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|summary|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* ``summary``: Brief port summary such as Device Name, Driver Name etc. 169 170* ``stats``: RX/TX statistics. 171 172* ``xstats``: RX/TX extended NIC statistics. 173 174* ``fdir``: Flow Director information and statistics. 175 176* ``stat_qmap``: Queue statistics mapping. 177 178* ``dcb_tc``: DCB information such as TC mapping. 179 180* ``cap``: Supported offload capabilities. 181 182For example: 183 184.. code-block:: console 185 186 testpmd> show port info 0 187 188 ********************* Infos for port 0 ********************* 189 190 MAC address: XX:XX:XX:XX:XX:XX 191 Connect to socket: 0 192 memory allocation on the socket: 0 193 Link status: up 194 Link speed: 40000 Mbps 195 Link duplex: full-duplex 196 Promiscuous mode: enabled 197 Allmulticast mode: disabled 198 Maximum number of MAC addresses: 64 199 Maximum number of MAC addresses of hash filtering: 0 200 VLAN offload: 201 strip on 202 filter on 203 qinq(extend) off 204 Redirection table size: 512 205 Supported flow types: 206 ipv4-frag 207 ipv4-tcp 208 ipv4-udp 209 ipv4-sctp 210 ipv4-other 211 ipv6-frag 212 ipv6-tcp 213 ipv6-udp 214 ipv6-sctp 215 ipv6-other 216 l2_payload 217 port 218 vxlan 219 geneve 220 nvgre 221 222show port rss reta 223~~~~~~~~~~~~~~~~~~ 224 225Display the rss redirection table entry indicated by masks on port X:: 226 227 testpmd> show port (port_id) rss reta (size) (mask0, mask1...) 228 229size is used to indicate the hardware supported reta size 230 231show port rss-hash 232~~~~~~~~~~~~~~~~~~ 233 234Display the RSS hash functions and RSS hash key of a port:: 235 236 testpmd> show port (port_id) rss-hash [key] 237 238clear port 239~~~~~~~~~~ 240 241Clear the port statistics for a given port or for all ports:: 242 243 testpmd> clear port (info|stats|xstats|fdir|stat_qmap) (port_id|all) 244 245For example:: 246 247 testpmd> clear port stats all 248 249show (rxq|txq) 250~~~~~~~~~~~~~~ 251 252Display information for a given port's RX/TX queue:: 253 254 testpmd> show (rxq|txq) info (port_id) (queue_id) 255 256show config 257~~~~~~~~~~~ 258 259Displays the configuration of the application. 260The configuration comes from the command-line, the runtime or the application defaults:: 261 262 testpmd> show config (rxtx|cores|fwd|txpkts) 263 264The available information categories are: 265 266* ``rxtx``: RX/TX configuration items. 267 268* ``cores``: List of forwarding cores. 269 270* ``fwd``: Packet forwarding configuration. 271 272* ``txpkts``: Packets to TX configuration. 273 274For example: 275 276.. code-block:: console 277 278 testpmd> show config rxtx 279 280 io packet forwarding - CRC stripping disabled - packets/burst=16 281 nb forwarding cores=2 - nb forwarding ports=1 282 RX queues=1 - RX desc=128 - RX free threshold=0 283 RX threshold registers: pthresh=8 hthresh=8 wthresh=4 284 TX queues=1 - TX desc=512 - TX free threshold=0 285 TX threshold registers: pthresh=36 hthresh=0 wthresh=0 286 TX RS bit threshold=0 - TXQ flags=0x0 287 288set fwd 289~~~~~~~ 290 291Set the packet forwarding mode:: 292 293 testpmd> set fwd (io|mac|macswap|flowgen| \ 294 rxonly|txonly|csum|icmpecho|noisy) (""|retry) 295 296``retry`` can be specified for forwarding engines except ``rx_only``. 297 298The available information categories are: 299 300* ``io``: Forwards packets "as-is" in I/O mode. 301 This is the fastest possible forwarding operation as it does not access packets data. 302 This is the default mode. 303 304* ``mac``: Changes the source and the destination Ethernet addresses of packets before forwarding them. 305 Default application behaviour is to set source Ethernet address to that of the transmitting interface, and destination 306 address to a dummy value (set during init). The user may specify a target destination Ethernet address via the 'eth-peer' or 307 'eth-peer-configfile' command-line options. It is not currently possible to specify a specific source Ethernet address. 308 309* ``macswap``: MAC swap forwarding mode. 310 Swaps the source and the destination Ethernet addresses of packets before forwarding them. 311 312* ``flowgen``: Multi-flow generation mode. 313 Originates a number of flows (with varying destination IP addresses), and terminate receive traffic. 314 315* ``rxonly``: Receives packets but doesn't transmit them. 316 317* ``txonly``: Generates and transmits packets without receiving any. 318 319* ``csum``: Changes the checksum field with hardware or software methods depending on the offload flags on the packet. 320 321* ``icmpecho``: Receives a burst of packets, lookup for IMCP echo requests and, if any, send back ICMP echo replies. 322 323* ``ieee1588``: Demonstrate L2 IEEE1588 V2 PTP timestamping for RX and TX. Requires ``CONFIG_RTE_LIBRTE_IEEE1588=y``. 324 325* ``softnic``: Demonstrates the softnic forwarding operation. In this mode, packet forwarding is 326 similar to I/O mode except for the fact that packets are loopback to the softnic ports only. Therefore, portmask parameter should be set to softnic port only. The various software based custom NIC pipelines specified through the softnic firmware (DPDK packet framework script) can be tested in this mode. Furthermore, it allows to build 5-level hierarchical QoS scheduler as a default option that can be enabled through CLI once testpmd application is initialised. The user can modify the default scheduler hierarchy or can specify the new QoS Scheduler hierarchy through CLI. Requires ``CONFIG_RTE_LIBRTE_PMD_SOFTNIC=y``. 327 328* ``noisy``: Noisy neighbour simulation. 329 Simulate more realistic behavior of a guest machine engaged in receiving 330 and sending packets performing Virtual Network Function (VNF). 331 332Example:: 333 334 testpmd> set fwd rxonly 335 336 Set rxonly packet forwarding mode 337 338 339read rxd 340~~~~~~~~ 341 342Display an RX descriptor for a port RX queue:: 343 344 testpmd> read rxd (port_id) (queue_id) (rxd_id) 345 346For example:: 347 348 testpmd> read rxd 0 0 4 349 0x0000000B - 0x001D0180 / 0x0000000B - 0x001D0180 350 351read txd 352~~~~~~~~ 353 354Display a TX descriptor for a port TX queue:: 355 356 testpmd> read txd (port_id) (queue_id) (txd_id) 357 358For example:: 359 360 testpmd> read txd 0 0 4 361 0x00000001 - 0x24C3C440 / 0x000F0000 - 0x2330003C 362 363ddp get list 364~~~~~~~~~~~~ 365 366Get loaded dynamic device personalization (DDP) package info list:: 367 368 testpmd> ddp get list (port_id) 369 370ddp get info 371~~~~~~~~~~~~ 372 373Display information about dynamic device personalization (DDP) profile:: 374 375 testpmd> ddp get info (profile_path) 376 377show vf stats 378~~~~~~~~~~~~~ 379 380Display VF statistics:: 381 382 testpmd> show vf stats (port_id) (vf_id) 383 384clear vf stats 385~~~~~~~~~~~~~~ 386 387Reset VF statistics:: 388 389 testpmd> clear vf stats (port_id) (vf_id) 390 391show port pctype mapping 392~~~~~~~~~~~~~~~~~~~~~~~~ 393 394List all items from the pctype mapping table:: 395 396 testpmd> show port (port_id) pctype mapping 397 398show rx offloading capabilities 399~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ 400 401List all per queue and per port Rx offloading capabilities of a port:: 402 403 testpmd> show port (port_id) rx_offload capabilities 404 405show rx offloading configuration 406~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ 407 408List port level and all queue level Rx offloading configuration:: 409 410 testpmd> show port (port_id) rx_offload configuration 411 412show tx offloading capabilities 413~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ 414 415List all per queue and per port Tx offloading capabilities of a port:: 416 417 testpmd> show port (port_id) tx_offload capabilities 418 419show tx offloading configuration 420~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ 421 422List port level and all queue level Tx offloading configuration:: 423 424 testpmd> show port (port_id) tx_offload configuration 425 426 427Configuration Functions 428----------------------- 429 430The testpmd application can be configured from the runtime as well as from the command-line. 431 432This section details the available configuration functions that are available. 433 434.. note:: 435 436 Configuration changes only become active when forwarding is started/restarted. 437 438set default 439~~~~~~~~~~~ 440 441Reset forwarding to the default configuration:: 442 443 testpmd> set default 444 445set verbose 446~~~~~~~~~~~ 447 448Set the debug verbosity level:: 449 450 testpmd> set verbose (level) 451 452Available levels are as following: 453 454* ``0`` silent except for error. 455* ``1`` fully verbose except for Tx packets. 456* ``2`` fully verbose except for Rx packets. 457* ``> 2`` fully verbose. 458 459set log 460~~~~~~~ 461 462Set the log level for a log type:: 463 464 testpmd> set log global|(type) (level) 465 466Where: 467 468* ``type`` is the log name. 469 470* ``level`` is the log level. 471 472For example, to change the global log level:: 473 testpmd> set log global (level) 474 475Regexes can also be used for type. To change log level of user1, user2 and user3:: 476 testpmd> set log user[1-3] (level) 477 478set nbport 479~~~~~~~~~~ 480 481Set the number of ports used by the application: 482 483set nbport (num) 484 485This is equivalent to the ``--nb-ports`` command-line option. 486 487set nbcore 488~~~~~~~~~~ 489 490Set the number of cores used by the application:: 491 492 testpmd> set nbcore (num) 493 494This is equivalent to the ``--nb-cores`` command-line option. 495 496.. note:: 497 498 The number of cores used must not be greater than number of ports used multiplied by the number of queues per port. 499 500set coremask 501~~~~~~~~~~~~ 502 503Set the forwarding cores hexadecimal mask:: 504 505 testpmd> set coremask (mask) 506 507This is equivalent to the ``--coremask`` command-line option. 508 509.. note:: 510 511 The master lcore is reserved for command line parsing only and cannot be masked on for packet forwarding. 512 513set portmask 514~~~~~~~~~~~~ 515 516Set the forwarding ports hexadecimal mask:: 517 518 testpmd> set portmask (mask) 519 520This is equivalent to the ``--portmask`` command-line option. 521 522set burst 523~~~~~~~~~ 524 525Set number of packets per burst:: 526 527 testpmd> set burst (num) 528 529This is equivalent to the ``--burst command-line`` option. 530 531When retry is enabled, the transmit delay time and number of retries can also be set:: 532 533 testpmd> set burst tx delay (microseconds) retry (num) 534 535set txpkts 536~~~~~~~~~~ 537 538Set the length of each segment of the TX-ONLY packets or length of packet for FLOWGEN mode:: 539 540 testpmd> set txpkts (x[,y]*) 541 542Where x[,y]* represents a CSV list of values, without white space. 543 544set txsplit 545~~~~~~~~~~~ 546 547Set the split policy for the TX packets, applicable for TX-ONLY and CSUM forwarding modes:: 548 549 testpmd> set txsplit (off|on|rand) 550 551Where: 552 553* ``off`` disable packet copy & split for CSUM mode. 554 555* ``on`` split outgoing packet into multiple segments. Size of each segment 556 and number of segments per packet is determined by ``set txpkts`` command 557 (see above). 558 559* ``rand`` same as 'on', but number of segments per each packet is a random value between 1 and total number of segments. 560 561set corelist 562~~~~~~~~~~~~ 563 564Set the list of forwarding cores:: 565 566 testpmd> set corelist (x[,y]*) 567 568For example, to change the forwarding cores: 569 570.. code-block:: console 571 572 testpmd> set corelist 3,1 573 testpmd> show config fwd 574 575 io packet forwarding - ports=2 - cores=2 - streams=2 - NUMA support disabled 576 Logical Core 3 (socket 0) forwards packets on 1 streams: 577 RX P=0/Q=0 (socket 0) -> TX P=1/Q=0 (socket 0) peer=02:00:00:00:00:01 578 Logical Core 1 (socket 0) forwards packets on 1 streams: 579 RX P=1/Q=0 (socket 0) -> TX P=0/Q=0 (socket 0) peer=02:00:00:00:00:00 580 581.. note:: 582 583 The cores are used in the same order as specified on the command line. 584 585set portlist 586~~~~~~~~~~~~ 587 588Set the list of forwarding ports:: 589 590 testpmd> set portlist (x[,y]*) 591 592For example, to change the port forwarding: 593 594.. code-block:: console 595 596 testpmd> set portlist 0,2,1,3 597 testpmd> show config fwd 598 599 io packet forwarding - ports=4 - cores=1 - streams=4 600 Logical Core 3 (socket 0) forwards packets on 4 streams: 601 RX P=0/Q=0 (socket 0) -> TX P=2/Q=0 (socket 0) peer=02:00:00:00:00:01 602 RX P=2/Q=0 (socket 0) -> TX P=0/Q=0 (socket 0) peer=02:00:00:00:00:00 603 RX P=1/Q=0 (socket 0) -> TX P=3/Q=0 (socket 0) peer=02:00:00:00:00:03 604 RX P=3/Q=0 (socket 0) -> TX P=1/Q=0 (socket 0) peer=02:00:00:00:00:02 605 606set tx loopback 607~~~~~~~~~~~~~~~ 608 609Enable/disable tx loopback:: 610 611 testpmd> set tx loopback (port_id) (on|off) 612 613set drop enable 614~~~~~~~~~~~~~~~ 615 616set drop enable bit for all queues:: 617 618 testpmd> set all queues drop (port_id) (on|off) 619 620set split drop enable (for VF) 621~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ 622 623set split drop enable bit for VF from PF:: 624 625 testpmd> set vf split drop (port_id) (vf_id) (on|off) 626 627set mac antispoof (for VF) 628~~~~~~~~~~~~~~~~~~~~~~~~~~ 629 630Set mac antispoof for a VF from the PF:: 631 632 testpmd> set vf mac antispoof (port_id) (vf_id) (on|off) 633 634set macsec offload 635~~~~~~~~~~~~~~~~~~ 636 637Enable/disable MACsec offload:: 638 639 testpmd> set macsec offload (port_id) on encrypt (on|off) replay-protect (on|off) 640 testpmd> set macsec offload (port_id) off 641 642set macsec sc 643~~~~~~~~~~~~~ 644 645Configure MACsec secure connection (SC):: 646 647 testpmd> set macsec sc (tx|rx) (port_id) (mac) (pi) 648 649.. note:: 650 651 The pi argument is ignored for tx. 652 Check the NIC Datasheet for hardware limits. 653 654set macsec sa 655~~~~~~~~~~~~~ 656 657Configure MACsec secure association (SA):: 658 659 testpmd> set macsec sa (tx|rx) (port_id) (idx) (an) (pn) (key) 660 661.. note:: 662 663 The IDX value must be 0 or 1. 664 Check the NIC Datasheet for hardware limits. 665 666set broadcast mode (for VF) 667~~~~~~~~~~~~~~~~~~~~~~~~~~~ 668 669Set broadcast mode for a VF from the PF:: 670 671 testpmd> set vf broadcast (port_id) (vf_id) (on|off) 672 673vlan set strip 674~~~~~~~~~~~~~~ 675 676Set the VLAN strip on a port:: 677 678 testpmd> vlan set strip (on|off) (port_id) 679 680vlan set stripq 681~~~~~~~~~~~~~~~ 682 683Set the VLAN strip for a queue on a port:: 684 685 testpmd> vlan set stripq (on|off) (port_id,queue_id) 686 687vlan set stripq (for VF) 688~~~~~~~~~~~~~~~~~~~~~~~~ 689 690Set VLAN strip for all queues in a pool for a VF from the PF:: 691 692 testpmd> set vf vlan stripq (port_id) (vf_id) (on|off) 693 694vlan set insert (for VF) 695~~~~~~~~~~~~~~~~~~~~~~~~ 696 697Set VLAN insert for a VF from the PF:: 698 699 testpmd> set vf vlan insert (port_id) (vf_id) (vlan_id) 700 701vlan set tag (for VF) 702~~~~~~~~~~~~~~~~~~~~~ 703 704Set VLAN tag for a VF from the PF:: 705 706 testpmd> set vf vlan tag (port_id) (vf_id) (on|off) 707 708vlan set antispoof (for VF) 709~~~~~~~~~~~~~~~~~~~~~~~~~~~ 710 711Set VLAN antispoof for a VF from the PF:: 712 713 testpmd> set vf vlan antispoof (port_id) (vf_id) (on|off) 714 715vlan set filter 716~~~~~~~~~~~~~~~ 717 718Set the VLAN filter on a port:: 719 720 testpmd> vlan set filter (on|off) (port_id) 721 722vlan set qinq 723~~~~~~~~~~~~~ 724 725Set the VLAN QinQ (extended queue in queue) on for a port:: 726 727 testpmd> vlan set qinq (on|off) (port_id) 728 729vlan set tpid 730~~~~~~~~~~~~~ 731 732Set the inner or outer VLAN TPID for packet filtering on a port:: 733 734 testpmd> vlan set (inner|outer) tpid (value) (port_id) 735 736.. note:: 737 738 TPID value must be a 16-bit number (value <= 65536). 739 740rx_vlan add 741~~~~~~~~~~~ 742 743Add a VLAN ID, or all identifiers, to the set of VLAN identifiers filtered by port ID:: 744 745 testpmd> rx_vlan add (vlan_id|all) (port_id) 746 747.. note:: 748 749 VLAN filter must be set on that port. VLAN ID < 4096. 750 Depending on the NIC used, number of vlan_ids may be limited to the maximum entries 751 in VFTA table. This is important if enabling all vlan_ids. 752 753rx_vlan rm 754~~~~~~~~~~ 755 756Remove a VLAN ID, or all identifiers, from the set of VLAN identifiers filtered by port ID:: 757 758 testpmd> rx_vlan rm (vlan_id|all) (port_id) 759 760rx_vlan add (for VF) 761~~~~~~~~~~~~~~~~~~~~ 762 763Add a VLAN ID, to the set of VLAN identifiers filtered for VF(s) for port ID:: 764 765 testpmd> rx_vlan add (vlan_id) port (port_id) vf (vf_mask) 766 767rx_vlan rm (for VF) 768~~~~~~~~~~~~~~~~~~~ 769 770Remove a VLAN ID, from the set of VLAN identifiers filtered for VF(s) for port ID:: 771 772 testpmd> rx_vlan rm (vlan_id) port (port_id) vf (vf_mask) 773 774tunnel_filter add 775~~~~~~~~~~~~~~~~~ 776 777Add a tunnel filter on a port:: 778 779 testpmd> tunnel_filter add (port_id) (outer_mac) (inner_mac) (ip_addr) \ 780 (inner_vlan) (vxlan|nvgre|ipingre) (imac-ivlan|imac-ivlan-tenid|\ 781 imac-tenid|imac|omac-imac-tenid|oip|iip) (tenant_id) (queue_id) 782 783The available information categories are: 784 785* ``vxlan``: Set tunnel type as VXLAN. 786 787* ``nvgre``: Set tunnel type as NVGRE. 788 789* ``ipingre``: Set tunnel type as IP-in-GRE. 790 791* ``imac-ivlan``: Set filter type as Inner MAC and VLAN. 792 793* ``imac-ivlan-tenid``: Set filter type as Inner MAC, VLAN and tenant ID. 794 795* ``imac-tenid``: Set filter type as Inner MAC and tenant ID. 796 797* ``imac``: Set filter type as Inner MAC. 798 799* ``omac-imac-tenid``: Set filter type as Outer MAC, Inner MAC and tenant ID. 800 801* ``oip``: Set filter type as Outer IP. 802 803* ``iip``: Set filter type as Inner IP. 804 805Example:: 806 807 testpmd> tunnel_filter add 0 68:05:CA:28:09:82 00:00:00:00:00:00 \ 808 192.168.2.2 0 ipingre oip 1 1 809 810 Set an IP-in-GRE tunnel on port 0, and the filter type is Outer IP. 811 812tunnel_filter remove 813~~~~~~~~~~~~~~~~~~~~ 814 815Remove a tunnel filter on a port:: 816 817 testpmd> tunnel_filter rm (port_id) (outer_mac) (inner_mac) (ip_addr) \ 818 (inner_vlan) (vxlan|nvgre|ipingre) (imac-ivlan|imac-ivlan-tenid|\ 819 imac-tenid|imac|omac-imac-tenid|oip|iip) (tenant_id) (queue_id) 820 821rx_vxlan_port add 822~~~~~~~~~~~~~~~~~ 823 824Add an UDP port for VXLAN packet filter on a port:: 825 826 testpmd> rx_vxlan_port add (udp_port) (port_id) 827 828rx_vxlan_port remove 829~~~~~~~~~~~~~~~~~~~~ 830 831Remove an UDP port for VXLAN packet filter on a port:: 832 833 testpmd> rx_vxlan_port rm (udp_port) (port_id) 834 835tx_vlan set 836~~~~~~~~~~~ 837 838Set hardware insertion of VLAN IDs in packets sent on a port:: 839 840 testpmd> tx_vlan set (port_id) vlan_id[, vlan_id_outer] 841 842For example, set a single VLAN ID (5) insertion on port 0:: 843 844 tx_vlan set 0 5 845 846Or, set double VLAN ID (inner: 2, outer: 3) insertion on port 1:: 847 848 tx_vlan set 1 2 3 849 850 851tx_vlan set pvid 852~~~~~~~~~~~~~~~~ 853 854Set port based hardware insertion of VLAN ID in packets sent on a port:: 855 856 testpmd> tx_vlan set pvid (port_id) (vlan_id) (on|off) 857 858tx_vlan reset 859~~~~~~~~~~~~~ 860 861Disable hardware insertion of a VLAN header in packets sent on a port:: 862 863 testpmd> tx_vlan reset (port_id) 864 865csum set 866~~~~~~~~ 867 868Select hardware or software calculation of the checksum when 869transmitting a packet using the ``csum`` forwarding engine:: 870 871 testpmd> csum set (ip|udp|tcp|sctp|outer-ip|outer-udp) (hw|sw) (port_id) 872 873Where: 874 875* ``ip|udp|tcp|sctp`` always relate to the inner layer. 876 877* ``outer-ip`` relates to the outer IP layer (only for IPv4) in the case where the packet is recognized 878 as a tunnel packet by the forwarding engine (vxlan, gre and ipip are 879 supported). See also the ``csum parse-tunnel`` command. 880 881* ``outer-udp`` relates to the outer UDP layer in the case where the packet is recognized 882 as a tunnel packet by the forwarding engine (vxlan, vxlan-gpe are 883 supported). See also the ``csum parse-tunnel`` command. 884 885.. note:: 886 887 Check the NIC Datasheet for hardware limits. 888 889RSS queue region 890~~~~~~~~~~~~~~~~ 891 892Set RSS queue region span on a port:: 893 894 testpmd> set port (port_id) queue-region region_id (value) \ 895 queue_start_index (value) queue_num (value) 896 897Set flowtype mapping on a RSS queue region on a port:: 898 899 testpmd> set port (port_id) queue-region region_id (value) flowtype (value) 900 901where: 902 903* For the flowtype(pctype) of packet,the specific index for each type has 904 been defined in file i40e_type.h as enum i40e_filter_pctype. 905 906Set user priority mapping on a RSS queue region on a port:: 907 908 testpmd> set port (port_id) queue-region UP (value) region_id (value) 909 910Flush all queue region related configuration on a port:: 911 912 testpmd> set port (port_id) queue-region flush (on|off) 913 914where: 915 916* "on"is just an enable function which server for other configuration, 917 it is for all configuration about queue region from up layer, 918 at first will only keep in DPDK softwarestored in driver, 919 only after "flush on", it commit all configuration to HW. 920 "off" is just clean all configuration about queue region just now, 921 and restore all to DPDK i40e driver default config when start up. 922 923Show all queue region related configuration info on a port:: 924 925 testpmd> show port (port_id) queue-region 926 927.. note:: 928 929 Queue region only support on PF by now, so these command is 930 only for configuration of queue region on PF port. 931 932csum parse-tunnel 933~~~~~~~~~~~~~~~~~ 934 935Define how tunneled packets should be handled by the csum forward 936engine:: 937 938 testpmd> csum parse-tunnel (on|off) (tx_port_id) 939 940If enabled, the csum forward engine will try to recognize supported 941tunnel headers (vxlan, gre, ipip). 942 943If disabled, treat tunnel packets as non-tunneled packets (a inner 944header is handled as a packet payload). 945 946.. note:: 947 948 The port argument is the TX port like in the ``csum set`` command. 949 950Example: 951 952Consider a packet in packet like the following:: 953 954 eth_out/ipv4_out/udp_out/vxlan/eth_in/ipv4_in/tcp_in 955 956* If parse-tunnel is enabled, the ``ip|udp|tcp|sctp`` parameters of ``csum set`` 957 command relate to the inner headers (here ``ipv4_in`` and ``tcp_in``), and the 958 ``outer-ip|outer-udp`` parameter relates to the outer headers (here ``ipv4_out`` and ``udp_out``). 959 960* If parse-tunnel is disabled, the ``ip|udp|tcp|sctp`` parameters of ``csum set`` 961 command relate to the outer headers, here ``ipv4_out`` and ``udp_out``. 962 963csum show 964~~~~~~~~~ 965 966Display tx checksum offload configuration:: 967 968 testpmd> csum show (port_id) 969 970tso set 971~~~~~~~ 972 973Enable TCP Segmentation Offload (TSO) in the ``csum`` forwarding engine:: 974 975 testpmd> tso set (segsize) (port_id) 976 977.. note:: 978 979 Check the NIC datasheet for hardware limits. 980 981tso show 982~~~~~~~~ 983 984Display the status of TCP Segmentation Offload:: 985 986 testpmd> tso show (port_id) 987 988set port - gro 989~~~~~~~~~~~~~~ 990 991Enable or disable GRO in ``csum`` forwarding engine:: 992 993 testpmd> set port <port_id> gro on|off 994 995If enabled, the csum forwarding engine will perform GRO on the TCP/IPv4 996packets received from the given port. 997 998If disabled, packets received from the given port won't be performed 999GRO. By default, GRO is disabled for all ports. 1000 1001.. note:: 1002 1003 When enable GRO for a port, TCP/IPv4 packets received from the port 1004 will be performed GRO. After GRO, all merged packets have bad 1005 checksums, since the GRO library doesn't re-calculate checksums for 1006 the merged packets. Therefore, if users want the merged packets to 1007 have correct checksums, please select HW IP checksum calculation and 1008 HW TCP checksum calculation for the port which the merged packets are 1009 transmitted to. 1010 1011show port - gro 1012~~~~~~~~~~~~~~~ 1013 1014Display GRO configuration for a given port:: 1015 1016 testpmd> show port <port_id> gro 1017 1018set gro flush 1019~~~~~~~~~~~~~ 1020 1021Set the cycle to flush the GROed packets from reassembly tables:: 1022 1023 testpmd> set gro flush <cycles> 1024 1025When enable GRO, the csum forwarding engine performs GRO on received 1026packets, and the GROed packets are stored in reassembly tables. Users 1027can use this command to determine when the GROed packets are flushed 1028from the reassembly tables. 1029 1030The ``cycles`` is measured in GRO operation times. The csum forwarding 1031engine flushes the GROed packets from the tables every ``cycles`` GRO 1032operations. 1033 1034By default, the value of ``cycles`` is 1, which means flush GROed packets 1035from the reassembly tables as soon as one GRO operation finishes. The value 1036of ``cycles`` should be in the range of 1 to ``GRO_MAX_FLUSH_CYCLES``. 1037 1038Please note that the large value of ``cycles`` may cause the poor TCP/IP 1039stack performance. Because the GROed packets are delayed to arrive the 1040stack, thus causing more duplicated ACKs and TCP retransmissions. 1041 1042set port - gso 1043~~~~~~~~~~~~~~ 1044 1045Toggle per-port GSO support in ``csum`` forwarding engine:: 1046 1047 testpmd> set port <port_id> gso on|off 1048 1049If enabled, the csum forwarding engine will perform GSO on supported IPv4 1050packets, transmitted on the given port. 1051 1052If disabled, packets transmitted on the given port will not undergo GSO. 1053By default, GSO is disabled for all ports. 1054 1055.. note:: 1056 1057 When GSO is enabled on a port, supported IPv4 packets transmitted on that 1058 port undergo GSO. Afterwards, the segmented packets are represented by 1059 multi-segment mbufs; however, the csum forwarding engine doesn't calculation 1060 of checksums for GSO'd segments in SW. As a result, if users want correct 1061 checksums in GSO segments, they should enable HW checksum calculation for 1062 GSO-enabled ports. 1063 1064 For example, HW checksum calculation for VxLAN GSO'd packets may be enabled 1065 by setting the following options in the csum forwarding engine: 1066 1067 testpmd> csum set outer_ip hw <port_id> 1068 1069 testpmd> csum set ip hw <port_id> 1070 1071 testpmd> csum set tcp hw <port_id> 1072 1073 UDP GSO is the same as IP fragmentation, which treats the UDP header 1074 as the payload and does not modify it during segmentation. That is, 1075 after UDP GSO, only the first output fragment has the original UDP 1076 header. Therefore, users need to enable HW IP checksum calculation 1077 and SW UDP checksum calculation for GSO-enabled ports, if they want 1078 correct checksums for UDP/IPv4 packets. 1079 1080set gso segsz 1081~~~~~~~~~~~~~ 1082 1083Set the maximum GSO segment size (measured in bytes), which includes the 1084packet header and the packet payload for GSO-enabled ports (global):: 1085 1086 testpmd> set gso segsz <length> 1087 1088show port - gso 1089~~~~~~~~~~~~~~~ 1090 1091Display the status of Generic Segmentation Offload for a given port:: 1092 1093 testpmd> show port <port_id> gso 1094 1095mac_addr add 1096~~~~~~~~~~~~ 1097 1098Add an alternative MAC address to a port:: 1099 1100 testpmd> mac_addr add (port_id) (XX:XX:XX:XX:XX:XX) 1101 1102mac_addr remove 1103~~~~~~~~~~~~~~~ 1104 1105Remove a MAC address from a port:: 1106 1107 testpmd> mac_addr remove (port_id) (XX:XX:XX:XX:XX:XX) 1108 1109mac_addr add (for VF) 1110~~~~~~~~~~~~~~~~~~~~~ 1111 1112Add an alternative MAC address for a VF to a port:: 1113 1114 testpmd> mac_add add port (port_id) vf (vf_id) (XX:XX:XX:XX:XX:XX) 1115 1116mac_addr set 1117~~~~~~~~~~~~ 1118 1119Set the default MAC address for a port:: 1120 1121 testpmd> mac_addr set (port_id) (XX:XX:XX:XX:XX:XX) 1122 1123mac_addr set (for VF) 1124~~~~~~~~~~~~~~~~~~~~~ 1125 1126Set the MAC address for a VF from the PF:: 1127 1128 testpmd> set vf mac addr (port_id) (vf_id) (XX:XX:XX:XX:XX:XX) 1129 1130set eth-peer 1131~~~~~~~~~~~~ 1132 1133Set the forwarding peer address for certain port:: 1134 1135 testpmd> set eth-peer (port_id) (perr_addr) 1136 1137This is equivalent to the ``--eth-peer`` command-line option. 1138 1139set port-uta 1140~~~~~~~~~~~~ 1141 1142Set the unicast hash filter(s) on/off for a port:: 1143 1144 testpmd> set port (port_id) uta (XX:XX:XX:XX:XX:XX|all) (on|off) 1145 1146set promisc 1147~~~~~~~~~~~ 1148 1149Set the promiscuous mode on for a port or for all ports. 1150In promiscuous mode packets are not dropped if they aren't for the specified MAC address:: 1151 1152 testpmd> set promisc (port_id|all) (on|off) 1153 1154set allmulti 1155~~~~~~~~~~~~ 1156 1157Set the allmulti mode for a port or for all ports:: 1158 1159 testpmd> set allmulti (port_id|all) (on|off) 1160 1161Same as the ifconfig (8) option. Controls how multicast packets are handled. 1162 1163set promisc (for VF) 1164~~~~~~~~~~~~~~~~~~~~ 1165 1166Set the unicast promiscuous mode for a VF from PF. 1167It's supported by Intel i40e NICs now. 1168In promiscuous mode packets are not dropped if they aren't for the specified MAC address:: 1169 1170 testpmd> set vf promisc (port_id) (vf_id) (on|off) 1171 1172set allmulticast (for VF) 1173~~~~~~~~~~~~~~~~~~~~~~~~~ 1174 1175Set the multicast promiscuous mode for a VF from PF. 1176It's supported by Intel i40e NICs now. 1177In promiscuous mode packets are not dropped if they aren't for the specified MAC address:: 1178 1179 testpmd> set vf allmulti (port_id) (vf_id) (on|off) 1180 1181set tx max bandwidth (for VF) 1182~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ 1183 1184Set TX max absolute bandwidth (Mbps) for a VF from PF:: 1185 1186 testpmd> set vf tx max-bandwidth (port_id) (vf_id) (max_bandwidth) 1187 1188set tc tx min bandwidth (for VF) 1189~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ 1190 1191Set all TCs' TX min relative bandwidth (%) for a VF from PF:: 1192 1193 testpmd> set vf tc tx min-bandwidth (port_id) (vf_id) (bw1, bw2, ...) 1194 1195set tc tx max bandwidth (for VF) 1196~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ 1197 1198Set a TC's TX max absolute bandwidth (Mbps) for a VF from PF:: 1199 1200 testpmd> set vf tc tx max-bandwidth (port_id) (vf_id) (tc_no) (max_bandwidth) 1201 1202set tc strict link priority mode 1203~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ 1204 1205Set some TCs' strict link priority mode on a physical port:: 1206 1207 testpmd> set tx strict-link-priority (port_id) (tc_bitmap) 1208 1209set tc tx min bandwidth 1210~~~~~~~~~~~~~~~~~~~~~~~ 1211 1212Set all TCs' TX min relative bandwidth (%) globally for all PF and VFs:: 1213 1214 testpmd> set tc tx min-bandwidth (port_id) (bw1, bw2, ...) 1215 1216set flow_ctrl rx 1217~~~~~~~~~~~~~~~~ 1218 1219Set the link flow control parameter on a port:: 1220 1221 testpmd> set flow_ctrl rx (on|off) tx (on|off) (high_water) (low_water) \ 1222 (pause_time) (send_xon) mac_ctrl_frame_fwd (on|off) \ 1223 autoneg (on|off) (port_id) 1224 1225Where: 1226 1227* ``high_water`` (integer): High threshold value to trigger XOFF. 1228 1229* ``low_water`` (integer): Low threshold value to trigger XON. 1230 1231* ``pause_time`` (integer): Pause quota in the Pause frame. 1232 1233* ``send_xon`` (0/1): Send XON frame. 1234 1235* ``mac_ctrl_frame_fwd``: Enable receiving MAC control frames. 1236 1237* ``autoneg``: Change the auto-negotiation parameter. 1238 1239set pfc_ctrl rx 1240~~~~~~~~~~~~~~~ 1241 1242Set the priority flow control parameter on a port:: 1243 1244 testpmd> set pfc_ctrl rx (on|off) tx (on|off) (high_water) (low_water) \ 1245 (pause_time) (priority) (port_id) 1246 1247Where: 1248 1249* ``high_water`` (integer): High threshold value. 1250 1251* ``low_water`` (integer): Low threshold value. 1252 1253* ``pause_time`` (integer): Pause quota in the Pause frame. 1254 1255* ``priority`` (0-7): VLAN User Priority. 1256 1257set stat_qmap 1258~~~~~~~~~~~~~ 1259 1260Set statistics mapping (qmapping 0..15) for RX/TX queue on port:: 1261 1262 testpmd> set stat_qmap (tx|rx) (port_id) (queue_id) (qmapping) 1263 1264For example, to set rx queue 2 on port 0 to mapping 5:: 1265 1266 testpmd>set stat_qmap rx 0 2 5 1267 1268set xstats-hide-zero 1269~~~~~~~~~~~~~~~~~~~~ 1270 1271Set the option to hide zero values for xstats display:: 1272 1273 testpmd> set xstats-hide-zero on|off 1274 1275.. note:: 1276 1277 By default, the zero values are displayed for xstats. 1278 1279set port - rx/tx (for VF) 1280~~~~~~~~~~~~~~~~~~~~~~~~~ 1281 1282Set VF receive/transmit from a port:: 1283 1284 testpmd> set port (port_id) vf (vf_id) (rx|tx) (on|off) 1285 1286set port - mac address filter (for VF) 1287~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ 1288 1289Add/Remove unicast or multicast MAC addr filter for a VF:: 1290 1291 testpmd> set port (port_id) vf (vf_id) (mac_addr) \ 1292 (exact-mac|exact-mac-vlan|hashmac|hashmac-vlan) (on|off) 1293 1294set port - rx mode(for VF) 1295~~~~~~~~~~~~~~~~~~~~~~~~~~ 1296 1297Set the VF receive mode of a port:: 1298 1299 testpmd> set port (port_id) vf (vf_id) \ 1300 rxmode (AUPE|ROPE|BAM|MPE) (on|off) 1301 1302The available receive modes are: 1303 1304* ``AUPE``: Accepts untagged VLAN. 1305 1306* ``ROPE``: Accepts unicast hash. 1307 1308* ``BAM``: Accepts broadcast packets. 1309 1310* ``MPE``: Accepts all multicast packets. 1311 1312set port - tx_rate (for Queue) 1313~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ 1314 1315Set TX rate limitation for a queue on a port:: 1316 1317 testpmd> set port (port_id) queue (queue_id) rate (rate_value) 1318 1319set port - tx_rate (for VF) 1320~~~~~~~~~~~~~~~~~~~~~~~~~~~ 1321 1322Set TX rate limitation for queues in VF on a port:: 1323 1324 testpmd> set port (port_id) vf (vf_id) rate (rate_value) queue_mask (queue_mask) 1325 1326set port - mirror rule 1327~~~~~~~~~~~~~~~~~~~~~~ 1328 1329Set pool or vlan type mirror rule for a port:: 1330 1331 testpmd> set port (port_id) mirror-rule (rule_id) \ 1332 (pool-mirror-up|pool-mirror-down|vlan-mirror) \ 1333 (poolmask|vlanid[,vlanid]*) dst-pool (pool_id) (on|off) 1334 1335Set link mirror rule for a port:: 1336 1337 testpmd> set port (port_id) mirror-rule (rule_id) \ 1338 (uplink-mirror|downlink-mirror) dst-pool (pool_id) (on|off) 1339 1340For example to enable mirror traffic with vlan 0,1 to pool 0:: 1341 1342 set port 0 mirror-rule 0 vlan-mirror 0,1 dst-pool 0 on 1343 1344reset port - mirror rule 1345~~~~~~~~~~~~~~~~~~~~~~~~ 1346 1347Reset a mirror rule for a port:: 1348 1349 testpmd> reset port (port_id) mirror-rule (rule_id) 1350 1351set flush_rx 1352~~~~~~~~~~~~ 1353 1354Set the flush on RX streams before forwarding. 1355The default is flush ``on``. 1356Mainly used with PCAP drivers to turn off the default behavior of flushing the first 512 packets on RX streams:: 1357 1358 testpmd> set flush_rx off 1359 1360set bypass mode 1361~~~~~~~~~~~~~~~ 1362 1363Set the bypass mode for the lowest port on bypass enabled NIC:: 1364 1365 testpmd> set bypass mode (normal|bypass|isolate) (port_id) 1366 1367set bypass event 1368~~~~~~~~~~~~~~~~ 1369 1370Set the event required to initiate specified bypass mode for the lowest port on a bypass enabled:: 1371 1372 testpmd> set bypass event (timeout|os_on|os_off|power_on|power_off) \ 1373 mode (normal|bypass|isolate) (port_id) 1374 1375Where: 1376 1377* ``timeout``: Enable bypass after watchdog timeout. 1378 1379* ``os_on``: Enable bypass when OS/board is powered on. 1380 1381* ``os_off``: Enable bypass when OS/board is powered off. 1382 1383* ``power_on``: Enable bypass when power supply is turned on. 1384 1385* ``power_off``: Enable bypass when power supply is turned off. 1386 1387 1388set bypass timeout 1389~~~~~~~~~~~~~~~~~~ 1390 1391Set the bypass watchdog timeout to ``n`` seconds where 0 = instant:: 1392 1393 testpmd> set bypass timeout (0|1.5|2|3|4|8|16|32) 1394 1395show bypass config 1396~~~~~~~~~~~~~~~~~~ 1397 1398Show the bypass configuration for a bypass enabled NIC using the lowest port on the NIC:: 1399 1400 testpmd> show bypass config (port_id) 1401 1402set link up 1403~~~~~~~~~~~ 1404 1405Set link up for a port:: 1406 1407 testpmd> set link-up port (port id) 1408 1409set link down 1410~~~~~~~~~~~~~ 1411 1412Set link down for a port:: 1413 1414 testpmd> set link-down port (port id) 1415 1416E-tag set 1417~~~~~~~~~ 1418 1419Enable E-tag insertion for a VF on a port:: 1420 1421 testpmd> E-tag set insertion on port-tag-id (value) port (port_id) vf (vf_id) 1422 1423Disable E-tag insertion for a VF on a port:: 1424 1425 testpmd> E-tag set insertion off port (port_id) vf (vf_id) 1426 1427Enable/disable E-tag stripping on a port:: 1428 1429 testpmd> E-tag set stripping (on|off) port (port_id) 1430 1431Enable/disable E-tag based forwarding on a port:: 1432 1433 testpmd> E-tag set forwarding (on|off) port (port_id) 1434 1435Add an E-tag forwarding filter on a port:: 1436 1437 testpmd> E-tag set filter add e-tag-id (value) dst-pool (pool_id) port (port_id) 1438 1439Delete an E-tag forwarding filter on a port:: 1440 testpmd> E-tag set filter del e-tag-id (value) port (port_id) 1441 1442ddp add 1443~~~~~~~ 1444 1445Load a dynamic device personalization (DDP) profile and store backup profile:: 1446 1447 testpmd> ddp add (port_id) (profile_path[,backup_profile_path]) 1448 1449ddp del 1450~~~~~~~ 1451 1452Delete a dynamic device personalization profile and restore backup profile:: 1453 1454 testpmd> ddp del (port_id) (backup_profile_path) 1455 1456ptype mapping 1457~~~~~~~~~~~~~ 1458 1459List all items from the ptype mapping table:: 1460 1461 testpmd> ptype mapping get (port_id) (valid_only) 1462 1463Where: 1464 1465* ``valid_only``: A flag indicates if only list valid items(=1) or all itemss(=0). 1466 1467Replace a specific or a group of software defined ptype with a new one:: 1468 1469 testpmd> ptype mapping replace (port_id) (target) (mask) (pkt_type) 1470 1471where: 1472 1473* ``target``: A specific software ptype or a mask to represent a group of software ptypes. 1474 1475* ``mask``: A flag indicate if "target" is a specific software ptype(=0) or a ptype mask(=1). 1476 1477* ``pkt_type``: The new software ptype to replace the old ones. 1478 1479Update hardware defined ptype to software defined packet type mapping table:: 1480 1481 testpmd> ptype mapping update (port_id) (hw_ptype) (sw_ptype) 1482 1483where: 1484 1485* ``hw_ptype``: hardware ptype as the index of the ptype mapping table. 1486 1487* ``sw_ptype``: software ptype as the value of the ptype mapping table. 1488 1489Reset ptype mapping table:: 1490 1491 testpmd> ptype mapping reset (port_id) 1492 1493config per port Rx offloading 1494~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ 1495 1496Enable or disable a per port Rx offloading on all Rx queues of a port:: 1497 1498 testpmd> port config (port_id) rx_offload (offloading) on|off 1499 1500* ``offloading``: can be any of these offloading capability: 1501 vlan_strip, ipv4_cksum, udp_cksum, tcp_cksum, tcp_lro, 1502 qinq_strip, outer_ipv4_cksum, macsec_strip, 1503 header_split, vlan_filter, vlan_extend, jumbo_frame, 1504 crc_strip, scatter, timestamp, security, keep_crc 1505 1506This command should be run when the port is stopped, or else it will fail. 1507 1508config per queue Rx offloading 1509~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ 1510 1511Enable or disable a per queue Rx offloading only on a specific Rx queue:: 1512 1513 testpmd> port (port_id) rxq (queue_id) rx_offload (offloading) on|off 1514 1515* ``offloading``: can be any of these offloading capability: 1516 vlan_strip, ipv4_cksum, udp_cksum, tcp_cksum, tcp_lro, 1517 qinq_strip, outer_ipv4_cksum, macsec_strip, 1518 header_split, vlan_filter, vlan_extend, jumbo_frame, 1519 crc_strip, scatter, timestamp, security, keep_crc 1520 1521This command should be run when the port is stopped, or else it will fail. 1522 1523config per port Tx offloading 1524~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ 1525 1526Enable or disable a per port Tx offloading on all Tx queues of a port:: 1527 1528 testpmd> port config (port_id) tx_offload (offloading) on|off 1529 1530* ``offloading``: can be any of these offloading capability: 1531 vlan_insert, ipv4_cksum, udp_cksum, tcp_cksum, 1532 sctp_cksum, tcp_tso, udp_tso, outer_ipv4_cksum, 1533 qinq_insert, vxlan_tnl_tso, gre_tnl_tso, 1534 ipip_tnl_tso, geneve_tnl_tso, macsec_insert, 1535 mt_lockfree, multi_segs, mbuf_fast_free, security 1536 1537This command should be run when the port is stopped, or else it will fail. 1538 1539config per queue Tx offloading 1540~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ 1541 1542Enable or disable a per queue Tx offloading only on a specific Tx queue:: 1543 1544 testpmd> port (port_id) txq (queue_id) tx_offload (offloading) on|off 1545 1546* ``offloading``: can be any of these offloading capability: 1547 vlan_insert, ipv4_cksum, udp_cksum, tcp_cksum, 1548 sctp_cksum, tcp_tso, udp_tso, outer_ipv4_cksum, 1549 qinq_insert, vxlan_tnl_tso, gre_tnl_tso, 1550 ipip_tnl_tso, geneve_tnl_tso, macsec_insert, 1551 mt_lockfree, multi_segs, mbuf_fast_free, security 1552 1553This command should be run when the port is stopped, or else it will fail. 1554 1555Config VXLAN Encap outer layers 1556~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ 1557 1558Configure the outer layer to encapsulate a packet inside a VXLAN tunnel:: 1559 1560 set vxlan ip-version (ipv4|ipv6) vni (vni) udp-src (udp-src) \ 1561 udp-dst (udp-dst) ip-src (ip-src) ip-dst (ip-dst) eth-src (eth-src) \ 1562 eth-dst (eth-dst) 1563 1564 set vxlan-with-vlan ip-version (ipv4|ipv6) vni (vni) udp-src (udp-src) \ 1565 udp-dst (udp-dst) ip-src (ip-src) ip-dst (ip-dst) vlan-tci (vlan-tci) \ 1566 eth-src (eth-src) eth-dst (eth-dst) 1567 1568Those command will set an internal configuration inside testpmd, any following 1569flow rule using the action vxlan_encap will use the last configuration set. 1570To have a different encapsulation header, one of those commands must be called 1571before the flow rule creation. 1572 1573Config NVGRE Encap outer layers 1574~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ 1575 1576Configure the outer layer to encapsulate a packet inside a NVGRE tunnel:: 1577 1578 set nvgre ip-version (ipv4|ipv6) tni (tni) ip-src (ip-src) ip-dst (ip-dst) \ 1579 eth-src (eth-src) eth-dst (eth-dst) 1580 set nvgre-with-vlan ip-version (ipv4|ipv6) tni (tni) ip-src (ip-src) \ 1581 ip-dst (ip-dst) vlan-tci (vlan-tci) eth-src (eth-src) eth-dst (eth-dst) 1582 1583Those command will set an internal configuration inside testpmd, any following 1584flow rule using the action nvgre_encap will use the last configuration set. 1585To have a different encapsulation header, one of those commands must be called 1586before the flow rule creation. 1587 1588Port Functions 1589-------------- 1590 1591The following sections show functions for configuring ports. 1592 1593.. note:: 1594 1595 Port configuration changes only become active when forwarding is started/restarted. 1596 1597port attach 1598~~~~~~~~~~~ 1599 1600Attach a port specified by pci address or virtual device args:: 1601 1602 testpmd> port attach (identifier) 1603 1604To attach a new pci device, the device should be recognized by kernel first. 1605Then it should be moved under DPDK management. 1606Finally the port can be attached to testpmd. 1607 1608For example, to move a pci device using ixgbe under DPDK management: 1609 1610.. code-block:: console 1611 1612 # Check the status of the available devices. 1613 ./usertools/dpdk-devbind.py --status 1614 1615 Network devices using DPDK-compatible driver 1616 ============================================ 1617 <none> 1618 1619 Network devices using kernel driver 1620 =================================== 1621 0000:0a:00.0 '82599ES 10-Gigabit' if=eth2 drv=ixgbe unused= 1622 1623 1624 # Bind the device to igb_uio. 1625 sudo ./usertools/dpdk-devbind.py -b igb_uio 0000:0a:00.0 1626 1627 1628 # Recheck the status of the devices. 1629 ./usertools/dpdk-devbind.py --status 1630 Network devices using DPDK-compatible driver 1631 ============================================ 1632 0000:0a:00.0 '82599ES 10-Gigabit' drv=igb_uio unused= 1633 1634To attach a port created by virtual device, above steps are not needed. 1635 1636For example, to attach a port whose pci address is 0000:0a:00.0. 1637 1638.. code-block:: console 1639 1640 testpmd> port attach 0000:0a:00.0 1641 Attaching a new port... 1642 EAL: PCI device 0000:0a:00.0 on NUMA socket -1 1643 EAL: probe driver: 8086:10fb rte_ixgbe_pmd 1644 EAL: PCI memory mapped at 0x7f83bfa00000 1645 EAL: PCI memory mapped at 0x7f83bfa80000 1646 PMD: eth_ixgbe_dev_init(): MAC: 2, PHY: 18, SFP+: 5 1647 PMD: eth_ixgbe_dev_init(): port 0 vendorID=0x8086 deviceID=0x10fb 1648 Port 0 is attached. Now total ports is 1 1649 Done 1650 1651For example, to attach a port created by pcap PMD. 1652 1653.. code-block:: console 1654 1655 testpmd> port attach net_pcap0 1656 Attaching a new port... 1657 PMD: Initializing pmd_pcap for net_pcap0 1658 PMD: Creating pcap-backed ethdev on numa socket 0 1659 Port 0 is attached. Now total ports is 1 1660 Done 1661 1662In this case, identifier is ``net_pcap0``. 1663This identifier format is the same as ``--vdev`` format of DPDK applications. 1664 1665For example, to re-attach a bonded port which has been previously detached, 1666the mode and slave parameters must be given. 1667 1668.. code-block:: console 1669 1670 testpmd> port attach net_bond_0,mode=0,slave=1 1671 Attaching a new port... 1672 EAL: Initializing pmd_bond for net_bond_0 1673 EAL: Create bonded device net_bond_0 on port 0 in mode 0 on socket 0. 1674 Port 0 is attached. Now total ports is 1 1675 Done 1676 1677 1678port detach 1679~~~~~~~~~~~ 1680 1681Detach a specific port:: 1682 1683 testpmd> port detach (port_id) 1684 1685Before detaching a port, the port should be stopped and closed. 1686 1687For example, to detach a pci device port 0. 1688 1689.. code-block:: console 1690 1691 testpmd> port stop 0 1692 Stopping ports... 1693 Done 1694 testpmd> port close 0 1695 Closing ports... 1696 Done 1697 1698 testpmd> port detach 0 1699 Detaching a port... 1700 EAL: PCI device 0000:0a:00.0 on NUMA socket -1 1701 EAL: remove driver: 8086:10fb rte_ixgbe_pmd 1702 EAL: PCI memory unmapped at 0x7f83bfa00000 1703 EAL: PCI memory unmapped at 0x7f83bfa80000 1704 Done 1705 1706 1707For example, to detach a virtual device port 0. 1708 1709.. code-block:: console 1710 1711 testpmd> port stop 0 1712 Stopping ports... 1713 Done 1714 testpmd> port close 0 1715 Closing ports... 1716 Done 1717 1718 testpmd> port detach 0 1719 Detaching a port... 1720 PMD: Closing pcap ethdev on numa socket 0 1721 Port 'net_pcap0' is detached. Now total ports is 0 1722 Done 1723 1724To remove a pci device completely from the system, first detach the port from testpmd. 1725Then the device should be moved under kernel management. 1726Finally the device can be removed using kernel pci hotplug functionality. 1727 1728For example, to move a pci device under kernel management: 1729 1730.. code-block:: console 1731 1732 sudo ./usertools/dpdk-devbind.py -b ixgbe 0000:0a:00.0 1733 1734 ./usertools/dpdk-devbind.py --status 1735 1736 Network devices using DPDK-compatible driver 1737 ============================================ 1738 <none> 1739 1740 Network devices using kernel driver 1741 =================================== 1742 0000:0a:00.0 '82599ES 10-Gigabit' if=eth2 drv=ixgbe unused=igb_uio 1743 1744To remove a port created by a virtual device, above steps are not needed. 1745 1746port start 1747~~~~~~~~~~ 1748 1749Start all ports or a specific port:: 1750 1751 testpmd> port start (port_id|all) 1752 1753port stop 1754~~~~~~~~~ 1755 1756Stop all ports or a specific port:: 1757 1758 testpmd> port stop (port_id|all) 1759 1760port close 1761~~~~~~~~~~ 1762 1763Close all ports or a specific port:: 1764 1765 testpmd> port close (port_id|all) 1766 1767port config - queue ring size 1768~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ 1769 1770Configure a rx/tx queue ring size:: 1771 1772 testpmd> port (port_id) (rxq|txq) (queue_id) ring_size (value) 1773 1774Only take effect after command that (re-)start the port or command that setup specific queue. 1775 1776port start/stop queue 1777~~~~~~~~~~~~~~~~~~~~~ 1778 1779Start/stop a rx/tx queue on a specific port:: 1780 1781 testpmd> port (port_id) (rxq|txq) (queue_id) (start|stop) 1782 1783port config - queue deferred start 1784~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ 1785 1786Switch on/off deferred start of a specific port queue:: 1787 1788 testpmd> port (port_id) (rxq|txq) (queue_id) deferred_start (on|off) 1789 1790port setup queue 1791~~~~~~~~~~~~~~~~~~~~~ 1792 1793Setup a rx/tx queue on a specific port:: 1794 1795 testpmd> port (port_id) (rxq|txq) (queue_id) setup 1796 1797Only take effect when port is started. 1798 1799port config - speed 1800~~~~~~~~~~~~~~~~~~~ 1801 1802Set the speed and duplex mode for all ports or a specific port:: 1803 1804 testpmd> port config (port_id|all) speed (10|100|1000|10000|25000|40000|50000|100000|auto) \ 1805 duplex (half|full|auto) 1806 1807port config - queues/descriptors 1808~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ 1809 1810Set number of queues/descriptors for rxq, txq, rxd and txd:: 1811 1812 testpmd> port config all (rxq|txq|rxd|txd) (value) 1813 1814This is equivalent to the ``--rxq``, ``--txq``, ``--rxd`` and ``--txd`` command-line options. 1815 1816port config - max-pkt-len 1817~~~~~~~~~~~~~~~~~~~~~~~~~ 1818 1819Set the maximum packet length:: 1820 1821 testpmd> port config all max-pkt-len (value) 1822 1823This is equivalent to the ``--max-pkt-len`` command-line option. 1824 1825port config - CRC Strip 1826~~~~~~~~~~~~~~~~~~~~~~~ 1827 1828Set hardware CRC stripping on or off for all ports:: 1829 1830 testpmd> port config all crc-strip (on|off) 1831 1832CRC stripping is on by default. 1833 1834The ``off`` option is equivalent to the ``--disable-crc-strip`` command-line option. 1835 1836port config - scatter 1837~~~~~~~~~~~~~~~~~~~~~~~ 1838 1839Set RX scatter mode on or off for all ports:: 1840 1841 testpmd> port config all scatter (on|off) 1842 1843RX scatter mode is off by default. 1844 1845The ``on`` option is equivalent to the ``--enable-scatter`` command-line option. 1846 1847port config - RX Checksum 1848~~~~~~~~~~~~~~~~~~~~~~~~~ 1849 1850Set hardware RX checksum offload to on or off for all ports:: 1851 1852 testpmd> port config all rx-cksum (on|off) 1853 1854Checksum offload is off by default. 1855 1856The ``on`` option is equivalent to the ``--enable-rx-cksum`` command-line option. 1857 1858port config - VLAN 1859~~~~~~~~~~~~~~~~~~ 1860 1861Set hardware VLAN on or off for all ports:: 1862 1863 testpmd> port config all hw-vlan (on|off) 1864 1865Hardware VLAN is off by default. 1866 1867The ``on`` option is equivalent to the ``--enable-hw-vlan`` command-line option. 1868 1869port config - VLAN filter 1870~~~~~~~~~~~~~~~~~~~~~~~~~ 1871 1872Set hardware VLAN filter on or off for all ports:: 1873 1874 testpmd> port config all hw-vlan-filter (on|off) 1875 1876Hardware VLAN filter is off by default. 1877 1878The ``on`` option is equivalent to the ``--enable-hw-vlan-filter`` command-line option. 1879 1880port config - VLAN strip 1881~~~~~~~~~~~~~~~~~~~~~~~~ 1882 1883Set hardware VLAN strip on or off for all ports:: 1884 1885 testpmd> port config all hw-vlan-strip (on|off) 1886 1887Hardware VLAN strip is off by default. 1888 1889The ``on`` option is equivalent to the ``--enable-hw-vlan-strip`` command-line option. 1890 1891port config - VLAN extend 1892~~~~~~~~~~~~~~~~~~~~~~~~~ 1893 1894Set hardware VLAN extend on or off for all ports:: 1895 1896 testpmd> port config all hw-vlan-extend (on|off) 1897 1898Hardware VLAN extend is off by default. 1899 1900The ``on`` option is equivalent to the ``--enable-hw-vlan-extend`` command-line option. 1901 1902port config - Drop Packets 1903~~~~~~~~~~~~~~~~~~~~~~~~~~ 1904 1905Set packet drop for packets with no descriptors on or off for all ports:: 1906 1907 testpmd> port config all drop-en (on|off) 1908 1909Packet dropping for packets with no descriptors is off by default. 1910 1911The ``on`` option is equivalent to the ``--enable-drop-en`` command-line option. 1912 1913port config - RSS 1914~~~~~~~~~~~~~~~~~ 1915 1916Set the RSS (Receive Side Scaling) mode on or off:: 1917 1918 testpmd> port config all rss (all|default|ip|tcp|udp|sctp|ether|port|vxlan|geneve|nvgre|none) 1919 1920RSS is on by default. 1921 1922The ``all`` option is equivalent to ip|tcp|udp|sctp|ether. 1923The ``default`` option enables all supported RSS types reported by device info. 1924The ``none`` option is equivalent to the ``--disable-rss`` command-line option. 1925 1926port config - RSS Reta 1927~~~~~~~~~~~~~~~~~~~~~~ 1928 1929Set the RSS (Receive Side Scaling) redirection table:: 1930 1931 testpmd> port config all rss reta (hash,queue)[,(hash,queue)] 1932 1933port config - DCB 1934~~~~~~~~~~~~~~~~~ 1935 1936Set the DCB mode for an individual port:: 1937 1938 testpmd> port config (port_id) dcb vt (on|off) (traffic_class) pfc (on|off) 1939 1940The traffic class should be 4 or 8. 1941 1942port config - Burst 1943~~~~~~~~~~~~~~~~~~~ 1944 1945Set the number of packets per burst:: 1946 1947 testpmd> port config all burst (value) 1948 1949This is equivalent to the ``--burst`` command-line option. 1950 1951port config - Threshold 1952~~~~~~~~~~~~~~~~~~~~~~~ 1953 1954Set thresholds for TX/RX queues:: 1955 1956 testpmd> port config all (threshold) (value) 1957 1958Where the threshold type can be: 1959 1960* ``txpt:`` Set the prefetch threshold register of the TX rings, 0 <= value <= 255. 1961 1962* ``txht:`` Set the host threshold register of the TX rings, 0 <= value <= 255. 1963 1964* ``txwt:`` Set the write-back threshold register of the TX rings, 0 <= value <= 255. 1965 1966* ``rxpt:`` Set the prefetch threshold register of the RX rings, 0 <= value <= 255. 1967 1968* ``rxht:`` Set the host threshold register of the RX rings, 0 <= value <= 255. 1969 1970* ``rxwt:`` Set the write-back threshold register of the RX rings, 0 <= value <= 255. 1971 1972* ``txfreet:`` Set the transmit free threshold of the TX rings, 0 <= value <= txd. 1973 1974* ``rxfreet:`` Set the transmit free threshold of the RX rings, 0 <= value <= rxd. 1975 1976* ``txrst:`` Set the transmit RS bit threshold of TX rings, 0 <= value <= txd. 1977 1978These threshold options are also available from the command-line. 1979 1980port config - E-tag 1981~~~~~~~~~~~~~~~~~~~ 1982 1983Set the value of ether-type for E-tag:: 1984 1985 testpmd> port config (port_id|all) l2-tunnel E-tag ether-type (value) 1986 1987Enable/disable the E-tag support:: 1988 1989 testpmd> port config (port_id|all) l2-tunnel E-tag (enable|disable) 1990 1991port config pctype mapping 1992~~~~~~~~~~~~~~~~~~~~~~~~~~ 1993 1994Reset pctype mapping table:: 1995 1996 testpmd> port config (port_id) pctype mapping reset 1997 1998Update hardware defined pctype to software defined flow type mapping table:: 1999 2000 testpmd> port config (port_id) pctype mapping update (pctype_id_0[,pctype_id_1]*) (flow_type_id) 2001 2002where: 2003 2004* ``pctype_id_x``: hardware pctype id as index of bit in bitmask value of the pctype mapping table. 2005 2006* ``flow_type_id``: software flow type id as the index of the pctype mapping table. 2007 2008port config input set 2009~~~~~~~~~~~~~~~~~~~~~ 2010 2011Config RSS/FDIR/FDIR flexible payload input set for some pctype:: 2012 testpmd> port config (port_id) pctype (pctype_id) \ 2013 (hash_inset|fdir_inset|fdir_flx_inset) \ 2014 (get|set|clear) field (field_idx) 2015 2016Clear RSS/FDIR/FDIR flexible payload input set for some pctype:: 2017 testpmd> port config (port_id) pctype (pctype_id) \ 2018 (hash_inset|fdir_inset|fdir_flx_inset) clear all 2019 2020where: 2021 2022* ``pctype_id``: hardware packet classification types. 2023* ``field_idx``: hardware field index. 2024 2025port config udp_tunnel_port 2026~~~~~~~~~~~~~~~~~~~~~~~~~~~ 2027 2028Add/remove UDP tunnel port for VXLAN/GENEVE tunneling protocols:: 2029 testpmd> port config (port_id) udp_tunnel_port add|rm vxlan|geneve (udp_port) 2030 2031Link Bonding Functions 2032---------------------- 2033 2034The Link Bonding functions make it possible to dynamically create and 2035manage link bonding devices from within testpmd interactive prompt. 2036 2037create bonded device 2038~~~~~~~~~~~~~~~~~~~~ 2039 2040Create a new bonding device:: 2041 2042 testpmd> create bonded device (mode) (socket) 2043 2044For example, to create a bonded device in mode 1 on socket 0:: 2045 2046 testpmd> create bonded device 1 0 2047 created new bonded device (port X) 2048 2049add bonding slave 2050~~~~~~~~~~~~~~~~~ 2051 2052Adds Ethernet device to a Link Bonding device:: 2053 2054 testpmd> add bonding slave (slave id) (port id) 2055 2056For example, to add Ethernet device (port 6) to a Link Bonding device (port 10):: 2057 2058 testpmd> add bonding slave 6 10 2059 2060 2061remove bonding slave 2062~~~~~~~~~~~~~~~~~~~~ 2063 2064Removes an Ethernet slave device from a Link Bonding device:: 2065 2066 testpmd> remove bonding slave (slave id) (port id) 2067 2068For example, to remove Ethernet slave device (port 6) to a Link Bonding device (port 10):: 2069 2070 testpmd> remove bonding slave 6 10 2071 2072set bonding mode 2073~~~~~~~~~~~~~~~~ 2074 2075Set the Link Bonding mode of a Link Bonding device:: 2076 2077 testpmd> set bonding mode (value) (port id) 2078 2079For example, to set the bonding mode of a Link Bonding device (port 10) to broadcast (mode 3):: 2080 2081 testpmd> set bonding mode 3 10 2082 2083set bonding primary 2084~~~~~~~~~~~~~~~~~~~ 2085 2086Set an Ethernet slave device as the primary device on a Link Bonding device:: 2087 2088 testpmd> set bonding primary (slave id) (port id) 2089 2090For example, to set the Ethernet slave device (port 6) as the primary port of a Link Bonding device (port 10):: 2091 2092 testpmd> set bonding primary 6 10 2093 2094set bonding mac 2095~~~~~~~~~~~~~~~ 2096 2097Set the MAC address of a Link Bonding device:: 2098 2099 testpmd> set bonding mac (port id) (mac) 2100 2101For example, to set the MAC address of a Link Bonding device (port 10) to 00:00:00:00:00:01:: 2102 2103 testpmd> set bonding mac 10 00:00:00:00:00:01 2104 2105set bonding xmit_balance_policy 2106~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ 2107 2108Set the transmission policy for a Link Bonding device when it is in Balance XOR mode:: 2109 2110 testpmd> set bonding xmit_balance_policy (port_id) (l2|l23|l34) 2111 2112For example, set a Link Bonding device (port 10) to use a balance policy of layer 3+4 (IP addresses & UDP ports):: 2113 2114 testpmd> set bonding xmit_balance_policy 10 l34 2115 2116 2117set bonding mon_period 2118~~~~~~~~~~~~~~~~~~~~~~ 2119 2120Set the link status monitoring polling period in milliseconds for a bonding device. 2121 2122This adds support for PMD slave devices which do not support link status interrupts. 2123When the mon_period is set to a value greater than 0 then all PMD's which do not support 2124link status ISR will be queried every polling interval to check if their link status has changed:: 2125 2126 testpmd> set bonding mon_period (port_id) (value) 2127 2128For example, to set the link status monitoring polling period of bonded device (port 5) to 150ms:: 2129 2130 testpmd> set bonding mon_period 5 150 2131 2132 2133set bonding lacp dedicated_queue 2134~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ 2135 2136Enable dedicated tx/rx queues on bonding devices slaves to handle LACP control plane traffic 2137when in mode 4 (link-aggregration-802.3ad):: 2138 2139 testpmd> set bonding lacp dedicated_queues (port_id) (enable|disable) 2140 2141 2142set bonding agg_mode 2143~~~~~~~~~~~~~~~~~~~~ 2144 2145Enable one of the specific aggregators mode when in mode 4 (link-aggregration-802.3ad):: 2146 2147 testpmd> set bonding agg_mode (port_id) (bandwidth|count|stable) 2148 2149 2150show bonding config 2151~~~~~~~~~~~~~~~~~~~ 2152 2153Show the current configuration of a Link Bonding device:: 2154 2155 testpmd> show bonding config (port id) 2156 2157For example, 2158to show the configuration a Link Bonding device (port 9) with 3 slave devices (1, 3, 4) 2159in balance mode with a transmission policy of layer 2+3:: 2160 2161 testpmd> show bonding config 9 2162 Bonding mode: 2 2163 Balance Xmit Policy: BALANCE_XMIT_POLICY_LAYER23 2164 Slaves (3): [1 3 4] 2165 Active Slaves (3): [1 3 4] 2166 Primary: [3] 2167 2168 2169Register Functions 2170------------------ 2171 2172The Register Functions can be used to read from and write to registers on the network card referenced by a port number. 2173This is mainly useful for debugging purposes. 2174Reference should be made to the appropriate datasheet for the network card for details on the register addresses 2175and fields that can be accessed. 2176 2177read reg 2178~~~~~~~~ 2179 2180Display the value of a port register:: 2181 2182 testpmd> read reg (port_id) (address) 2183 2184For example, to examine the Flow Director control register (FDIRCTL, 0x0000EE000) on an Intel 82599 10 GbE Controller:: 2185 2186 testpmd> read reg 0 0xEE00 2187 port 0 PCI register at offset 0xEE00: 0x4A060029 (1241907241) 2188 2189read regfield 2190~~~~~~~~~~~~~ 2191 2192Display a port register bit field:: 2193 2194 testpmd> read regfield (port_id) (address) (bit_x) (bit_y) 2195 2196For example, reading the lowest two bits from the register in the example above:: 2197 2198 testpmd> read regfield 0 0xEE00 0 1 2199 port 0 PCI register at offset 0xEE00: bits[0, 1]=0x1 (1) 2200 2201read regbit 2202~~~~~~~~~~~ 2203 2204Display a single port register bit:: 2205 2206 testpmd> read regbit (port_id) (address) (bit_x) 2207 2208For example, reading the lowest bit from the register in the example above:: 2209 2210 testpmd> read regbit 0 0xEE00 0 2211 port 0 PCI register at offset 0xEE00: bit 0=1 2212 2213write reg 2214~~~~~~~~~ 2215 2216Set the value of a port register:: 2217 2218 testpmd> write reg (port_id) (address) (value) 2219 2220For example, to clear a register:: 2221 2222 testpmd> write reg 0 0xEE00 0x0 2223 port 0 PCI register at offset 0xEE00: 0x00000000 (0) 2224 2225write regfield 2226~~~~~~~~~~~~~~ 2227 2228Set bit field of a port register:: 2229 2230 testpmd> write regfield (port_id) (address) (bit_x) (bit_y) (value) 2231 2232For example, writing to the register cleared in the example above:: 2233 2234 testpmd> write regfield 0 0xEE00 0 1 2 2235 port 0 PCI register at offset 0xEE00: 0x00000002 (2) 2236 2237write regbit 2238~~~~~~~~~~~~ 2239 2240Set single bit value of a port register:: 2241 2242 testpmd> write regbit (port_id) (address) (bit_x) (value) 2243 2244For example, to set the high bit in the register from the example above:: 2245 2246 testpmd> write regbit 0 0xEE00 31 1 2247 port 0 PCI register at offset 0xEE00: 0x8000000A (2147483658) 2248 2249Traffic Metering and Policing 2250----------------------------- 2251 2252The following section shows functions for configuring traffic metering and 2253policing on the ethernet device through the use of generic ethdev API. 2254 2255show port traffic management capability 2256~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ 2257 2258Show traffic metering and policing capability of the port:: 2259 2260 testpmd> show port meter cap (port_id) 2261 2262add port meter profile (srTCM rfc2967) 2263~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ 2264 2265Add meter profile (srTCM rfc2697) to the ethernet device:: 2266 2267 testpmd> add port meter profile srtcm_rfc2697 (port_id) (profile_id) \ 2268 (cir) (cbs) (ebs) 2269 2270where: 2271 2272* ``profile_id``: ID for the meter profile. 2273* ``cir``: Committed Information Rate (CIR) (bytes/second). 2274* ``cbs``: Committed Burst Size (CBS) (bytes). 2275* ``ebs``: Excess Burst Size (EBS) (bytes). 2276 2277add port meter profile (trTCM rfc2968) 2278~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ 2279 2280Add meter profile (srTCM rfc2698) to the ethernet device:: 2281 2282 testpmd> add port meter profile trtcm_rfc2698 (port_id) (profile_id) \ 2283 (cir) (pir) (cbs) (pbs) 2284 2285where: 2286 2287* ``profile_id``: ID for the meter profile. 2288* ``cir``: Committed information rate (bytes/second). 2289* ``pir``: Peak information rate (bytes/second). 2290* ``cbs``: Committed burst size (bytes). 2291* ``pbs``: Peak burst size (bytes). 2292 2293add port meter profile (trTCM rfc4115) 2294~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ 2295 2296Add meter profile (trTCM rfc4115) to the ethernet device:: 2297 2298 testpmd> add port meter profile trtcm_rfc4115 (port_id) (profile_id) \ 2299 (cir) (eir) (cbs) (ebs) 2300 2301where: 2302 2303* ``profile_id``: ID for the meter profile. 2304* ``cir``: Committed information rate (bytes/second). 2305* ``eir``: Excess information rate (bytes/second). 2306* ``cbs``: Committed burst size (bytes). 2307* ``ebs``: Excess burst size (bytes). 2308 2309delete port meter profile 2310~~~~~~~~~~~~~~~~~~~~~~~~~ 2311 2312Delete meter profile from the ethernet device:: 2313 2314 testpmd> del port meter profile (port_id) (profile_id) 2315 2316create port meter 2317~~~~~~~~~~~~~~~~~ 2318 2319Create new meter object for the ethernet device:: 2320 2321 testpmd> create port meter (port_id) (mtr_id) (profile_id) \ 2322 (meter_enable) (g_action) (y_action) (r_action) (stats_mask) (shared) \ 2323 (use_pre_meter_color) [(dscp_tbl_entry0) (dscp_tbl_entry1)...\ 2324 (dscp_tbl_entry63)] 2325 2326where: 2327 2328* ``mtr_id``: meter object ID. 2329* ``profile_id``: ID for the meter profile. 2330* ``meter_enable``: When this parameter has a non-zero value, the meter object 2331 gets enabled at the time of creation, otherwise remains disabled. 2332* ``g_action``: Policer action for the packet with green color. 2333* ``y_action``: Policer action for the packet with yellow color. 2334* ``r_action``: Policer action for the packet with red color. 2335* ``stats_mask``: Mask of statistics counter types to be enabled for the 2336 meter object. 2337* ``shared``: When this parameter has a non-zero value, the meter object is 2338 shared by multiple flows. Otherwise, meter object is used by single flow. 2339* ``use_pre_meter_color``: When this parameter has a non-zero value, the 2340 input color for the current meter object is determined by the latest meter 2341 object in the same flow. Otherwise, the current meter object uses the 2342 *dscp_table* to determine the input color. 2343* ``dscp_tbl_entryx``: DSCP table entry x providing meter providing input 2344 color, 0 <= x <= 63. 2345 2346enable port meter 2347~~~~~~~~~~~~~~~~~ 2348 2349Enable meter for the ethernet device:: 2350 2351 testpmd> enable port meter (port_id) (mtr_id) 2352 2353disable port meter 2354~~~~~~~~~~~~~~~~~~ 2355 2356Disable meter for the ethernet device:: 2357 2358 testpmd> disable port meter (port_id) (mtr_id) 2359 2360delete port meter 2361~~~~~~~~~~~~~~~~~ 2362 2363Delete meter for the ethernet device:: 2364 2365 testpmd> del port meter (port_id) (mtr_id) 2366 2367Set port meter profile 2368~~~~~~~~~~~~~~~~~~~~~~ 2369 2370Set meter profile for the ethernet device:: 2371 2372 testpmd> set port meter profile (port_id) (mtr_id) (profile_id) 2373 2374set port meter dscp table 2375~~~~~~~~~~~~~~~~~~~~~~~~~ 2376 2377Set meter dscp table for the ethernet device:: 2378 2379 testpmd> set port meter dscp table (port_id) (mtr_id) [(dscp_tbl_entry0) \ 2380 (dscp_tbl_entry1)...(dscp_tbl_entry63)] 2381 2382set port meter policer action 2383~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ 2384 2385Set meter policer action for the ethernet device:: 2386 2387 testpmd> set port meter policer action (port_id) (mtr_id) (action_mask) \ 2388 (action0) [(action1) (action1)] 2389 2390where: 2391 2392* ``action_mask``: Bit mask indicating which policer actions need to be 2393 updated. One or more policer actions can be updated in a single function 2394 invocation. To update the policer action associated with color C, bit 2395 (1 << C) needs to be set in *action_mask* and element at position C 2396 in the *actions* array needs to be valid. 2397* ``actionx``: Policer action for the color x, 2398 RTE_MTR_GREEN <= x < RTE_MTR_COLORS 2399 2400set port meter stats mask 2401~~~~~~~~~~~~~~~~~~~~~~~~~ 2402 2403Set meter stats mask for the ethernet device:: 2404 2405 testpmd> set port meter stats mask (port_id) (mtr_id) (stats_mask) 2406 2407where: 2408 2409* ``stats_mask``: Bit mask indicating statistics counter types to be enabled. 2410 2411show port meter stats 2412~~~~~~~~~~~~~~~~~~~~~ 2413 2414Show meter stats of the ethernet device:: 2415 2416 testpmd> show port meter stats (port_id) (mtr_id) (clear) 2417 2418where: 2419 2420* ``clear``: Flag that indicates whether the statistics counters should 2421 be cleared (i.e. set to zero) immediately after they have been read or not. 2422 2423Traffic Management 2424------------------ 2425 2426The following section shows functions for configuring traffic management on 2427on the ethernet device through the use of generic TM API. 2428 2429show port traffic management capability 2430~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ 2431 2432Show traffic management capability of the port:: 2433 2434 testpmd> show port tm cap (port_id) 2435 2436show port traffic management capability (hierarchy level) 2437~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ 2438 2439Show traffic management hierarchy level capability of the port:: 2440 2441 testpmd> show port tm level cap (port_id) (level_id) 2442 2443show port traffic management capability (hierarchy node level) 2444~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ 2445 2446Show the traffic management hierarchy node capability of the port:: 2447 2448 testpmd> show port tm node cap (port_id) (node_id) 2449 2450show port traffic management hierarchy node type 2451~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ 2452 2453Show the port traffic management hierarchy node type:: 2454 2455 testpmd> show port tm node type (port_id) (node_id) 2456 2457show port traffic management hierarchy node stats 2458~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ 2459 2460Show the port traffic management hierarchy node statistics:: 2461 2462 testpmd> show port tm node stats (port_id) (node_id) (clear) 2463 2464where: 2465 2466* ``clear``: When this parameter has a non-zero value, the statistics counters 2467 are cleared (i.e. set to zero) immediately after they have been read, 2468 otherwise the statistics counters are left untouched. 2469 2470Add port traffic management private shaper profile 2471~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ 2472 2473Add the port traffic management private shaper profile:: 2474 2475 testpmd> add port tm node shaper profile (port_id) (shaper_profile_id) \ 2476 (tb_rate) (tb_size) (packet_length_adjust) 2477 2478where: 2479 2480* ``shaper_profile id``: Shaper profile ID for the new profile. 2481* ``tb_rate``: Token bucket rate (bytes per second). 2482* ``tb_size``: Token bucket size (bytes). 2483* ``packet_length_adjust``: The value (bytes) to be added to the length of 2484 each packet for the purpose of shaping. This parameter value can be used to 2485 correct the packet length with the framing overhead bytes that are consumed 2486 on the wire. 2487 2488Delete port traffic management private shaper profile 2489~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ 2490 2491Delete the port traffic management private shaper:: 2492 2493 testpmd> del port tm node shaper profile (port_id) (shaper_profile_id) 2494 2495where: 2496 2497* ``shaper_profile id``: Shaper profile ID that needs to be deleted. 2498 2499Add port traffic management shared shaper 2500~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ 2501 2502Create the port traffic management shared shaper:: 2503 2504 testpmd> add port tm node shared shaper (port_id) (shared_shaper_id) \ 2505 (shaper_profile_id) 2506 2507where: 2508 2509* ``shared_shaper_id``: Shared shaper ID to be created. 2510* ``shaper_profile id``: Shaper profile ID for shared shaper. 2511 2512Set port traffic management shared shaper 2513~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ 2514 2515Update the port traffic management shared shaper:: 2516 2517 testpmd> set port tm node shared shaper (port_id) (shared_shaper_id) \ 2518 (shaper_profile_id) 2519 2520where: 2521 2522* ``shared_shaper_id``: Shared shaper ID to be update. 2523* ``shaper_profile id``: Shaper profile ID for shared shaper. 2524 2525Delete port traffic management shared shaper 2526~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ 2527 2528Delete the port traffic management shared shaper:: 2529 2530 testpmd> del port tm node shared shaper (port_id) (shared_shaper_id) 2531 2532where: 2533 2534* ``shared_shaper_id``: Shared shaper ID to be deleted. 2535 2536Set port traffic management hiearchy node private shaper 2537~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ 2538 2539set the port traffic management hierarchy node private shaper:: 2540 2541 testpmd> set port tm node shaper profile (port_id) (node_id) \ 2542 (shaper_profile_id) 2543 2544where: 2545 2546* ``shaper_profile id``: Private shaper profile ID to be enabled on the 2547 hierarchy node. 2548 2549Add port traffic management WRED profile 2550~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ 2551 2552Create a new WRED profile:: 2553 2554 testpmd> add port tm node wred profile (port_id) (wred_profile_id) \ 2555 (color_g) (min_th_g) (max_th_g) (maxp_inv_g) (wq_log2_g) \ 2556 (color_y) (min_th_y) (max_th_y) (maxp_inv_y) (wq_log2_y) \ 2557 (color_r) (min_th_r) (max_th_r) (maxp_inv_r) (wq_log2_r) 2558 2559where: 2560 2561* ``wred_profile id``: Identifier for the newly create WRED profile 2562* ``color_g``: Packet color (green) 2563* ``min_th_g``: Minimum queue threshold for packet with green color 2564* ``max_th_g``: Minimum queue threshold for packet with green color 2565* ``maxp_inv_g``: Inverse of packet marking probability maximum value (maxp) 2566* ``wq_log2_g``: Negated log2 of queue weight (wq) 2567* ``color_y``: Packet color (yellow) 2568* ``min_th_y``: Minimum queue threshold for packet with yellow color 2569* ``max_th_y``: Minimum queue threshold for packet with yellow color 2570* ``maxp_inv_y``: Inverse of packet marking probability maximum value (maxp) 2571* ``wq_log2_y``: Negated log2 of queue weight (wq) 2572* ``color_r``: Packet color (red) 2573* ``min_th_r``: Minimum queue threshold for packet with yellow color 2574* ``max_th_r``: Minimum queue threshold for packet with yellow color 2575* ``maxp_inv_r``: Inverse of packet marking probability maximum value (maxp) 2576* ``wq_log2_r``: Negated log2 of queue weight (wq) 2577 2578Delete port traffic management WRED profile 2579~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ 2580 2581Delete the WRED profile:: 2582 2583 testpmd> del port tm node wred profile (port_id) (wred_profile_id) 2584 2585Add port traffic management hierarchy nonleaf node 2586~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ 2587 2588Add nonleaf node to port traffic management hiearchy:: 2589 2590 testpmd> add port tm nonleaf node (port_id) (node_id) (parent_node_id) \ 2591 (priority) (weight) (level_id) (shaper_profile_id) \ 2592 (n_sp_priorities) (stats_mask) (n_shared_shapers) \ 2593 [(shared_shaper_0) (shared_shaper_1) ...] \ 2594 2595where: 2596 2597* ``parent_node_id``: Node ID of the parent. 2598* ``priority``: Node priority (highest node priority is zero). This is used by 2599 the SP algorithm running on the parent node for scheduling this node. 2600* ``weight``: Node weight (lowest weight is one). The node weight is relative 2601 to the weight sum of all siblings that have the same priority. It is used by 2602 the WFQ algorithm running on the parent node for scheduling this node. 2603* ``level_id``: Hiearchy level of the node. 2604* ``shaper_profile_id``: Shaper profile ID of the private shaper to be used by 2605 the node. 2606* ``n_sp_priorities``: Number of strict priorities. 2607* ``stats_mask``: Mask of statistics counter types to be enabled for this node. 2608* ``n_shared_shapers``: Number of shared shapers. 2609* ``shared_shaper_id``: Shared shaper id. 2610 2611Add port traffic management hierarchy leaf node 2612~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ 2613 2614Add leaf node to port traffic management hiearchy:: 2615 2616 testpmd> add port tm leaf node (port_id) (node_id) (parent_node_id) \ 2617 (priority) (weight) (level_id) (shaper_profile_id) \ 2618 (cman_mode) (wred_profile_id) (stats_mask) (n_shared_shapers) \ 2619 [(shared_shaper_id) (shared_shaper_id) ...] \ 2620 2621where: 2622 2623* ``parent_node_id``: Node ID of the parent. 2624* ``priority``: Node priority (highest node priority is zero). This is used by 2625 the SP algorithm running on the parent node for scheduling this node. 2626* ``weight``: Node weight (lowest weight is one). The node weight is relative 2627 to the weight sum of all siblings that have the same priority. It is used by 2628 the WFQ algorithm running on the parent node for scheduling this node. 2629* ``level_id``: Hiearchy level of the node. 2630* ``shaper_profile_id``: Shaper profile ID of the private shaper to be used by 2631 the node. 2632* ``cman_mode``: Congestion management mode to be enabled for this node. 2633* ``wred_profile_id``: WRED profile id to be enabled for this node. 2634* ``stats_mask``: Mask of statistics counter types to be enabled for this node. 2635* ``n_shared_shapers``: Number of shared shapers. 2636* ``shared_shaper_id``: Shared shaper id. 2637 2638Delete port traffic management hierarchy node 2639~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ 2640 2641Delete node from port traffic management hiearchy:: 2642 2643 testpmd> del port tm node (port_id) (node_id) 2644 2645Update port traffic management hierarchy parent node 2646~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ 2647 2648Update port traffic management hierarchy parent node:: 2649 2650 testpmd> set port tm node parent (port_id) (node_id) (parent_node_id) \ 2651 (priority) (weight) 2652 2653This function can only be called after the hierarchy commit invocation. Its 2654success depends on the port support for this operation, as advertised through 2655the port capability set. This function is valid for all nodes of the traffic 2656management hierarchy except root node. 2657 2658Suspend port traffic management hierarchy node 2659~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ 2660 2661 testpmd> suspend port tm node (port_id) (node_id) 2662 2663Resume port traffic management hierarchy node 2664~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ 2665 2666 testpmd> resume port tm node (port_id) (node_id) 2667 2668Commit port traffic management hierarchy 2669~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ 2670 2671Commit the traffic management hierarchy on the port:: 2672 2673 testpmd> port tm hierarchy commit (port_id) (clean_on_fail) 2674 2675where: 2676 2677* ``clean_on_fail``: When set to non-zero, hierarchy is cleared on function 2678 call failure. On the other hand, hierarchy is preserved when this parameter 2679 is equal to zero. 2680 2681Set port traffic management mark VLAN dei 2682~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ 2683 2684Enables/Disables the traffic management marking on the port for VLAN packets:: 2685 2686 testpmd> set port tm mark vlan_dei <port_id> <green> <yellow> <red> 2687 2688where: 2689 2690* ``port_id``: The port which on which VLAN packets marked as ``green`` or 2691 ``yellow`` or ``red`` will have dei bit enabled 2692 2693* ``green`` enable 1, disable 0 marking for dei bit of VLAN packets marked as green 2694 2695* ``yellow`` enable 1, disable 0 marking for dei bit of VLAN packets marked as yellow 2696 2697* ``red`` enable 1, disable 0 marking for dei bit of VLAN packets marked as red 2698 2699Set port traffic management mark IP dscp 2700~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ 2701 2702Enables/Disables the traffic management marking on the port for IP dscp packets:: 2703 2704 testpmd> set port tm mark ip_dscp <port_id> <green> <yellow> <red> 2705 2706where: 2707 2708* ``port_id``: The port which on which IP packets marked as ``green`` or 2709 ``yellow`` or ``red`` will have IP dscp bits updated 2710 2711* ``green`` enable 1, disable 0 marking IP dscp to low drop precedence for green packets 2712 2713* ``yellow`` enable 1, disable 0 marking IP dscp to medium drop precedence for yellow packets 2714 2715* ``red`` enable 1, disable 0 marking IP dscp to high drop precedence for red packets 2716 2717Set port traffic management mark IP ecn 2718~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ 2719 2720Enables/Disables the traffic management marking on the port for IP ecn packets:: 2721 2722 testpmd> set port tm mark ip_ecn <port_id> <green> <yellow> <red> 2723 2724where: 2725 2726* ``port_id``: The port which on which IP packets marked as ``green`` or 2727 ``yellow`` or ``red`` will have IP ecn bits updated 2728 2729* ``green`` enable 1, disable 0 marking IP ecn for green marked packets with ecn of 2'b01 or 2'b10 2730 to ecn of 2'b11 when IP is caring TCP or SCTP 2731 2732* ``yellow`` enable 1, disable 0 marking IP ecn for yellow marked packets with ecn of 2'b01 or 2'b10 2733 to ecn of 2'b11 when IP is caring TCP or SCTP 2734 2735* ``red`` enable 1, disable 0 marking IP ecn for yellow marked packets with ecn of 2'b01 or 2'b10 2736 to ecn of 2'b11 when IP is caring TCP or SCTP 2737 2738Set port traffic management default hierarchy (softnic forwarding mode) 2739~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ 2740 2741set the traffic management default hierarchy on the port:: 2742 2743 testpmd> set port tm hierarchy default (port_id) 2744 2745Filter Functions 2746---------------- 2747 2748This section details the available filter functions that are available. 2749 2750Note these functions interface the deprecated legacy filtering framework, 2751superseded by *rte_flow*. See `Flow rules management`_. 2752 2753ethertype_filter 2754~~~~~~~~~~~~~~~~~~~~ 2755 2756Add or delete a L2 Ethertype filter, which identify packets by their L2 Ethertype mainly assign them to a receive queue:: 2757 2758 ethertype_filter (port_id) (add|del) (mac_addr|mac_ignr) (mac_address) \ 2759 ethertype (ether_type) (drop|fwd) queue (queue_id) 2760 2761The available information parameters are: 2762 2763* ``port_id``: The port which the Ethertype filter assigned on. 2764 2765* ``mac_addr``: Compare destination mac address. 2766 2767* ``mac_ignr``: Ignore destination mac address match. 2768 2769* ``mac_address``: Destination mac address to match. 2770 2771* ``ether_type``: The EtherType value want to match, 2772 for example 0x0806 for ARP packet. 0x0800 (IPv4) and 0x86DD (IPv6) are invalid. 2773 2774* ``queue_id``: The receive queue associated with this EtherType filter. 2775 It is meaningless when deleting or dropping. 2776 2777Example, to add/remove an ethertype filter rule:: 2778 2779 testpmd> ethertype_filter 0 add mac_ignr 00:11:22:33:44:55 \ 2780 ethertype 0x0806 fwd queue 3 2781 2782 testpmd> ethertype_filter 0 del mac_ignr 00:11:22:33:44:55 \ 2783 ethertype 0x0806 fwd queue 3 2784 27852tuple_filter 2786~~~~~~~~~~~~~~~~~ 2787 2788Add or delete a 2-tuple filter, 2789which identifies packets by specific protocol and destination TCP/UDP port 2790and forwards packets into one of the receive queues:: 2791 2792 2tuple_filter (port_id) (add|del) dst_port (dst_port_value) \ 2793 protocol (protocol_value) mask (mask_value) \ 2794 tcp_flags (tcp_flags_value) priority (prio_value) \ 2795 queue (queue_id) 2796 2797The available information parameters are: 2798 2799* ``port_id``: The port which the 2-tuple filter assigned on. 2800 2801* ``dst_port_value``: Destination port in L4. 2802 2803* ``protocol_value``: IP L4 protocol. 2804 2805* ``mask_value``: Participates in the match or not by bit for field above, 1b means participate. 2806 2807* ``tcp_flags_value``: TCP control bits. The non-zero value is invalid, when the pro_value is not set to 0x06 (TCP). 2808 2809* ``prio_value``: Priority of this filter. 2810 2811* ``queue_id``: The receive queue associated with this 2-tuple filter. 2812 2813Example, to add/remove an 2tuple filter rule:: 2814 2815 testpmd> 2tuple_filter 0 add dst_port 32 protocol 0x06 mask 0x03 \ 2816 tcp_flags 0x02 priority 3 queue 3 2817 2818 testpmd> 2tuple_filter 0 del dst_port 32 protocol 0x06 mask 0x03 \ 2819 tcp_flags 0x02 priority 3 queue 3 2820 28215tuple_filter 2822~~~~~~~~~~~~~~~~~ 2823 2824Add or delete a 5-tuple filter, 2825which consists of a 5-tuple (protocol, source and destination IP addresses, source and destination TCP/UDP/SCTP port) 2826and routes packets into one of the receive queues:: 2827 2828 5tuple_filter (port_id) (add|del) dst_ip (dst_address) src_ip \ 2829 (src_address) dst_port (dst_port_value) \ 2830 src_port (src_port_value) protocol (protocol_value) \ 2831 mask (mask_value) tcp_flags (tcp_flags_value) \ 2832 priority (prio_value) queue (queue_id) 2833 2834The available information parameters are: 2835 2836* ``port_id``: The port which the 5-tuple filter assigned on. 2837 2838* ``dst_address``: Destination IP address. 2839 2840* ``src_address``: Source IP address. 2841 2842* ``dst_port_value``: TCP/UDP destination port. 2843 2844* ``src_port_value``: TCP/UDP source port. 2845 2846* ``protocol_value``: L4 protocol. 2847 2848* ``mask_value``: Participates in the match or not by bit for field above, 1b means participate 2849 2850* ``tcp_flags_value``: TCP control bits. The non-zero value is invalid, when the protocol_value is not set to 0x06 (TCP). 2851 2852* ``prio_value``: The priority of this filter. 2853 2854* ``queue_id``: The receive queue associated with this 5-tuple filter. 2855 2856Example, to add/remove an 5tuple filter rule:: 2857 2858 testpmd> 5tuple_filter 0 add dst_ip 2.2.2.5 src_ip 2.2.2.4 \ 2859 dst_port 64 src_port 32 protocol 0x06 mask 0x1F \ 2860 flags 0x0 priority 3 queue 3 2861 2862 testpmd> 5tuple_filter 0 del dst_ip 2.2.2.5 src_ip 2.2.2.4 \ 2863 dst_port 64 src_port 32 protocol 0x06 mask 0x1F \ 2864 flags 0x0 priority 3 queue 3 2865 2866syn_filter 2867~~~~~~~~~~ 2868 2869Using the SYN filter, TCP packets whose *SYN* flag is set can be forwarded to a separate queue:: 2870 2871 syn_filter (port_id) (add|del) priority (high|low) queue (queue_id) 2872 2873The available information parameters are: 2874 2875* ``port_id``: The port which the SYN filter assigned on. 2876 2877* ``high``: This SYN filter has higher priority than other filters. 2878 2879* ``low``: This SYN filter has lower priority than other filters. 2880 2881* ``queue_id``: The receive queue associated with this SYN filter 2882 2883Example:: 2884 2885 testpmd> syn_filter 0 add priority high queue 3 2886 2887flex_filter 2888~~~~~~~~~~~ 2889 2890With flex filter, packets can be recognized by any arbitrary pattern within the first 128 bytes of the packet 2891and routed into one of the receive queues:: 2892 2893 flex_filter (port_id) (add|del) len (len_value) bytes (bytes_value) \ 2894 mask (mask_value) priority (prio_value) queue (queue_id) 2895 2896The available information parameters are: 2897 2898* ``port_id``: The port which the Flex filter is assigned on. 2899 2900* ``len_value``: Filter length in bytes, no greater than 128. 2901 2902* ``bytes_value``: A string in hexadecimal, means the value the flex filter needs to match. 2903 2904* ``mask_value``: A string in hexadecimal, bit 1 means corresponding byte participates in the match. 2905 2906* ``prio_value``: The priority of this filter. 2907 2908* ``queue_id``: The receive queue associated with this Flex filter. 2909 2910Example:: 2911 2912 testpmd> flex_filter 0 add len 16 bytes 0x00000000000000000000000008060000 \ 2913 mask 000C priority 3 queue 3 2914 2915 testpmd> flex_filter 0 del len 16 bytes 0x00000000000000000000000008060000 \ 2916 mask 000C priority 3 queue 3 2917 2918 2919.. _testpmd_flow_director: 2920 2921flow_director_filter 2922~~~~~~~~~~~~~~~~~~~~ 2923 2924The Flow Director works in receive mode to identify specific flows or sets of flows and route them to specific queues. 2925 2926Four types of filtering are supported which are referred to as Perfect Match, Signature, Perfect-mac-vlan and 2927Perfect-tunnel filters, the match mode is set by the ``--pkt-filter-mode`` command-line parameter: 2928 2929* Perfect match filters. 2930 The hardware checks a match between the masked fields of the received packets and the programmed filters. 2931 The masked fields are for IP flow. 2932 2933* Signature filters. 2934 The hardware checks a match between a hash-based signature of the masked fields of the received packet. 2935 2936* Perfect-mac-vlan match filters. 2937 The hardware checks a match between the masked fields of the received packets and the programmed filters. 2938 The masked fields are for MAC VLAN flow. 2939 2940* Perfect-tunnel match filters. 2941 The hardware checks a match between the masked fields of the received packets and the programmed filters. 2942 The masked fields are for tunnel flow. 2943 2944* Perfect-raw-flow-type match filters. 2945 The hardware checks a match between the masked fields of the received packets and pre-loaded raw (template) packet. 2946 The masked fields are specified by input sets. 2947 2948The Flow Director filters can match the different fields for different type of packet: flow type, specific input set 2949per flow type and the flexible payload. 2950 2951The Flow Director can also mask out parts of all of these fields so that filters 2952are only applied to certain fields or parts of the fields. 2953 2954Note that for raw flow type mode the source and destination fields in the 2955raw packet buffer need to be presented in a reversed order with respect 2956to the expected received packets. 2957For example: IP source and destination addresses or TCP/UDP/SCTP 2958source and destination ports 2959 2960Different NICs may have different capabilities, command show port fdir (port_id) can be used to acquire the information. 2961 2962# Commands to add flow director filters of different flow types:: 2963 2964 flow_director_filter (port_id) mode IP (add|del|update) \ 2965 flow (ipv4-other|ipv4-frag|ipv6-other|ipv6-frag) \ 2966 src (src_ip_address) dst (dst_ip_address) \ 2967 tos (tos_value) proto (proto_value) ttl (ttl_value) \ 2968 vlan (vlan_value) flexbytes (flexbytes_value) \ 2969 (drop|fwd) pf|vf(vf_id) queue (queue_id) \ 2970 fd_id (fd_id_value) 2971 2972 flow_director_filter (port_id) mode IP (add|del|update) \ 2973 flow (ipv4-tcp|ipv4-udp|ipv6-tcp|ipv6-udp) \ 2974 src (src_ip_address) (src_port) \ 2975 dst (dst_ip_address) (dst_port) \ 2976 tos (tos_value) ttl (ttl_value) \ 2977 vlan (vlan_value) flexbytes (flexbytes_value) \ 2978 (drop|fwd) queue pf|vf(vf_id) (queue_id) \ 2979 fd_id (fd_id_value) 2980 2981 flow_director_filter (port_id) mode IP (add|del|update) \ 2982 flow (ipv4-sctp|ipv6-sctp) \ 2983 src (src_ip_address) (src_port) \ 2984 dst (dst_ip_address) (dst_port) \ 2985 tos (tos_value) ttl (ttl_value) \ 2986 tag (verification_tag) vlan (vlan_value) \ 2987 flexbytes (flexbytes_value) (drop|fwd) \ 2988 pf|vf(vf_id) queue (queue_id) fd_id (fd_id_value) 2989 2990 flow_director_filter (port_id) mode IP (add|del|update) flow l2_payload \ 2991 ether (ethertype) flexbytes (flexbytes_value) \ 2992 (drop|fwd) pf|vf(vf_id) queue (queue_id) 2993 fd_id (fd_id_value) 2994 2995 flow_director_filter (port_id) mode MAC-VLAN (add|del|update) \ 2996 mac (mac_address) vlan (vlan_value) \ 2997 flexbytes (flexbytes_value) (drop|fwd) \ 2998 queue (queue_id) fd_id (fd_id_value) 2999 3000 flow_director_filter (port_id) mode Tunnel (add|del|update) \ 3001 mac (mac_address) vlan (vlan_value) \ 3002 tunnel (NVGRE|VxLAN) tunnel-id (tunnel_id_value) \ 3003 flexbytes (flexbytes_value) (drop|fwd) \ 3004 queue (queue_id) fd_id (fd_id_value) 3005 3006 flow_director_filter (port_id) mode raw (add|del|update) flow (flow_id) \ 3007 (drop|fwd) queue (queue_id) fd_id (fd_id_value) \ 3008 packet (packet file name) 3009 3010For example, to add an ipv4-udp flow type filter:: 3011 3012 testpmd> flow_director_filter 0 mode IP add flow ipv4-udp src 2.2.2.3 32 \ 3013 dst 2.2.2.5 33 tos 2 ttl 40 vlan 0x1 flexbytes (0x88,0x48) \ 3014 fwd pf queue 1 fd_id 1 3015 3016For example, add an ipv4-other flow type filter:: 3017 3018 testpmd> flow_director_filter 0 mode IP add flow ipv4-other src 2.2.2.3 \ 3019 dst 2.2.2.5 tos 2 proto 20 ttl 40 vlan 0x1 \ 3020 flexbytes (0x88,0x48) fwd pf queue 1 fd_id 1 3021 3022flush_flow_director 3023~~~~~~~~~~~~~~~~~~~ 3024 3025Flush all flow director filters on a device:: 3026 3027 testpmd> flush_flow_director (port_id) 3028 3029Example, to flush all flow director filter on port 0:: 3030 3031 testpmd> flush_flow_director 0 3032 3033flow_director_mask 3034~~~~~~~~~~~~~~~~~~ 3035 3036Set flow director's input masks:: 3037 3038 flow_director_mask (port_id) mode IP vlan (vlan_value) \ 3039 src_mask (ipv4_src) (ipv6_src) (src_port) \ 3040 dst_mask (ipv4_dst) (ipv6_dst) (dst_port) 3041 3042 flow_director_mask (port_id) mode MAC-VLAN vlan (vlan_value) 3043 3044 flow_director_mask (port_id) mode Tunnel vlan (vlan_value) \ 3045 mac (mac_value) tunnel-type (tunnel_type_value) \ 3046 tunnel-id (tunnel_id_value) 3047 3048Example, to set flow director mask on port 0:: 3049 3050 testpmd> flow_director_mask 0 mode IP vlan 0xefff \ 3051 src_mask 255.255.255.255 \ 3052 FFFF:FFFF:FFFF:FFFF:FFFF:FFFF:FFFF:FFFF 0xFFFF \ 3053 dst_mask 255.255.255.255 \ 3054 FFFF:FFFF:FFFF:FFFF:FFFF:FFFF:FFFF:FFFF 0xFFFF 3055 3056flow_director_flex_mask 3057~~~~~~~~~~~~~~~~~~~~~~~ 3058 3059set masks of flow director's flexible payload based on certain flow type:: 3060 3061 testpmd> flow_director_flex_mask (port_id) \ 3062 flow (none|ipv4-other|ipv4-frag|ipv4-tcp|ipv4-udp|ipv4-sctp| \ 3063 ipv6-other|ipv6-frag|ipv6-tcp|ipv6-udp|ipv6-sctp| \ 3064 l2_payload|all) (mask) 3065 3066Example, to set flow director's flex mask for all flow type on port 0:: 3067 3068 testpmd> flow_director_flex_mask 0 flow all \ 3069 (0xff,0xff,0,0,0,0,0,0,0,0,0,0,0,0,0,0) 3070 3071 3072flow_director_flex_payload 3073~~~~~~~~~~~~~~~~~~~~~~~~~~ 3074 3075Configure flexible payload selection:: 3076 3077 flow_director_flex_payload (port_id) (raw|l2|l3|l4) (config) 3078 3079For example, to select the first 16 bytes from the offset 4 (bytes) of packet's payload as flexible payload:: 3080 3081 testpmd> flow_director_flex_payload 0 l4 \ 3082 (4,5,6,7,8,9,10,11,12,13,14,15,16,17,18,19) 3083 3084get_sym_hash_ena_per_port 3085~~~~~~~~~~~~~~~~~~~~~~~~~ 3086 3087Get symmetric hash enable configuration per port:: 3088 3089 get_sym_hash_ena_per_port (port_id) 3090 3091For example, to get symmetric hash enable configuration of port 1:: 3092 3093 testpmd> get_sym_hash_ena_per_port 1 3094 3095set_sym_hash_ena_per_port 3096~~~~~~~~~~~~~~~~~~~~~~~~~ 3097 3098Set symmetric hash enable configuration per port to enable or disable:: 3099 3100 set_sym_hash_ena_per_port (port_id) (enable|disable) 3101 3102For example, to set symmetric hash enable configuration of port 1 to enable:: 3103 3104 testpmd> set_sym_hash_ena_per_port 1 enable 3105 3106get_hash_global_config 3107~~~~~~~~~~~~~~~~~~~~~~ 3108 3109Get the global configurations of hash filters:: 3110 3111 get_hash_global_config (port_id) 3112 3113For example, to get the global configurations of hash filters of port 1:: 3114 3115 testpmd> get_hash_global_config 1 3116 3117set_hash_global_config 3118~~~~~~~~~~~~~~~~~~~~~~ 3119 3120Set the global configurations of hash filters:: 3121 3122 set_hash_global_config (port_id) (toeplitz|simple_xor|default) \ 3123 (ipv4|ipv4-frag|ipv4-tcp|ipv4-udp|ipv4-sctp|ipv4-other|ipv6|ipv6-frag| \ 3124 ipv6-tcp|ipv6-udp|ipv6-sctp|ipv6-other|l2_payload|<flow_id>) \ 3125 (enable|disable) 3126 3127For example, to enable simple_xor for flow type of ipv6 on port 2:: 3128 3129 testpmd> set_hash_global_config 2 simple_xor ipv6 enable 3130 3131set_hash_input_set 3132~~~~~~~~~~~~~~~~~~ 3133 3134Set the input set for hash:: 3135 3136 set_hash_input_set (port_id) (ipv4-frag|ipv4-tcp|ipv4-udp|ipv4-sctp| \ 3137 ipv4-other|ipv6-frag|ipv6-tcp|ipv6-udp|ipv6-sctp|ipv6-other| \ 3138 l2_payload|<flow_id>) (ovlan|ivlan|src-ipv4|dst-ipv4|src-ipv6|dst-ipv6| \ 3139 ipv4-tos|ipv4-proto|ipv6-tc|ipv6-next-header|udp-src-port|udp-dst-port| \ 3140 tcp-src-port|tcp-dst-port|sctp-src-port|sctp-dst-port|sctp-veri-tag| \ 3141 udp-key|gre-key|fld-1st|fld-2nd|fld-3rd|fld-4th|fld-5th|fld-6th|fld-7th| \ 3142 fld-8th|none) (select|add) 3143 3144For example, to add source IP to hash input set for flow type of ipv4-udp on port 0:: 3145 3146 testpmd> set_hash_input_set 0 ipv4-udp src-ipv4 add 3147 3148set_fdir_input_set 3149~~~~~~~~~~~~~~~~~~ 3150 3151The Flow Director filters can match the different fields for different type of packet, i.e. specific input set 3152on per flow type and the flexible payload. This command can be used to change input set for each flow type. 3153 3154Set the input set for flow director:: 3155 3156 set_fdir_input_set (port_id) (ipv4-frag|ipv4-tcp|ipv4-udp|ipv4-sctp| \ 3157 ipv4-other|ipv6|ipv6-frag|ipv6-tcp|ipv6-udp|ipv6-sctp|ipv6-other| \ 3158 l2_payload|<flow_id>) (ivlan|ethertype|src-ipv4|dst-ipv4|src-ipv6|dst-ipv6| \ 3159 ipv4-tos|ipv4-proto|ipv4-ttl|ipv6-tc|ipv6-next-header|ipv6-hop-limits| \ 3160 tudp-src-port|udp-dst-port|cp-src-port|tcp-dst-port|sctp-src-port| \ 3161 sctp-dst-port|sctp-veri-tag|none) (select|add) 3162 3163For example to add source IP to FD input set for flow type of ipv4-udp on port 0:: 3164 3165 testpmd> set_fdir_input_set 0 ipv4-udp src-ipv4 add 3166 3167global_config 3168~~~~~~~~~~~~~ 3169 3170Set different GRE key length for input set:: 3171 3172 global_config (port_id) gre-key-len (number in bytes) 3173 3174For example to set GRE key length for input set to 4 bytes on port 0:: 3175 3176 testpmd> global_config 0 gre-key-len 4 3177 3178 3179.. _testpmd_rte_flow: 3180 3181Flow rules management 3182--------------------- 3183 3184Control of the generic flow API (*rte_flow*) is fully exposed through the 3185``flow`` command (validation, creation, destruction, queries and operation 3186modes). 3187 3188Considering *rte_flow* overlaps with all `Filter Functions`_, using both 3189features simultaneously may cause undefined side-effects and is therefore 3190not recommended. 3191 3192``flow`` syntax 3193~~~~~~~~~~~~~~~ 3194 3195Because the ``flow`` command uses dynamic tokens to handle the large number 3196of possible flow rules combinations, its behavior differs slightly from 3197other commands, in particular: 3198 3199- Pressing *?* or the *<tab>* key displays contextual help for the current 3200 token, not that of the entire command. 3201 3202- Optional and repeated parameters are supported (provided they are listed 3203 in the contextual help). 3204 3205The first parameter stands for the operation mode. Possible operations and 3206their general syntax are described below. They are covered in detail in the 3207following sections. 3208 3209- Check whether a flow rule can be created:: 3210 3211 flow validate {port_id} 3212 [group {group_id}] [priority {level}] [ingress] [egress] [transfer] 3213 pattern {item} [/ {item} [...]] / end 3214 actions {action} [/ {action} [...]] / end 3215 3216- Create a flow rule:: 3217 3218 flow create {port_id} 3219 [group {group_id}] [priority {level}] [ingress] [egress] [transfer] 3220 pattern {item} [/ {item} [...]] / end 3221 actions {action} [/ {action} [...]] / end 3222 3223- Destroy specific flow rules:: 3224 3225 flow destroy {port_id} rule {rule_id} [...] 3226 3227- Destroy all flow rules:: 3228 3229 flow flush {port_id} 3230 3231- Query an existing flow rule:: 3232 3233 flow query {port_id} {rule_id} {action} 3234 3235- List existing flow rules sorted by priority, filtered by group 3236 identifiers:: 3237 3238 flow list {port_id} [group {group_id}] [...] 3239 3240- Restrict ingress traffic to the defined flow rules:: 3241 3242 flow isolate {port_id} {boolean} 3243 3244Validating flow rules 3245~~~~~~~~~~~~~~~~~~~~~ 3246 3247``flow validate`` reports whether a flow rule would be accepted by the 3248underlying device in its current state but stops short of creating it. It is 3249bound to ``rte_flow_validate()``:: 3250 3251 flow validate {port_id} 3252 [group {group_id}] [priority {level}] [ingress] [egress] [transfer] 3253 pattern {item} [/ {item} [...]] / end 3254 actions {action} [/ {action} [...]] / end 3255 3256If successful, it will show:: 3257 3258 Flow rule validated 3259 3260Otherwise it will show an error message of the form:: 3261 3262 Caught error type [...] ([...]): [...] 3263 3264This command uses the same parameters as ``flow create``, their format is 3265described in `Creating flow rules`_. 3266 3267Check whether redirecting any Ethernet packet received on port 0 to RX queue 3268index 6 is supported:: 3269 3270 testpmd> flow validate 0 ingress pattern eth / end 3271 actions queue index 6 / end 3272 Flow rule validated 3273 testpmd> 3274 3275Port 0 does not support TCPv6 rules:: 3276 3277 testpmd> flow validate 0 ingress pattern eth / ipv6 / tcp / end 3278 actions drop / end 3279 Caught error type 9 (specific pattern item): Invalid argument 3280 testpmd> 3281 3282Creating flow rules 3283~~~~~~~~~~~~~~~~~~~ 3284 3285``flow create`` validates and creates the specified flow rule. It is bound 3286to ``rte_flow_create()``:: 3287 3288 flow create {port_id} 3289 [group {group_id}] [priority {level}] [ingress] [egress] [transfer] 3290 pattern {item} [/ {item} [...]] / end 3291 actions {action} [/ {action} [...]] / end 3292 3293If successful, it will return a flow rule ID usable with other commands:: 3294 3295 Flow rule #[...] created 3296 3297Otherwise it will show an error message of the form:: 3298 3299 Caught error type [...] ([...]): [...] 3300 3301Parameters describe in the following order: 3302 3303- Attributes (*group*, *priority*, *ingress*, *egress*, *transfer* tokens). 3304- A matching pattern, starting with the *pattern* token and terminated by an 3305 *end* pattern item. 3306- Actions, starting with the *actions* token and terminated by an *end* 3307 action. 3308 3309These translate directly to *rte_flow* objects provided as-is to the 3310underlying functions. 3311 3312The shortest valid definition only comprises mandatory tokens:: 3313 3314 testpmd> flow create 0 pattern end actions end 3315 3316Note that PMDs may refuse rules that essentially do nothing such as this 3317one. 3318 3319**All unspecified object values are automatically initialized to 0.** 3320 3321Attributes 3322^^^^^^^^^^ 3323 3324These tokens affect flow rule attributes (``struct rte_flow_attr``) and are 3325specified before the ``pattern`` token. 3326 3327- ``group {group id}``: priority group. 3328- ``priority {level}``: priority level within group. 3329- ``ingress``: rule applies to ingress traffic. 3330- ``egress``: rule applies to egress traffic. 3331- ``transfer``: apply rule directly to endpoints found in pattern. 3332 3333Each instance of an attribute specified several times overrides the previous 3334value as shown below (group 4 is used):: 3335 3336 testpmd> flow create 0 group 42 group 24 group 4 [...] 3337 3338Note that once enabled, ``ingress`` and ``egress`` cannot be disabled. 3339 3340While not specifying a direction is an error, some rules may allow both 3341simultaneously. 3342 3343Most rules affect RX therefore contain the ``ingress`` token:: 3344 3345 testpmd> flow create 0 ingress pattern [...] 3346 3347Matching pattern 3348^^^^^^^^^^^^^^^^ 3349 3350A matching pattern starts after the ``pattern`` token. It is made of pattern 3351items and is terminated by a mandatory ``end`` item. 3352 3353Items are named after their type (*RTE_FLOW_ITEM_TYPE_* from ``enum 3354rte_flow_item_type``). 3355 3356The ``/`` token is used as a separator between pattern items as shown 3357below:: 3358 3359 testpmd> flow create 0 ingress pattern eth / ipv4 / udp / end [...] 3360 3361Note that protocol items like these must be stacked from lowest to highest 3362layer to make sense. For instance, the following rule is either invalid or 3363unlikely to match any packet:: 3364 3365 testpmd> flow create 0 ingress pattern eth / udp / ipv4 / end [...] 3366 3367More information on these restrictions can be found in the *rte_flow* 3368documentation. 3369 3370Several items support additional specification structures, for example 3371``ipv4`` allows specifying source and destination addresses as follows:: 3372 3373 testpmd> flow create 0 ingress pattern eth / ipv4 src is 10.1.1.1 3374 dst is 10.2.0.0 / end [...] 3375 3376This rule matches all IPv4 traffic with the specified properties. 3377 3378In this example, ``src`` and ``dst`` are field names of the underlying 3379``struct rte_flow_item_ipv4`` object. All item properties can be specified 3380in a similar fashion. 3381 3382The ``is`` token means that the subsequent value must be matched exactly, 3383and assigns ``spec`` and ``mask`` fields in ``struct rte_flow_item`` 3384accordingly. Possible assignment tokens are: 3385 3386- ``is``: match value perfectly (with full bit-mask). 3387- ``spec``: match value according to configured bit-mask. 3388- ``last``: specify upper bound to establish a range. 3389- ``mask``: specify bit-mask with relevant bits set to one. 3390- ``prefix``: generate bit-mask from a prefix length. 3391 3392These yield identical results:: 3393 3394 ipv4 src is 10.1.1.1 3395 3396:: 3397 3398 ipv4 src spec 10.1.1.1 src mask 255.255.255.255 3399 3400:: 3401 3402 ipv4 src spec 10.1.1.1 src prefix 32 3403 3404:: 3405 3406 ipv4 src is 10.1.1.1 src last 10.1.1.1 # range with a single value 3407 3408:: 3409 3410 ipv4 src is 10.1.1.1 src last 0 # 0 disables range 3411 3412Inclusive ranges can be defined with ``last``:: 3413 3414 ipv4 src is 10.1.1.1 src last 10.2.3.4 # 10.1.1.1 to 10.2.3.4 3415 3416Note that ``mask`` affects both ``spec`` and ``last``:: 3417 3418 ipv4 src is 10.1.1.1 src last 10.2.3.4 src mask 255.255.0.0 3419 # matches 10.1.0.0 to 10.2.255.255 3420 3421Properties can be modified multiple times:: 3422 3423 ipv4 src is 10.1.1.1 src is 10.1.2.3 src is 10.2.3.4 # matches 10.2.3.4 3424 3425:: 3426 3427 ipv4 src is 10.1.1.1 src prefix 24 src prefix 16 # matches 10.1.0.0/16 3428 3429Pattern items 3430^^^^^^^^^^^^^ 3431 3432This section lists supported pattern items and their attributes, if any. 3433 3434- ``end``: end list of pattern items. 3435 3436- ``void``: no-op pattern item. 3437 3438- ``invert``: perform actions when pattern does not match. 3439 3440- ``any``: match any protocol for the current layer. 3441 3442 - ``num {unsigned}``: number of layers covered. 3443 3444- ``pf``: match traffic from/to the physical function. 3445 3446- ``vf``: match traffic from/to a virtual function ID. 3447 3448 - ``id {unsigned}``: VF ID. 3449 3450- ``phy_port``: match traffic from/to a specific physical port. 3451 3452 - ``index {unsigned}``: physical port index. 3453 3454- ``port_id``: match traffic from/to a given DPDK port ID. 3455 3456 - ``id {unsigned}``: DPDK port ID. 3457 3458- ``mark``: match value set in previously matched flow rule using the mark action. 3459 3460 - ``id {unsigned}``: arbitrary integer value. 3461 3462- ``raw``: match an arbitrary byte string. 3463 3464 - ``relative {boolean}``: look for pattern after the previous item. 3465 - ``search {boolean}``: search pattern from offset (see also limit). 3466 - ``offset {integer}``: absolute or relative offset for pattern. 3467 - ``limit {unsigned}``: search area limit for start of pattern. 3468 - ``pattern {string}``: byte string to look for. 3469 3470- ``eth``: match Ethernet header. 3471 3472 - ``dst {MAC-48}``: destination MAC. 3473 - ``src {MAC-48}``: source MAC. 3474 - ``type {unsigned}``: EtherType or TPID. 3475 3476- ``vlan``: match 802.1Q/ad VLAN tag. 3477 3478 - ``tci {unsigned}``: tag control information. 3479 - ``pcp {unsigned}``: priority code point. 3480 - ``dei {unsigned}``: drop eligible indicator. 3481 - ``vid {unsigned}``: VLAN identifier. 3482 - ``inner_type {unsigned}``: inner EtherType or TPID. 3483 3484- ``ipv4``: match IPv4 header. 3485 3486 - ``tos {unsigned}``: type of service. 3487 - ``ttl {unsigned}``: time to live. 3488 - ``proto {unsigned}``: next protocol ID. 3489 - ``src {ipv4 address}``: source address. 3490 - ``dst {ipv4 address}``: destination address. 3491 3492- ``ipv6``: match IPv6 header. 3493 3494 - ``tc {unsigned}``: traffic class. 3495 - ``flow {unsigned}``: flow label. 3496 - ``proto {unsigned}``: protocol (next header). 3497 - ``hop {unsigned}``: hop limit. 3498 - ``src {ipv6 address}``: source address. 3499 - ``dst {ipv6 address}``: destination address. 3500 3501- ``icmp``: match ICMP header. 3502 3503 - ``type {unsigned}``: ICMP packet type. 3504 - ``code {unsigned}``: ICMP packet code. 3505 3506- ``udp``: match UDP header. 3507 3508 - ``src {unsigned}``: UDP source port. 3509 - ``dst {unsigned}``: UDP destination port. 3510 3511- ``tcp``: match TCP header. 3512 3513 - ``src {unsigned}``: TCP source port. 3514 - ``dst {unsigned}``: TCP destination port. 3515 3516- ``sctp``: match SCTP header. 3517 3518 - ``src {unsigned}``: SCTP source port. 3519 - ``dst {unsigned}``: SCTP destination port. 3520 - ``tag {unsigned}``: validation tag. 3521 - ``cksum {unsigned}``: checksum. 3522 3523- ``vxlan``: match VXLAN header. 3524 3525 - ``vni {unsigned}``: VXLAN identifier. 3526 3527- ``e_tag``: match IEEE 802.1BR E-Tag header. 3528 3529 - ``grp_ecid_b {unsigned}``: GRP and E-CID base. 3530 3531- ``nvgre``: match NVGRE header. 3532 3533 - ``tni {unsigned}``: virtual subnet ID. 3534 3535- ``mpls``: match MPLS header. 3536 3537 - ``label {unsigned}``: MPLS label. 3538 3539- ``gre``: match GRE header. 3540 3541 - ``protocol {unsigned}``: protocol type. 3542 3543- ``fuzzy``: fuzzy pattern match, expect faster than default. 3544 3545 - ``thresh {unsigned}``: accuracy threshold. 3546 3547- ``gtp``, ``gtpc``, ``gtpu``: match GTPv1 header. 3548 3549 - ``teid {unsigned}``: tunnel endpoint identifier. 3550 3551- ``geneve``: match GENEVE header. 3552 3553 - ``vni {unsigned}``: virtual network identifier. 3554 - ``protocol {unsigned}``: protocol type. 3555 3556- ``vxlan-gpe``: match VXLAN-GPE header. 3557 3558 - ``vni {unsigned}``: VXLAN-GPE identifier. 3559 3560- ``arp_eth_ipv4``: match ARP header for Ethernet/IPv4. 3561 3562 - ``sha {MAC-48}``: sender hardware address. 3563 - ``spa {ipv4 address}``: sender IPv4 address. 3564 - ``tha {MAC-48}``: target hardware address. 3565 - ``tpa {ipv4 address}``: target IPv4 address. 3566 3567- ``ipv6_ext``: match presence of any IPv6 extension header. 3568 3569 - ``next_hdr {unsigned}``: next header. 3570 3571- ``icmp6``: match any ICMPv6 header. 3572 3573 - ``type {unsigned}``: ICMPv6 type. 3574 - ``code {unsigned}``: ICMPv6 code. 3575 3576- ``icmp6_nd_ns``: match ICMPv6 neighbor discovery solicitation. 3577 3578 - ``target_addr {ipv6 address}``: target address. 3579 3580- ``icmp6_nd_na``: match ICMPv6 neighbor discovery advertisement. 3581 3582 - ``target_addr {ipv6 address}``: target address. 3583 3584- ``icmp6_nd_opt``: match presence of any ICMPv6 neighbor discovery option. 3585 3586 - ``type {unsigned}``: ND option type. 3587 3588- ``icmp6_nd_opt_sla_eth``: match ICMPv6 neighbor discovery source Ethernet 3589 link-layer address option. 3590 3591 - ``sla {MAC-48}``: source Ethernet LLA. 3592 3593- ``icmp6_nd_opt_sla_eth``: match ICMPv6 neighbor discovery target Ethernet 3594 link-layer address option. 3595 3596 - ``tla {MAC-48}``: target Ethernet LLA. 3597 3598Actions list 3599^^^^^^^^^^^^ 3600 3601A list of actions starts after the ``actions`` token in the same fashion as 3602`Matching pattern`_; actions are separated by ``/`` tokens and the list is 3603terminated by a mandatory ``end`` action. 3604 3605Actions are named after their type (*RTE_FLOW_ACTION_TYPE_* from ``enum 3606rte_flow_action_type``). 3607 3608Dropping all incoming UDPv4 packets can be expressed as follows:: 3609 3610 testpmd> flow create 0 ingress pattern eth / ipv4 / udp / end 3611 actions drop / end 3612 3613Several actions have configurable properties which must be specified when 3614there is no valid default value. For example, ``queue`` requires a target 3615queue index. 3616 3617This rule redirects incoming UDPv4 traffic to queue index 6:: 3618 3619 testpmd> flow create 0 ingress pattern eth / ipv4 / udp / end 3620 actions queue index 6 / end 3621 3622While this one could be rejected by PMDs (unspecified queue index):: 3623 3624 testpmd> flow create 0 ingress pattern eth / ipv4 / udp / end 3625 actions queue / end 3626 3627As defined by *rte_flow*, the list is not ordered, all actions of a given 3628rule are performed simultaneously. These are equivalent:: 3629 3630 queue index 6 / void / mark id 42 / end 3631 3632:: 3633 3634 void / mark id 42 / queue index 6 / end 3635 3636All actions in a list should have different types, otherwise only the last 3637action of a given type is taken into account:: 3638 3639 queue index 4 / queue index 5 / queue index 6 / end # will use queue 6 3640 3641:: 3642 3643 drop / drop / drop / end # drop is performed only once 3644 3645:: 3646 3647 mark id 42 / queue index 3 / mark id 24 / end # mark will be 24 3648 3649Considering they are performed simultaneously, opposite and overlapping 3650actions can sometimes be combined when the end result is unambiguous:: 3651 3652 drop / queue index 6 / end # drop has no effect 3653 3654:: 3655 3656 queue index 6 / rss queues 6 7 8 / end # queue has no effect 3657 3658:: 3659 3660 drop / passthru / end # drop has no effect 3661 3662Note that PMDs may still refuse such combinations. 3663 3664Actions 3665^^^^^^^ 3666 3667This section lists supported actions and their attributes, if any. 3668 3669- ``end``: end list of actions. 3670 3671- ``void``: no-op action. 3672 3673- ``passthru``: let subsequent rule process matched packets. 3674 3675- ``jump``: redirect traffic to group on device. 3676 3677 - ``group {unsigned}``: group to redirect to. 3678 3679- ``mark``: attach 32 bit value to packets. 3680 3681 - ``id {unsigned}``: 32 bit value to return with packets. 3682 3683- ``flag``: flag packets. 3684 3685- ``queue``: assign packets to a given queue index. 3686 3687 - ``index {unsigned}``: queue index to use. 3688 3689- ``drop``: drop packets (note: passthru has priority). 3690 3691- ``count``: enable counters for this rule. 3692 3693- ``rss``: spread packets among several queues. 3694 3695 - ``func {hash function}``: RSS hash function to apply, allowed tokens are 3696 the same as `set_hash_global_config`_. 3697 3698 - ``level {unsigned}``: encapsulation level for ``types``. 3699 3700 - ``types [{RSS hash type} [...]] end``: specific RSS hash types, allowed 3701 tokens are the same as `set_hash_input_set`_, except that an empty list 3702 does not disable RSS but instead requests unspecified "best-effort" 3703 settings. 3704 3705 - ``key {string}``: RSS hash key, overrides ``key_len``. 3706 3707 - ``key_len {unsigned}``: RSS hash key length in bytes, can be used in 3708 conjunction with ``key`` to pad or truncate it. 3709 3710 - ``queues [{unsigned} [...]] end``: queue indices to use. 3711 3712- ``pf``: direct traffic to physical function. 3713 3714- ``vf``: direct traffic to a virtual function ID. 3715 3716 - ``original {boolean}``: use original VF ID if possible. 3717 - ``id {unsigned}``: VF ID. 3718 3719- ``phy_port``: direct packets to physical port index. 3720 3721 - ``original {boolean}``: use original port index if possible. 3722 - ``index {unsigned}``: physical port index. 3723 3724- ``port_id``: direct matching traffic to a given DPDK port ID. 3725 3726 - ``original {boolean}``: use original DPDK port ID if possible. 3727 - ``id {unsigned}``: DPDK port ID. 3728 3729- ``of_set_mpls_ttl``: OpenFlow's ``OFPAT_SET_MPLS_TTL``. 3730 3731 - ``mpls_ttl``: MPLS TTL. 3732 3733- ``of_dec_mpls_ttl``: OpenFlow's ``OFPAT_DEC_MPLS_TTL``. 3734 3735- ``of_set_nw_ttl``: OpenFlow's ``OFPAT_SET_NW_TTL``. 3736 3737 - ``nw_ttl``: IP TTL. 3738 3739- ``of_dec_nw_ttl``: OpenFlow's ``OFPAT_DEC_NW_TTL``. 3740 3741- ``of_copy_ttl_out``: OpenFlow's ``OFPAT_COPY_TTL_OUT``. 3742 3743- ``of_copy_ttl_in``: OpenFlow's ``OFPAT_COPY_TTL_IN``. 3744 3745- ``of_pop_vlan``: OpenFlow's ``OFPAT_POP_VLAN``. 3746 3747- ``of_push_vlan``: OpenFlow's ``OFPAT_PUSH_VLAN``. 3748 3749 - ``ethertype``: Ethertype. 3750 3751- ``of_set_vlan_vid``: OpenFlow's ``OFPAT_SET_VLAN_VID``. 3752 3753 - ``vlan_vid``: VLAN id. 3754 3755- ``of_set_vlan_pcp``: OpenFlow's ``OFPAT_SET_VLAN_PCP``. 3756 3757 - ``vlan_pcp``: VLAN priority. 3758 3759- ``of_pop_mpls``: OpenFlow's ``OFPAT_POP_MPLS``. 3760 3761 - ``ethertype``: Ethertype. 3762 3763- ``of_push_mpls``: OpenFlow's ``OFPAT_PUSH_MPLS``. 3764 3765 - ``ethertype``: Ethertype. 3766 3767- ``vxlan_encap``: Performs a VXLAN encapsulation, outer layer configuration 3768 is done through `Config VXLAN Encap outer layers`_. 3769 3770- ``vxlan_decap``: Performs a decapsulation action by stripping all headers of 3771 the VXLAN tunnel network overlay from the matched flow. 3772 3773- ``nvgre_encap``: Performs a NVGRE encapsulation, outer layer configuration 3774 is done through `Config NVGRE Encap outer layers`_. 3775 3776- ``nvgre_decap``: Performs a decapsulation action by stripping all headers of 3777 the NVGRE tunnel network overlay from the matched flow. 3778 3779- ``set_ipv4_src``: Set a new IPv4 source address in the outermost IPv4 header. 3780 3781 - ``ipv4_addr``: New IPv4 source address. 3782 3783- ``set_ipv4_dst``: Set a new IPv4 destination address in the outermost IPv4 3784 header. 3785 3786 - ``ipv4_addr``: New IPv4 destination address. 3787 3788- ``set_ipv6_src``: Set a new IPv6 source address in the outermost IPv6 header. 3789 3790 - ``ipv6_addr``: New IPv6 source address. 3791 3792- ``set_ipv6_dst``: Set a new IPv6 destination address in the outermost IPv6 3793 header. 3794 3795 - ``ipv6_addr``: New IPv6 destination address. 3796 3797- ``of_set_tp_src``: Set a new source port number in the outermost TCP/UDP 3798 header. 3799 3800 - ``port``: New TCP/UDP source port number. 3801 3802- ``of_set_tp_dst``: Set a new destination port number in the outermost TCP/UDP 3803 header. 3804 3805 - ``port``: New TCP/UDP destination port number. 3806 3807- ``mac_swap``: Swap the source and destination MAC addresses in the outermost 3808 Ethernet header. 3809 3810- ``dec_ttl``: Performs a decrease TTL value action 3811 3812- ``set_ttl``: Set TTL value with specificed value 3813 - ``ttl_value {unsigned}``: The new TTL value to be set 3814 3815- ``set_mac_src``: set source MAC address 3816 3817 - ``mac_addr {MAC-48}``: new source MAC address 3818 3819- ``set_mac_dst``: set destination MAC address 3820 3821 - ``mac_addr {MAC-48}``: new destination MAC address 3822 3823Destroying flow rules 3824~~~~~~~~~~~~~~~~~~~~~ 3825 3826``flow destroy`` destroys one or more rules from their rule ID (as returned 3827by ``flow create``), this command calls ``rte_flow_destroy()`` as many 3828times as necessary:: 3829 3830 flow destroy {port_id} rule {rule_id} [...] 3831 3832If successful, it will show:: 3833 3834 Flow rule #[...] destroyed 3835 3836It does not report anything for rule IDs that do not exist. The usual error 3837message is shown when a rule cannot be destroyed:: 3838 3839 Caught error type [...] ([...]): [...] 3840 3841``flow flush`` destroys all rules on a device and does not take extra 3842arguments. It is bound to ``rte_flow_flush()``:: 3843 3844 flow flush {port_id} 3845 3846Any errors are reported as above. 3847 3848Creating several rules and destroying them:: 3849 3850 testpmd> flow create 0 ingress pattern eth / ipv6 / end 3851 actions queue index 2 / end 3852 Flow rule #0 created 3853 testpmd> flow create 0 ingress pattern eth / ipv4 / end 3854 actions queue index 3 / end 3855 Flow rule #1 created 3856 testpmd> flow destroy 0 rule 0 rule 1 3857 Flow rule #1 destroyed 3858 Flow rule #0 destroyed 3859 testpmd> 3860 3861The same result can be achieved using ``flow flush``:: 3862 3863 testpmd> flow create 0 ingress pattern eth / ipv6 / end 3864 actions queue index 2 / end 3865 Flow rule #0 created 3866 testpmd> flow create 0 ingress pattern eth / ipv4 / end 3867 actions queue index 3 / end 3868 Flow rule #1 created 3869 testpmd> flow flush 0 3870 testpmd> 3871 3872Non-existent rule IDs are ignored:: 3873 3874 testpmd> flow create 0 ingress pattern eth / ipv6 / end 3875 actions queue index 2 / end 3876 Flow rule #0 created 3877 testpmd> flow create 0 ingress pattern eth / ipv4 / end 3878 actions queue index 3 / end 3879 Flow rule #1 created 3880 testpmd> flow destroy 0 rule 42 rule 10 rule 2 3881 testpmd> 3882 testpmd> flow destroy 0 rule 0 3883 Flow rule #0 destroyed 3884 testpmd> 3885 3886Querying flow rules 3887~~~~~~~~~~~~~~~~~~~ 3888 3889``flow query`` queries a specific action of a flow rule having that 3890ability. Such actions collect information that can be reported using this 3891command. It is bound to ``rte_flow_query()``:: 3892 3893 flow query {port_id} {rule_id} {action} 3894 3895If successful, it will display either the retrieved data for known actions 3896or the following message:: 3897 3898 Cannot display result for action type [...] ([...]) 3899 3900Otherwise, it will complain either that the rule does not exist or that some 3901error occurred:: 3902 3903 Flow rule #[...] not found 3904 3905:: 3906 3907 Caught error type [...] ([...]): [...] 3908 3909Currently only the ``count`` action is supported. This action reports the 3910number of packets that hit the flow rule and the total number of bytes. Its 3911output has the following format:: 3912 3913 count: 3914 hits_set: [...] # whether "hits" contains a valid value 3915 bytes_set: [...] # whether "bytes" contains a valid value 3916 hits: [...] # number of packets 3917 bytes: [...] # number of bytes 3918 3919Querying counters for TCPv6 packets redirected to queue 6:: 3920 3921 testpmd> flow create 0 ingress pattern eth / ipv6 / tcp / end 3922 actions queue index 6 / count / end 3923 Flow rule #4 created 3924 testpmd> flow query 0 4 count 3925 count: 3926 hits_set: 1 3927 bytes_set: 0 3928 hits: 386446 3929 bytes: 0 3930 testpmd> 3931 3932Listing flow rules 3933~~~~~~~~~~~~~~~~~~ 3934 3935``flow list`` lists existing flow rules sorted by priority and optionally 3936filtered by group identifiers:: 3937 3938 flow list {port_id} [group {group_id}] [...] 3939 3940This command only fails with the following message if the device does not 3941exist:: 3942 3943 Invalid port [...] 3944 3945Output consists of a header line followed by a short description of each 3946flow rule, one per line. There is no output at all when no flow rules are 3947configured on the device:: 3948 3949 ID Group Prio Attr Rule 3950 [...] [...] [...] [...] [...] 3951 3952``Attr`` column flags: 3953 3954- ``i`` for ``ingress``. 3955- ``e`` for ``egress``. 3956 3957Creating several flow rules and listing them:: 3958 3959 testpmd> flow create 0 ingress pattern eth / ipv4 / end 3960 actions queue index 6 / end 3961 Flow rule #0 created 3962 testpmd> flow create 0 ingress pattern eth / ipv6 / end 3963 actions queue index 2 / end 3964 Flow rule #1 created 3965 testpmd> flow create 0 priority 5 ingress pattern eth / ipv4 / udp / end 3966 actions rss queues 6 7 8 end / end 3967 Flow rule #2 created 3968 testpmd> flow list 0 3969 ID Group Prio Attr Rule 3970 0 0 0 i- ETH IPV4 => QUEUE 3971 1 0 0 i- ETH IPV6 => QUEUE 3972 2 0 5 i- ETH IPV4 UDP => RSS 3973 testpmd> 3974 3975Rules are sorted by priority (i.e. group ID first, then priority level):: 3976 3977 testpmd> flow list 1 3978 ID Group Prio Attr Rule 3979 0 0 0 i- ETH => COUNT 3980 6 0 500 i- ETH IPV6 TCP => DROP COUNT 3981 5 0 1000 i- ETH IPV6 ICMP => QUEUE 3982 1 24 0 i- ETH IPV4 UDP => QUEUE 3983 4 24 10 i- ETH IPV4 TCP => DROP 3984 3 24 20 i- ETH IPV4 => DROP 3985 2 24 42 i- ETH IPV4 UDP => QUEUE 3986 7 63 0 i- ETH IPV6 UDP VXLAN => MARK QUEUE 3987 testpmd> 3988 3989Output can be limited to specific groups:: 3990 3991 testpmd> flow list 1 group 0 group 63 3992 ID Group Prio Attr Rule 3993 0 0 0 i- ETH => COUNT 3994 6 0 500 i- ETH IPV6 TCP => DROP COUNT 3995 5 0 1000 i- ETH IPV6 ICMP => QUEUE 3996 7 63 0 i- ETH IPV6 UDP VXLAN => MARK QUEUE 3997 testpmd> 3998 3999Toggling isolated mode 4000~~~~~~~~~~~~~~~~~~~~~~ 4001 4002``flow isolate`` can be used to tell the underlying PMD that ingress traffic 4003must only be injected from the defined flow rules; that no default traffic 4004is expected outside those rules and the driver is free to assign more 4005resources to handle them. It is bound to ``rte_flow_isolate()``:: 4006 4007 flow isolate {port_id} {boolean} 4008 4009If successful, enabling or disabling isolated mode shows either:: 4010 4011 Ingress traffic on port [...] 4012 is now restricted to the defined flow rules 4013 4014Or:: 4015 4016 Ingress traffic on port [...] 4017 is not restricted anymore to the defined flow rules 4018 4019Otherwise, in case of error:: 4020 4021 Caught error type [...] ([...]): [...] 4022 4023Mainly due to its side effects, PMDs supporting this mode may not have the 4024ability to toggle it more than once without reinitializing affected ports 4025first (e.g. by exiting testpmd). 4026 4027Enabling isolated mode:: 4028 4029 testpmd> flow isolate 0 true 4030 Ingress traffic on port 0 is now restricted to the defined flow rules 4031 testpmd> 4032 4033Disabling isolated mode:: 4034 4035 testpmd> flow isolate 0 false 4036 Ingress traffic on port 0 is not restricted anymore to the defined flow rules 4037 testpmd> 4038 4039Sample QinQ flow rules 4040~~~~~~~~~~~~~~~~~~~~~~ 4041 4042Before creating QinQ rule(s) the following commands should be issued to enable QinQ:: 4043 4044 testpmd> port stop 0 4045 testpmd> vlan set qinq on 0 4046 4047The above command sets the inner and outer TPID's to 0x8100. 4048 4049To change the TPID's the following commands should be used:: 4050 4051 testpmd> vlan set outer tpid 0xa100 0 4052 testpmd> vlan set inner tpid 0x9100 0 4053 testpmd> port start 0 4054 4055Validate and create a QinQ rule on port 0 to steer traffic to a VF queue in a VM. 4056 4057:: 4058 4059 testpmd> flow validate 0 ingress pattern eth / vlan tci is 123 / 4060 vlan tci is 456 / end actions vf id 1 / queue index 0 / end 4061 Flow rule #0 validated 4062 4063 testpmd> flow create 0 ingress pattern eth / vlan tci is 4 / 4064 vlan tci is 456 / end actions vf id 123 / queue index 0 / end 4065 Flow rule #0 created 4066 4067 testpmd> flow list 0 4068 ID Group Prio Attr Rule 4069 0 0 0 i- ETH VLAN VLAN=>VF QUEUE 4070 4071Validate and create a QinQ rule on port 0 to steer traffic to a queue on the host. 4072 4073:: 4074 4075 testpmd> flow validate 0 ingress pattern eth / vlan tci is 321 / 4076 vlan tci is 654 / end actions pf / queue index 0 / end 4077 Flow rule #1 validated 4078 4079 testpmd> flow create 0 ingress pattern eth / vlan tci is 321 / 4080 vlan tci is 654 / end actions pf / queue index 1 / end 4081 Flow rule #1 created 4082 4083 testpmd> flow list 0 4084 ID Group Prio Attr Rule 4085 0 0 0 i- ETH VLAN VLAN=>VF QUEUE 4086 1 0 0 i- ETH VLAN VLAN=>PF QUEUE 4087 4088Sample VXLAN encapsulation rule 4089~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ 4090 4091VXLAN encapsulation outer layer has default value pre-configured in testpmd 4092source code, those can be changed by using the following commands 4093 4094IPv4 VXLAN outer header:: 4095 4096 testpmd> set vxlan ip-version ipv4 vni 4 udp-src 4 udp-dst 4 ip-src 127.0.0.1 4097 ip-dst 128.0.0.1 eth-src 11:11:11:11:11:11 eth-dst 22:22:22:22:22:22 4098 testpmd> flow create 0 ingress pattern end actions vxlan_encap / 4099 queue index 0 / end 4100 4101 testpmd> set vxlan-with-vlan ip-version ipv4 vni 4 udp-src 4 udp-dst 4 ip-src 4102 127.0.0.1 ip-dst 128.0.0.1 vlan-tci 34 eth-src 11:11:11:11:11:11 4103 eth-dst 22:22:22:22:22:22 4104 testpmd> flow create 0 ingress pattern end actions vxlan_encap / 4105 queue index 0 / end 4106 4107IPv6 VXLAN outer header:: 4108 4109 testpmd> set vxlan ip-version ipv6 vni 4 udp-src 4 udp-dst 4 ip-src ::1 4110 ip-dst ::2222 eth-src 11:11:11:11:11:11 eth-dst 22:22:22:22:22:22 4111 testpmd> flow create 0 ingress pattern end actions vxlan_encap / 4112 queue index 0 / end 4113 4114 testpmd> set vxlan-with-vlan ip-version ipv6 vni 4 udp-src 4 udp-dst 4 4115 ip-src ::1 ip-dst ::2222 vlan-tci 34 eth-src 11:11:11:11:11:11 4116 eth-dst 22:22:22:22:22:22 4117 testpmd> flow create 0 ingress pattern end actions vxlan_encap / 4118 queue index 0 / end 4119 4120Sample NVGRE encapsulation rule 4121~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ 4122 4123NVGRE encapsulation outer layer has default value pre-configured in testpmd 4124source code, those can be changed by using the following commands 4125 4126IPv4 NVGRE outer header:: 4127 4128 testpmd> set nvgre ip-version ipv4 tni 4 ip-src 127.0.0.1 ip-dst 128.0.0.1 4129 eth-src 11:11:11:11:11:11 eth-dst 22:22:22:22:22:22 4130 testpmd> flow create 0 ingress pattern end actions nvgre_encap / 4131 queue index 0 / end 4132 4133 testpmd> set nvgre-with-vlan ip-version ipv4 tni 4 ip-src 127.0.0.1 4134 ip-dst 128.0.0.1 vlan-tci 34 eth-src 11:11:11:11:11:11 4135 eth-dst 22:22:22:22:22:22 4136 testpmd> flow create 0 ingress pattern end actions nvgre_encap / 4137 queue index 0 / end 4138 4139IPv6 NVGRE outer header:: 4140 4141 testpmd> set nvgre ip-version ipv6 tni 4 ip-src ::1 ip-dst ::2222 4142 eth-src 11:11:11:11:11:11 eth-dst 22:22:22:22:22:22 4143 testpmd> flow create 0 ingress pattern end actions nvgre_encap / 4144 queue index 0 / end 4145 4146 testpmd> set nvgre-with-vlan ip-version ipv6 tni 4 ip-src ::1 ip-dst ::2222 4147 vlan-tci 34 eth-src 11:11:11:11:11:11 eth-dst 22:22:22:22:22:22 4148 testpmd> flow create 0 ingress pattern end actions nvgre_encap / 4149 queue index 0 / end 4150 4151BPF Functions 4152-------------- 4153 4154The following sections show functions to load/unload eBPF based filters. 4155 4156bpf-load 4157~~~~~~~~ 4158 4159Load an eBPF program as a callback for partciular RX/TX queue:: 4160 4161 testpmd> bpf-load rx|tx (portid) (queueid) (load-flags) (bpf-prog-filename) 4162 4163The available load-flags are: 4164 4165* ``J``: use JIT generated native code, otherwise BPF interpreter will be used. 4166 4167* ``M``: assume input parameter is a pointer to rte_mbuf, otherwise assume it is a pointer to first segment's data. 4168 4169* ``-``: none. 4170 4171.. note:: 4172 4173 You'll need clang v3.7 or above to build bpf program you'd like to load 4174 4175For example: 4176 4177.. code-block:: console 4178 4179 cd test/bpf 4180 clang -O2 -target bpf -c t1.c 4181 4182Then to load (and JIT compile) t1.o at RX queue 0, port 1:: 4183 4184.. code-block:: console 4185 4186 testpmd> bpf-load rx 1 0 J ./dpdk.org/test/bpf/t1.o 4187 4188To load (not JITed) t1.o at TX queue 0, port 0:: 4189 4190.. code-block:: console 4191 4192 testpmd> bpf-load tx 0 0 - ./dpdk.org/test/bpf/t1.o 4193 4194bpf-unload 4195~~~~~~~~~~ 4196 4197Unload previously loaded eBPF program for partciular RX/TX queue:: 4198 4199 testpmd> bpf-unload rx|tx (portid) (queueid) 4200 4201For example to unload BPF filter from TX queue 0, port 0: 4202 4203.. code-block:: console 4204 4205 testpmd> bpf-load tx 0 0 - ./dpdk.org/test/bpf/t1.o 4206