 - ``protocol {unsigned}``: protocol type. 3307 3308Actions list 3309^^^^^^^^^^^^ 3310 3311A list of actions starts after the ``actions`` token in the same fashion as 3312`Matching pattern`_; actions are separated by ``/`` tokens and the list is 3313terminated by a mandatory ``end`` action. 3314 3315Actions are named after their type (*RTE_FLOW_ACTION_TYPE_* from ``enum 3316rte_flow_action_type``). 3317 3318Dropping all incoming UDPv4 packets can be expressed as follows:: 3319 3320 testpmd> flow create 0 ingress pattern eth / ipv4 / udp / end 3321 actions drop / end 3322 3323Several actions have configurable properties which must be specified when 3324there is no valid default value. For example, ``queue`` requires a target 3325queue index. 3326 3327This rule redirects incoming UDPv4 traffic to queue index 6:: 3328 3329 testpmd> flow create 0 ingress pattern eth / ipv4 / udp / end 3330 actions queue index 6 / end 3331 3332While this one could be rejected by PMDs (unspecified queue index):: 3333 3334 testpmd> flow create 0 ingress pattern eth / ipv4 / udp / end 3335 actions queue / end 3336 3337As defined by *rte_flow*, the list is not ordered, all actions of a given 3338rule are performed simultaneously. These are equivalent:: 3339 3340 queue index 6 / void / mark id 42 / end 3341 3342:: 3343 3344 void / mark id 42 / queue index 6 / end 3345 3346All actions in a list should have different types, otherwise only the last 3347action of a given type is taken into account:: 3348 3349 queue index 4 / queue index 5 / queue index 6 / end # will use queue 6 3350 3351:: 3352 3353 drop / drop / drop / end # drop is performed only once 3354 3355:: 3356 3357 mark id 42 / queue index 3 / mark id 24 / end # mark will be 24 3358 3359Considering they are performed simultaneously, opposite and overlapping 3360actions can sometimes be combined when the end result is unambiguous:: 3361 3362 drop / queue index 6 / end # drop has no effect 3363 3364:: 3365 3366 drop / dup index 6 / end # same as above 3367 3368:: 3369 3370 queue index 6 / rss queues 6 7 8 / end # queue has no effect 3371 3372:: 3373 3374 drop / passthru / end # drop has no effect 3375 3376Note that PMDs may still refuse such combinations. 3377 3378Actions 3379^^^^^^^ 3380 3381This section lists supported actions and their attributes, if any. 3382 3383- ``end``: end list of actions. 3384 3385- ``void``: no-op action. 3386 3387- ``passthru``: let subsequent rule process matched packets. 3388 3389- ``mark``: attach 32 bit value to packets. 3390 3391 - ``id {unsigned}``: 32 bit value to return with packets. 3392 3393- ``flag``: flag packets. 3394 3395- ``queue``: assign packets to a given queue index. 3396 3397 - ``index {unsigned}``: queue index to use. 3398 3399- ``drop``: drop packets (note: passthru has priority). 3400 3401- ``count``: enable counters for this rule. 3402 3403- ``dup``: duplicate packets to a given queue index. 3404 3405 - ``index {unsigned}``: queue index to duplicate packets to. 3406 3407- ``rss``: spread packets among several queues. 3408 3409 - ``types [{RSS hash type} [...]] end``: RSS hash types, allowed tokens 3410 are the same as `set_hash_input_set`_, an empty list means none (0). 3411 3412 - ``key {string}``: RSS hash key, overrides ``key_len``. 3413 3414 - ``key_len {unsigned}``: RSS hash key length in bytes, can be used in 3415 conjunction with ``key`` to pad or truncate it. 3416 3417 - ``queues [{unsigned} [...]] end``: queue indices to use. 3418 3419- ``pf``: redirect packets to physical device function. 3420 3421- ``vf``: redirect packets to virtual device function. 3422 3423 - ``original {boolean}``: use original VF ID if possible. 3424 - ``id {unsigned}``: VF ID to redirect packets to. 3425 3426Destroying flow rules 3427~~~~~~~~~~~~~~~~~~~~~ 3428 3429``flow destroy`` destroys one or more rules from their rule ID (as returned 3430by ``flow create``), this command calls ``rte_flow_destroy()`` as many 3431times as necessary:: 3432 3433 flow destroy {port_id} rule {rule_id} [...] 3434 3435If successful, it will show:: 3436 3437 Flow rule #[...] destroyed 3438 3439It does not report anything for rule IDs that do not exist. The usual error 3440message is shown when a rule cannot be destroyed:: 3441 3442 Caught error type [...] ([...]): [...] 3443 3444``flow flush`` destroys all rules on a device and does not take extra 3445arguments. It is bound to ``rte_flow_flush()``:: 3446 3447 flow flush {port_id} 3448 3449Any errors are reported as above. 3450 3451Creating several rules and destroying them:: 3452 3453 testpmd> flow create 0 ingress pattern eth / ipv6 / end 3454 actions queue index 2 / end 3455 Flow rule #0 created 3456 testpmd> flow create 0 ingress pattern eth / ipv4 / end 3457 actions queue index 3 / end 3458 Flow rule #1 created 3459 testpmd> flow destroy 0 rule 0 rule 1 3460 Flow rule #1 destroyed 3461 Flow rule #0 destroyed 3462 testpmd> 3463 3464The same result can be achieved using ``flow flush``:: 3465 3466 testpmd> flow create 0 ingress pattern eth / ipv6 / end 3467 actions queue index 2 / end 3468 Flow rule #0 created 3469 testpmd> flow create 0 ingress pattern eth / ipv4 / end 3470 actions queue index 3 / end 3471 Flow rule #1 created 3472 testpmd> flow flush 0 3473 testpmd> 3474 3475Non-existent rule IDs are ignored:: 3476 3477 testpmd> flow create 0 ingress pattern eth / ipv6 / end 3478 actions queue index 2 / end 3479 Flow rule #0 created 3480 testpmd> flow create 0 ingress pattern eth / ipv4 / end 3481 actions queue index 3 / end 3482 Flow rule #1 created 3483 testpmd> flow destroy 0 rule 42 rule 10 rule 2 3484 testpmd> 3485 testpmd> flow destroy 0 rule 0 3486 Flow rule #0 destroyed 3487 testpmd> 3488 3489Querying flow rules 3490~~~~~~~~~~~~~~~~~~~ 3491 3492``flow query`` queries a specific action of a flow rule having that 3493ability. Such actions collect information that can be reported using this 3494command. It is bound to ``rte_flow_query()``:: 3495 3496 flow query {port_id} {rule_id} {action} 3497 3498If successful, it will display either the retrieved data for known actions 3499or the following message:: 3500 3501 Cannot display result for action type [...] ([...]) 3502 3503Otherwise, it will complain either that the rule does not exist or that some 3504error occurred:: 3505 3506 Flow rule #[...] not found 3507 3508:: 3509 3510 Caught error type [...] ([...]): [...] 3511 3512Currently only the ``count`` action is supported. This action reports the 3513number of packets that hit the flow rule and the total number of bytes. Its 3514output has the following format:: 3515 3516 count: 3517 hits_set: [...] # whether "hits" contains a valid value 3518 bytes_set: [...] # whether "bytes" contains a valid value 3519 hits: [...] # number of packets 3520 bytes: [...] # number of bytes 3521 3522Querying counters for TCPv6 packets redirected to queue 6:: 3523 3524 testpmd> flow create 0 ingress pattern eth / ipv6 / tcp / end 3525 actions queue index 6 / count / end 3526 Flow rule #4 created 3527 testpmd> flow query 0 4 count 3528 count: 3529 hits_set: 1 3530 bytes_set: 0 3531 hits: 386446 3532 bytes: 0 3533 testpmd> 3534 3535Listing flow rules 3536~~~~~~~~~~~~~~~~~~ 3537 3538``flow list`` lists existing flow rules sorted by priority and optionally 3539filtered by group identifiers:: 3540 3541 flow list {port_id} [group {group_id}] [...] 3542 3543This command only fails with the following message if the device does not 3544exist:: 3545 3546 Invalid port [...] 3547 3548Output consists of a header line followed by a short description of each 3549flow rule, one per line. There is no output at all when no flow rules are 3550configured on the device:: 3551 3552 ID Group Prio Attr Rule 3553 [...] [...] [...] [...] [...] 3554 3555``Attr`` column flags: 3556 3557- ``i`` for ``ingress``. 3558- ``e`` for ``egress``. 3559 3560Creating several flow rules and listing them:: 3561 3562 testpmd> flow create 0 ingress pattern eth / ipv4 / end 3563 actions queue index 6 / end 3564 Flow rule #0 created 3565 testpmd> flow create 0 ingress pattern eth / ipv6 / end 3566 actions queue index 2 / end 3567 Flow rule #1 created 3568 testpmd> flow create 0 priority 5 ingress pattern eth / ipv4 / udp / end 3569 actions rss queues 6 7 8 end / end 3570 Flow rule #2 created 3571 testpmd> flow list 0 3572 ID Group Prio Attr Rule 3573 0 0 0 i- ETH IPV4 => QUEUE 3574 1 0 0 i- ETH IPV6 => QUEUE 3575 2 0 5 i- ETH IPV4 UDP => RSS 3576 testpmd> 3577 3578Rules are sorted by priority (i.e. group ID first, then priority level):: 3579 3580 testpmd> flow list 1 3581 ID Group Prio Attr Rule 3582 0 0 0 i- ETH => COUNT 3583 6 0 500 i- ETH IPV6 TCP => DROP COUNT 3584 5 0 1000 i- ETH IPV6 ICMP => QUEUE 3585 1 24 0 i- ETH IPV4 UDP => QUEUE 3586 4 24 10 i- ETH IPV4 TCP => DROP 3587 3 24 20 i- ETH IPV4 => DROP 3588 2 24 42 i- ETH IPV4 UDP => QUEUE 3589 7 63 0 i- ETH IPV6 UDP VXLAN => MARK QUEUE 3590 testpmd> 3591 3592Output can be limited to specific groups:: 3593 3594 testpmd> flow list 1 group 0 group 63 3595 ID Group Prio Attr Rule 3596 0 0 0 i- ETH => COUNT 3597 6 0 500 i- ETH IPV6 TCP => DROP COUNT 3598 5 0 1000 i- ETH IPV6 ICMP => QUEUE 3599 7 63 0 i- ETH IPV6 UDP VXLAN => MARK QUEUE 3600 testpmd> 3601 3602Toggling isolated mode 3603~~~~~~~~~~~~~~~~~~~~~~ 3604 3605``flow isolate`` can be used to tell the underlying PMD that ingress traffic 3606must only be injected from the defined flow rules; that no default traffic 3607is expected outside those rules and the driver is free to assign more 3608resources to handle them. It is bound to ``rte_flow_isolate()``:: 3609 3610 flow isolate {port_id} {boolean} 3611 3612If successful, enabling or disabling isolated mode shows either:: 3613 3614 Ingress traffic on port [...] 3615 is now restricted to the defined flow rules 3616 3617Or:: 3618 3619 Ingress traffic on port [...] 3620 is not restricted anymore to the defined flow rules 3621 3622Otherwise, in case of error:: 3623 3624 Caught error type [...] ([...]): [...] 3625 3626Mainly due to its side effects, PMDs supporting this mode may not have the 3627ability to toggle it more than once without reinitializing affected ports 3628first (e.g. by exiting testpmd). 3629 3630Enabling isolated mode:: 3631 3632 testpmd> flow isolate 0 true 3633 Ingress traffic on port 0 is now restricted to the defined flow rules 3634 testpmd> 3635 3636Disabling isolated mode:: 3637 3638 testpmd> flow isolate 0 false 3639 Ingress traffic on port 0 is not restricted anymore to the defined flow rules 3640 testpmd> 3641 3642Sample QinQ flow rules 3643~~~~~~~~~~~~~~~~~~~~~~ 3644 3645Before creating QinQ rule(s) the following commands should be issued to enable QinQ:: 3646 3647 testpmd> port stop 0 3648 testpmd> vlan set qinq on 0 3649 3650The above command sets the inner and outer TPID's to 0x8100. 3651 3652To change the TPID's the following commands should be used:: 3653 3654 testpmd> vlan set outer tpid 0xa100 0 3655 testpmd> vlan set inner tpid 0x9100 0 3656 testpmd> port start 0 3657 3658Validate and create a QinQ rule on port 0 to steer traffic to a VF queue in a VM. 3659 3660:: 3661 3662 testpmd> flow validate 0 ingress pattern eth / vlan tci is 123 / 3663 vlan tci is 456 / end actions vf id 1 / queue index 0 / end 3664 Flow rule #0 validated 3665 3666 testpmd> flow create 0 ingress pattern eth / vlan tci is 4 / 3667 vlan tci is 456 / end actions vf id 123 / queue index 0 / end 3668 Flow rule #0 created 3669 3670 testpmd> flow list 0 3671 ID Group Prio Attr Rule 3672 0 0 0 i- ETH VLAN VLAN=>VF QUEUE 3673 3674Validate and create a QinQ rule on port 0 to steer traffic to a queue on the host. 3675 3676:: 3677 3678 testpmd> flow validate 0 ingress pattern eth / vlan tci is 321 / 3679 vlan tci is 654 / end actions pf / queue index 0 / end 3680 Flow rule #1 validated 3681 3682 testpmd> flow create 0 ingress pattern eth / vlan tci is 321 / 3683 vlan tci is 654 / end actions pf / queue index 1 / end 3684 Flow rule #1 created 3685 3686 testpmd> flow list 0 3687 ID Group Prio Attr Rule 3688 0 0 0 i- ETH VLAN VLAN=>VF QUEUE 3689 1 0 0 i- ETH VLAN VLAN=>PF QUEUE 3690