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 452Currently the only available levels are 0 (silent except for error) and 1 (fully verbose). 453 454set log 455~~~~~~~ 456 457Set the log level for a log type:: 458 459 testpmd> set log global|(type) (level) 460 461Where: 462 463* ``type`` is the log name. 464 465* ``level`` is the log level. 466 467For example, to change the global log level:: 468 testpmd> set log global (level) 469 470Regexes can also be used for type. To change log level of user1, user2 and user3:: 471 testpmd> set log user[1-3] (level) 472 473set nbport 474~~~~~~~~~~ 475 476Set the number of ports used by the application: 477 478set nbport (num) 479 480This is equivalent to the ``--nb-ports`` command-line option. 481 482set nbcore 483~~~~~~~~~~ 484 485Set the number of cores used by the application:: 486 487 testpmd> set nbcore (num) 488 489This is equivalent to the ``--nb-cores`` command-line option. 490 491.. note:: 492 493 The number of cores used must not be greater than number of ports used multiplied by the number of queues per port. 494 495set coremask 496~~~~~~~~~~~~ 497 498Set the forwarding cores hexadecimal mask:: 499 500 testpmd> set coremask (mask) 501 502This is equivalent to the ``--coremask`` command-line option. 503 504.. note:: 505 506 The master lcore is reserved for command line parsing only and cannot be masked on for packet forwarding. 507 508set portmask 509~~~~~~~~~~~~ 510 511Set the forwarding ports hexadecimal mask:: 512 513 testpmd> set portmask (mask) 514 515This is equivalent to the ``--portmask`` command-line option. 516 517set burst 518~~~~~~~~~ 519 520Set number of packets per burst:: 521 522 testpmd> set burst (num) 523 524This is equivalent to the ``--burst command-line`` option. 525 526When retry is enabled, the transmit delay time and number of retries can also be set:: 527 528 testpmd> set burst tx delay (microseconds) retry (num) 529 530set txpkts 531~~~~~~~~~~ 532 533Set the length of each segment of the TX-ONLY packets or length of packet for FLOWGEN mode:: 534 535 testpmd> set txpkts (x[,y]*) 536 537Where x[,y]* represents a CSV list of values, without white space. 538 539set txsplit 540~~~~~~~~~~~ 541 542Set the split policy for the TX packets, applicable for TX-ONLY and CSUM forwarding modes:: 543 544 testpmd> set txsplit (off|on|rand) 545 546Where: 547 548* ``off`` disable packet copy & split for CSUM mode. 549 550* ``on`` split outgoing packet into multiple segments. Size of each segment 551 and number of segments per packet is determined by ``set txpkts`` command 552 (see above). 553 554* ``rand`` same as 'on', but number of segments per each packet is a random value between 1 and total number of segments. 555 556set corelist 557~~~~~~~~~~~~ 558 559Set the list of forwarding cores:: 560 561 testpmd> set corelist (x[,y]*) 562 563For example, to change the forwarding cores: 564 565.. code-block:: console 566 567 testpmd> set corelist 3,1 568 testpmd> show config fwd 569 570 io packet forwarding - ports=2 - cores=2 - streams=2 - NUMA support disabled 571 Logical Core 3 (socket 0) forwards packets on 1 streams: 572 RX P=0/Q=0 (socket 0) -> TX P=1/Q=0 (socket 0) peer=02:00:00:00:00:01 573 Logical Core 1 (socket 0) forwards packets on 1 streams: 574 RX P=1/Q=0 (socket 0) -> TX P=0/Q=0 (socket 0) peer=02:00:00:00:00:00 575 576.. note:: 577 578 The cores are used in the same order as specified on the command line. 579 580set portlist 581~~~~~~~~~~~~ 582 583Set the list of forwarding ports:: 584 585 testpmd> set portlist (x[,y]*) 586 587For example, to change the port forwarding: 588 589.. code-block:: console 590 591 testpmd> set portlist 0,2,1,3 592 testpmd> show config fwd 593 594 io packet forwarding - ports=4 - cores=1 - streams=4 595 Logical Core 3 (socket 0) forwards packets on 4 streams: 596 RX P=0/Q=0 (socket 0) -> TX P=2/Q=0 (socket 0) peer=02:00:00:00:00:01 597 RX P=2/Q=0 (socket 0) -> TX P=0/Q=0 (socket 0) peer=02:00:00:00:00:00 598 RX P=1/Q=0 (socket 0) -> TX P=3/Q=0 (socket 0) peer=02:00:00:00:00:03 599 RX P=3/Q=0 (socket 0) -> TX P=1/Q=0 (socket 0) peer=02:00:00:00:00:02 600 601set tx loopback 602~~~~~~~~~~~~~~~ 603 604Enable/disable tx loopback:: 605 606 testpmd> set tx loopback (port_id) (on|off) 607 608set drop enable 609~~~~~~~~~~~~~~~ 610 611set drop enable bit for all queues:: 612 613 testpmd> set all queues drop (port_id) (on|off) 614 615set split drop enable (for VF) 616~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ 617 618set split drop enable bit for VF from PF:: 619 620 testpmd> set vf split drop (port_id) (vf_id) (on|off) 621 622set mac antispoof (for VF) 623~~~~~~~~~~~~~~~~~~~~~~~~~~ 624 625Set mac antispoof for a VF from the PF:: 626 627 testpmd> set vf mac antispoof (port_id) (vf_id) (on|off) 628 629set macsec offload 630~~~~~~~~~~~~~~~~~~ 631 632Enable/disable MACsec offload:: 633 634 testpmd> set macsec offload (port_id) on encrypt (on|off) replay-protect (on|off) 635 testpmd> set macsec offload (port_id) off 636 637set macsec sc 638~~~~~~~~~~~~~ 639 640Configure MACsec secure connection (SC):: 641 642 testpmd> set macsec sc (tx|rx) (port_id) (mac) (pi) 643 644.. note:: 645 646 The pi argument is ignored for tx. 647 Check the NIC Datasheet for hardware limits. 648 649set macsec sa 650~~~~~~~~~~~~~ 651 652Configure MACsec secure association (SA):: 653 654 testpmd> set macsec sa (tx|rx) (port_id) (idx) (an) (pn) (key) 655 656.. note:: 657 658 The IDX value must be 0 or 1. 659 Check the NIC Datasheet for hardware limits. 660 661set broadcast mode (for VF) 662~~~~~~~~~~~~~~~~~~~~~~~~~~~ 663 664Set broadcast mode for a VF from the PF:: 665 666 testpmd> set vf broadcast (port_id) (vf_id) (on|off) 667 668vlan set strip 669~~~~~~~~~~~~~~ 670 671Set the VLAN strip on a port:: 672 673 testpmd> vlan set strip (on|off) (port_id) 674 675vlan set stripq 676~~~~~~~~~~~~~~~ 677 678Set the VLAN strip for a queue on a port:: 679 680 testpmd> vlan set stripq (on|off) (port_id,queue_id) 681 682vlan set stripq (for VF) 683~~~~~~~~~~~~~~~~~~~~~~~~ 684 685Set VLAN strip for all queues in a pool for a VF from the PF:: 686 687 testpmd> set vf vlan stripq (port_id) (vf_id) (on|off) 688 689vlan set insert (for VF) 690~~~~~~~~~~~~~~~~~~~~~~~~ 691 692Set VLAN insert for a VF from the PF:: 693 694 testpmd> set vf vlan insert (port_id) (vf_id) (vlan_id) 695 696vlan set tag (for VF) 697~~~~~~~~~~~~~~~~~~~~~ 698 699Set VLAN tag for a VF from the PF:: 700 701 testpmd> set vf vlan tag (port_id) (vf_id) (on|off) 702 703vlan set antispoof (for VF) 704~~~~~~~~~~~~~~~~~~~~~~~~~~~ 705 706Set VLAN antispoof for a VF from the PF:: 707 708 testpmd> set vf vlan antispoof (port_id) (vf_id) (on|off) 709 710vlan set filter 711~~~~~~~~~~~~~~~ 712 713Set the VLAN filter on a port:: 714 715 testpmd> vlan set filter (on|off) (port_id) 716 717vlan set qinq 718~~~~~~~~~~~~~ 719 720Set the VLAN QinQ (extended queue in queue) on for a port:: 721 722 testpmd> vlan set qinq (on|off) (port_id) 723 724vlan set tpid 725~~~~~~~~~~~~~ 726 727Set the inner or outer VLAN TPID for packet filtering on a port:: 728 729 testpmd> vlan set (inner|outer) tpid (value) (port_id) 730 731.. note:: 732 733 TPID value must be a 16-bit number (value <= 65536). 734 735rx_vlan add 736~~~~~~~~~~~ 737 738Add a VLAN ID, or all identifiers, to the set of VLAN identifiers filtered by port ID:: 739 740 testpmd> rx_vlan add (vlan_id|all) (port_id) 741 742.. note:: 743 744 VLAN filter must be set on that port. VLAN ID < 4096. 745 Depending on the NIC used, number of vlan_ids may be limited to the maximum entries 746 in VFTA table. This is important if enabling all vlan_ids. 747 748rx_vlan rm 749~~~~~~~~~~ 750 751Remove a VLAN ID, or all identifiers, from the set of VLAN identifiers filtered by port ID:: 752 753 testpmd> rx_vlan rm (vlan_id|all) (port_id) 754 755rx_vlan add (for VF) 756~~~~~~~~~~~~~~~~~~~~ 757 758Add a VLAN ID, to the set of VLAN identifiers filtered for VF(s) for port ID:: 759 760 testpmd> rx_vlan add (vlan_id) port (port_id) vf (vf_mask) 761 762rx_vlan rm (for VF) 763~~~~~~~~~~~~~~~~~~~ 764 765Remove a VLAN ID, from the set of VLAN identifiers filtered for VF(s) for port ID:: 766 767 testpmd> rx_vlan rm (vlan_id) port (port_id) vf (vf_mask) 768 769tunnel_filter add 770~~~~~~~~~~~~~~~~~ 771 772Add a tunnel filter on a port:: 773 774 testpmd> tunnel_filter add (port_id) (outer_mac) (inner_mac) (ip_addr) \ 775 (inner_vlan) (vxlan|nvgre|ipingre) (imac-ivlan|imac-ivlan-tenid|\ 776 imac-tenid|imac|omac-imac-tenid|oip|iip) (tenant_id) (queue_id) 777 778The available information categories are: 779 780* ``vxlan``: Set tunnel type as VXLAN. 781 782* ``nvgre``: Set tunnel type as NVGRE. 783 784* ``ipingre``: Set tunnel type as IP-in-GRE. 785 786* ``imac-ivlan``: Set filter type as Inner MAC and VLAN. 787 788* ``imac-ivlan-tenid``: Set filter type as Inner MAC, VLAN and tenant ID. 789 790* ``imac-tenid``: Set filter type as Inner MAC and tenant ID. 791 792* ``imac``: Set filter type as Inner MAC. 793 794* ``omac-imac-tenid``: Set filter type as Outer MAC, Inner MAC and tenant ID. 795 796* ``oip``: Set filter type as Outer IP. 797 798* ``iip``: Set filter type as Inner IP. 799 800Example:: 801 802 testpmd> tunnel_filter add 0 68:05:CA:28:09:82 00:00:00:00:00:00 \ 803 192.168.2.2 0 ipingre oip 1 1 804 805 Set an IP-in-GRE tunnel on port 0, and the filter type is Outer IP. 806 807tunnel_filter remove 808~~~~~~~~~~~~~~~~~~~~ 809 810Remove a tunnel filter on a port:: 811 812 testpmd> tunnel_filter rm (port_id) (outer_mac) (inner_mac) (ip_addr) \ 813 (inner_vlan) (vxlan|nvgre|ipingre) (imac-ivlan|imac-ivlan-tenid|\ 814 imac-tenid|imac|omac-imac-tenid|oip|iip) (tenant_id) (queue_id) 815 816rx_vxlan_port add 817~~~~~~~~~~~~~~~~~ 818 819Add an UDP port for VXLAN packet filter on a port:: 820 821 testpmd> rx_vxlan_port add (udp_port) (port_id) 822 823rx_vxlan_port remove 824~~~~~~~~~~~~~~~~~~~~ 825 826Remove an UDP port for VXLAN packet filter on a port:: 827 828 testpmd> rx_vxlan_port rm (udp_port) (port_id) 829 830tx_vlan set 831~~~~~~~~~~~ 832 833Set hardware insertion of VLAN IDs in packets sent on a port:: 834 835 testpmd> tx_vlan set (port_id) vlan_id[, vlan_id_outer] 836 837For example, set a single VLAN ID (5) insertion on port 0:: 838 839 tx_vlan set 0 5 840 841Or, set double VLAN ID (inner: 2, outer: 3) insertion on port 1:: 842 843 tx_vlan set 1 2 3 844 845 846tx_vlan set pvid 847~~~~~~~~~~~~~~~~ 848 849Set port based hardware insertion of VLAN ID in packets sent on a port:: 850 851 testpmd> tx_vlan set pvid (port_id) (vlan_id) (on|off) 852 853tx_vlan reset 854~~~~~~~~~~~~~ 855 856Disable hardware insertion of a VLAN header in packets sent on a port:: 857 858 testpmd> tx_vlan reset (port_id) 859 860csum set 861~~~~~~~~ 862 863Select hardware or software calculation of the checksum when 864transmitting a packet using the ``csum`` forwarding engine:: 865 866 testpmd> csum set (ip|udp|tcp|sctp|outer-ip|outer-udp) (hw|sw) (port_id) 867 868Where: 869 870* ``ip|udp|tcp|sctp`` always relate to the inner layer. 871 872* ``outer-ip`` relates to the outer IP layer (only for IPv4) in the case where the packet is recognized 873 as a tunnel packet by the forwarding engine (vxlan, gre and ipip are 874 supported). See also the ``csum parse-tunnel`` command. 875 876* ``outer-udp`` relates to the outer UDP layer in the case where the packet is recognized 877 as a tunnel packet by the forwarding engine (vxlan, vxlan-gpe are 878 supported). See also the ``csum parse-tunnel`` command. 879 880.. note:: 881 882 Check the NIC Datasheet for hardware limits. 883 884RSS queue region 885~~~~~~~~~~~~~~~~ 886 887Set RSS queue region span on a port:: 888 889 testpmd> set port (port_id) queue-region region_id (value) \ 890 queue_start_index (value) queue_num (value) 891 892Set flowtype mapping on a RSS queue region on a port:: 893 894 testpmd> set port (port_id) queue-region region_id (value) flowtype (value) 895 896where: 897 898* For the flowtype(pctype) of packet,the specific index for each type has 899 been defined in file i40e_type.h as enum i40e_filter_pctype. 900 901Set user priority mapping on a RSS queue region on a port:: 902 903 testpmd> set port (port_id) queue-region UP (value) region_id (value) 904 905Flush all queue region related configuration on a port:: 906 907 testpmd> set port (port_id) queue-region flush (on|off) 908 909where: 910 911* "on"is just an enable function which server for other configuration, 912 it is for all configuration about queue region from up layer, 913 at first will only keep in DPDK softwarestored in driver, 914 only after "flush on", it commit all configuration to HW. 915 "off" is just clean all configuration about queue region just now, 916 and restore all to DPDK i40e driver default config when start up. 917 918Show all queue region related configuration info on a port:: 919 920 testpmd> show port (port_id) queue-region 921 922.. note:: 923 924 Queue region only support on PF by now, so these command is 925 only for configuration of queue region on PF port. 926 927csum parse-tunnel 928~~~~~~~~~~~~~~~~~ 929 930Define how tunneled packets should be handled by the csum forward 931engine:: 932 933 testpmd> csum parse-tunnel (on|off) (tx_port_id) 934 935If enabled, the csum forward engine will try to recognize supported 936tunnel headers (vxlan, gre, ipip). 937 938If disabled, treat tunnel packets as non-tunneled packets (a inner 939header is handled as a packet payload). 940 941.. note:: 942 943 The port argument is the TX port like in the ``csum set`` command. 944 945Example: 946 947Consider a packet in packet like the following:: 948 949 eth_out/ipv4_out/udp_out/vxlan/eth_in/ipv4_in/tcp_in 950 951* If parse-tunnel is enabled, the ``ip|udp|tcp|sctp`` parameters of ``csum set`` 952 command relate to the inner headers (here ``ipv4_in`` and ``tcp_in``), and the 953 ``outer-ip|outer-udp`` parameter relates to the outer headers (here ``ipv4_out`` and ``udp_out``). 954 955* If parse-tunnel is disabled, the ``ip|udp|tcp|sctp`` parameters of ``csum set`` 956 command relate to the outer headers, here ``ipv4_out`` and ``udp_out``. 957 958csum show 959~~~~~~~~~ 960 961Display tx checksum offload configuration:: 962 963 testpmd> csum show (port_id) 964 965tso set 966~~~~~~~ 967 968Enable TCP Segmentation Offload (TSO) in the ``csum`` forwarding engine:: 969 970 testpmd> tso set (segsize) (port_id) 971 972.. note:: 973 974 Check the NIC datasheet for hardware limits. 975 976tso show 977~~~~~~~~ 978 979Display the status of TCP Segmentation Offload:: 980 981 testpmd> tso show (port_id) 982 983set port - gro 984~~~~~~~~~~~~~~ 985 986Enable or disable GRO in ``csum`` forwarding engine:: 987 988 testpmd> set port <port_id> gro on|off 989 990If enabled, the csum forwarding engine will perform GRO on the TCP/IPv4 991packets received from the given port. 992 993If disabled, packets received from the given port won't be performed 994GRO. By default, GRO is disabled for all ports. 995 996.. note:: 997 998 When enable GRO for a port, TCP/IPv4 packets received from the port 999 will be performed GRO. After GRO, all merged packets have bad 1000 checksums, since the GRO library doesn't re-calculate checksums for 1001 the merged packets. Therefore, if users want the merged packets to 1002 have correct checksums, please select HW IP checksum calculation and 1003 HW TCP checksum calculation for the port which the merged packets are 1004 transmitted to. 1005 1006show port - gro 1007~~~~~~~~~~~~~~~ 1008 1009Display GRO configuration for a given port:: 1010 1011 testpmd> show port <port_id> gro 1012 1013set gro flush 1014~~~~~~~~~~~~~ 1015 1016Set the cycle to flush the GROed packets from reassembly tables:: 1017 1018 testpmd> set gro flush <cycles> 1019 1020When enable GRO, the csum forwarding engine performs GRO on received 1021packets, and the GROed packets are stored in reassembly tables. Users 1022can use this command to determine when the GROed packets are flushed 1023from the reassembly tables. 1024 1025The ``cycles`` is measured in GRO operation times. The csum forwarding 1026engine flushes the GROed packets from the tables every ``cycles`` GRO 1027operations. 1028 1029By default, the value of ``cycles`` is 1, which means flush GROed packets 1030from the reassembly tables as soon as one GRO operation finishes. The value 1031of ``cycles`` should be in the range of 1 to ``GRO_MAX_FLUSH_CYCLES``. 1032 1033Please note that the large value of ``cycles`` may cause the poor TCP/IP 1034stack performance. Because the GROed packets are delayed to arrive the 1035stack, thus causing more duplicated ACKs and TCP retransmissions. 1036 1037set port - gso 1038~~~~~~~~~~~~~~ 1039 1040Toggle per-port GSO support in ``csum`` forwarding engine:: 1041 1042 testpmd> set port <port_id> gso on|off 1043 1044If enabled, the csum forwarding engine will perform GSO on supported IPv4 1045packets, transmitted on the given port. 1046 1047If disabled, packets transmitted on the given port will not undergo GSO. 1048By default, GSO is disabled for all ports. 1049 1050.. note:: 1051 1052 When GSO is enabled on a port, supported IPv4 packets transmitted on that 1053 port undergo GSO. Afterwards, the segmented packets are represented by 1054 multi-segment mbufs; however, the csum forwarding engine doesn't calculation 1055 of checksums for GSO'd segments in SW. As a result, if users want correct 1056 checksums in GSO segments, they should enable HW checksum calculation for 1057 GSO-enabled ports. 1058 1059 For example, HW checksum calculation for VxLAN GSO'd packets may be enabled 1060 by setting the following options in the csum forwarding engine: 1061 1062 testpmd> csum set outer_ip hw <port_id> 1063 1064 testpmd> csum set ip hw <port_id> 1065 1066 testpmd> csum set tcp hw <port_id> 1067 1068 UDP GSO is the same as IP fragmentation, which treats the UDP header 1069 as the payload and does not modify it during segmentation. That is, 1070 after UDP GSO, only the first output fragment has the original UDP 1071 header. Therefore, users need to enable HW IP checksum calculation 1072 and SW UDP checksum calculation for GSO-enabled ports, if they want 1073 correct checksums for UDP/IPv4 packets. 1074 1075set gso segsz 1076~~~~~~~~~~~~~ 1077 1078Set the maximum GSO segment size (measured in bytes), which includes the 1079packet header and the packet payload for GSO-enabled ports (global):: 1080 1081 testpmd> set gso segsz <length> 1082 1083show port - gso 1084~~~~~~~~~~~~~~~ 1085 1086Display the status of Generic Segmentation Offload for a given port:: 1087 1088 testpmd> show port <port_id> gso 1089 1090mac_addr add 1091~~~~~~~~~~~~ 1092 1093Add an alternative MAC address to a port:: 1094 1095 testpmd> mac_addr add (port_id) (XX:XX:XX:XX:XX:XX) 1096 1097mac_addr remove 1098~~~~~~~~~~~~~~~ 1099 1100Remove a MAC address from a port:: 1101 1102 testpmd> mac_addr remove (port_id) (XX:XX:XX:XX:XX:XX) 1103 1104mac_addr add (for VF) 1105~~~~~~~~~~~~~~~~~~~~~ 1106 1107Add an alternative MAC address for a VF to a port:: 1108 1109 testpmd> mac_add add port (port_id) vf (vf_id) (XX:XX:XX:XX:XX:XX) 1110 1111mac_addr set 1112~~~~~~~~~~~~ 1113 1114Set the default MAC address for a port:: 1115 1116 testpmd> mac_addr set (port_id) (XX:XX:XX:XX:XX:XX) 1117 1118mac_addr set (for VF) 1119~~~~~~~~~~~~~~~~~~~~~ 1120 1121Set the MAC address for a VF from the PF:: 1122 1123 testpmd> set vf mac addr (port_id) (vf_id) (XX:XX:XX:XX:XX:XX) 1124 1125set eth-peer 1126~~~~~~~~~~~~ 1127 1128Set the forwarding peer address for certain port:: 1129 1130 testpmd> set eth-peer (port_id) (perr_addr) 1131 1132This is equivalent to the ``--eth-peer`` command-line option. 1133 1134set port-uta 1135~~~~~~~~~~~~ 1136 1137Set the unicast hash filter(s) on/off for a port:: 1138 1139 testpmd> set port (port_id) uta (XX:XX:XX:XX:XX:XX|all) (on|off) 1140 1141set promisc 1142~~~~~~~~~~~ 1143 1144Set the promiscuous mode on for a port or for all ports. 1145In promiscuous mode packets are not dropped if they aren't for the specified MAC address:: 1146 1147 testpmd> set promisc (port_id|all) (on|off) 1148 1149set allmulti 1150~~~~~~~~~~~~ 1151 1152Set the allmulti mode for a port or for all ports:: 1153 1154 testpmd> set allmulti (port_id|all) (on|off) 1155 1156Same as the ifconfig (8) option. Controls how multicast packets are handled. 1157 1158set promisc (for VF) 1159~~~~~~~~~~~~~~~~~~~~ 1160 1161Set the unicast promiscuous mode for a VF from PF. 1162It's supported by Intel i40e NICs now. 1163In promiscuous mode packets are not dropped if they aren't for the specified MAC address:: 1164 1165 testpmd> set vf promisc (port_id) (vf_id) (on|off) 1166 1167set allmulticast (for VF) 1168~~~~~~~~~~~~~~~~~~~~~~~~~ 1169 1170Set the multicast promiscuous mode for a VF from PF. 1171It's supported by Intel i40e NICs now. 1172In promiscuous mode packets are not dropped if they aren't for the specified MAC address:: 1173 1174 testpmd> set vf allmulti (port_id) (vf_id) (on|off) 1175 1176set tx max bandwidth (for VF) 1177~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ 1178 1179Set TX max absolute bandwidth (Mbps) for a VF from PF:: 1180 1181 testpmd> set vf tx max-bandwidth (port_id) (vf_id) (max_bandwidth) 1182 1183set tc tx min bandwidth (for VF) 1184~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ 1185 1186Set all TCs' TX min relative bandwidth (%) for a VF from PF:: 1187 1188 testpmd> set vf tc tx min-bandwidth (port_id) (vf_id) (bw1, bw2, ...) 1189 1190set tc tx max bandwidth (for VF) 1191~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ 1192 1193Set a TC's TX max absolute bandwidth (Mbps) for a VF from PF:: 1194 1195 testpmd> set vf tc tx max-bandwidth (port_id) (vf_id) (tc_no) (max_bandwidth) 1196 1197set tc strict link priority mode 1198~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ 1199 1200Set some TCs' strict link priority mode on a physical port:: 1201 1202 testpmd> set tx strict-link-priority (port_id) (tc_bitmap) 1203 1204set tc tx min bandwidth 1205~~~~~~~~~~~~~~~~~~~~~~~ 1206 1207Set all TCs' TX min relative bandwidth (%) globally for all PF and VFs:: 1208 1209 testpmd> set tc tx min-bandwidth (port_id) (bw1, bw2, ...) 1210 1211set flow_ctrl rx 1212~~~~~~~~~~~~~~~~ 1213 1214Set the link flow control parameter on a port:: 1215 1216 testpmd> set flow_ctrl rx (on|off) tx (on|off) (high_water) (low_water) \ 1217 (pause_time) (send_xon) mac_ctrl_frame_fwd (on|off) \ 1218 autoneg (on|off) (port_id) 1219 1220Where: 1221 1222* ``high_water`` (integer): High threshold value to trigger XOFF. 1223 1224* ``low_water`` (integer): Low threshold value to trigger XON. 1225 1226* ``pause_time`` (integer): Pause quota in the Pause frame. 1227 1228* ``send_xon`` (0/1): Send XON frame. 1229 1230* ``mac_ctrl_frame_fwd``: Enable receiving MAC control frames. 1231 1232* ``autoneg``: Change the auto-negotiation parameter. 1233 1234set pfc_ctrl rx 1235~~~~~~~~~~~~~~~ 1236 1237Set the priority flow control parameter on a port:: 1238 1239 testpmd> set pfc_ctrl rx (on|off) tx (on|off) (high_water) (low_water) \ 1240 (pause_time) (priority) (port_id) 1241 1242Where: 1243 1244* ``high_water`` (integer): High threshold value. 1245 1246* ``low_water`` (integer): Low threshold value. 1247 1248* ``pause_time`` (integer): Pause quota in the Pause frame. 1249 1250* ``priority`` (0-7): VLAN User Priority. 1251 1252set stat_qmap 1253~~~~~~~~~~~~~ 1254 1255Set statistics mapping (qmapping 0..15) for RX/TX queue on port:: 1256 1257 testpmd> set stat_qmap (tx|rx) (port_id) (queue_id) (qmapping) 1258 1259For example, to set rx queue 2 on port 0 to mapping 5:: 1260 1261 testpmd>set stat_qmap rx 0 2 5 1262 1263set xstats-hide-zero 1264~~~~~~~~~~~~~~~~~~~~ 1265 1266Set the option to hide zero values for xstats display:: 1267 1268 testpmd> set xstats-hide-zero on|off 1269 1270.. note:: 1271 1272 By default, the zero values are displayed for xstats. 1273 1274set port - rx/tx (for VF) 1275~~~~~~~~~~~~~~~~~~~~~~~~~ 1276 1277Set VF receive/transmit from a port:: 1278 1279 testpmd> set port (port_id) vf (vf_id) (rx|tx) (on|off) 1280 1281set port - mac address filter (for VF) 1282~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ 1283 1284Add/Remove unicast or multicast MAC addr filter for a VF:: 1285 1286 testpmd> set port (port_id) vf (vf_id) (mac_addr) \ 1287 (exact-mac|exact-mac-vlan|hashmac|hashmac-vlan) (on|off) 1288 1289set port - rx mode(for VF) 1290~~~~~~~~~~~~~~~~~~~~~~~~~~ 1291 1292Set the VF receive mode of a port:: 1293 1294 testpmd> set port (port_id) vf (vf_id) \ 1295 rxmode (AUPE|ROPE|BAM|MPE) (on|off) 1296 1297The available receive modes are: 1298 1299* ``AUPE``: Accepts untagged VLAN. 1300 1301* ``ROPE``: Accepts unicast hash. 1302 1303* ``BAM``: Accepts broadcast packets. 1304 1305* ``MPE``: Accepts all multicast packets. 1306 1307set port - tx_rate (for Queue) 1308~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ 1309 1310Set TX rate limitation for a queue on a port:: 1311 1312 testpmd> set port (port_id) queue (queue_id) rate (rate_value) 1313 1314set port - tx_rate (for VF) 1315~~~~~~~~~~~~~~~~~~~~~~~~~~~ 1316 1317Set TX rate limitation for queues in VF on a port:: 1318 1319 testpmd> set port (port_id) vf (vf_id) rate (rate_value) queue_mask (queue_mask) 1320 1321set port - mirror rule 1322~~~~~~~~~~~~~~~~~~~~~~ 1323 1324Set pool or vlan type mirror rule for a port:: 1325 1326 testpmd> set port (port_id) mirror-rule (rule_id) \ 1327 (pool-mirror-up|pool-mirror-down|vlan-mirror) \ 1328 (poolmask|vlanid[,vlanid]*) dst-pool (pool_id) (on|off) 1329 1330Set link mirror rule for a port:: 1331 1332 testpmd> set port (port_id) mirror-rule (rule_id) \ 1333 (uplink-mirror|downlink-mirror) dst-pool (pool_id) (on|off) 1334 1335For example to enable mirror traffic with vlan 0,1 to pool 0:: 1336 1337 set port 0 mirror-rule 0 vlan-mirror 0,1 dst-pool 0 on 1338 1339reset port - mirror rule 1340~~~~~~~~~~~~~~~~~~~~~~~~ 1341 1342Reset a mirror rule for a port:: 1343 1344 testpmd> reset port (port_id) mirror-rule (rule_id) 1345 1346set flush_rx 1347~~~~~~~~~~~~ 1348 1349Set the flush on RX streams before forwarding. 1350The default is flush ``on``. 1351Mainly used with PCAP drivers to turn off the default behavior of flushing the first 512 packets on RX streams:: 1352 1353 testpmd> set flush_rx off 1354 1355set bypass mode 1356~~~~~~~~~~~~~~~ 1357 1358Set the bypass mode for the lowest port on bypass enabled NIC:: 1359 1360 testpmd> set bypass mode (normal|bypass|isolate) (port_id) 1361 1362set bypass event 1363~~~~~~~~~~~~~~~~ 1364 1365Set the event required to initiate specified bypass mode for the lowest port on a bypass enabled:: 1366 1367 testpmd> set bypass event (timeout|os_on|os_off|power_on|power_off) \ 1368 mode (normal|bypass|isolate) (port_id) 1369 1370Where: 1371 1372* ``timeout``: Enable bypass after watchdog timeout. 1373 1374* ``os_on``: Enable bypass when OS/board is powered on. 1375 1376* ``os_off``: Enable bypass when OS/board is powered off. 1377 1378* ``power_on``: Enable bypass when power supply is turned on. 1379 1380* ``power_off``: Enable bypass when power supply is turned off. 1381 1382 1383set bypass timeout 1384~~~~~~~~~~~~~~~~~~ 1385 1386Set the bypass watchdog timeout to ``n`` seconds where 0 = instant:: 1387 1388 testpmd> set bypass timeout (0|1.5|2|3|4|8|16|32) 1389 1390show bypass config 1391~~~~~~~~~~~~~~~~~~ 1392 1393Show the bypass configuration for a bypass enabled NIC using the lowest port on the NIC:: 1394 1395 testpmd> show bypass config (port_id) 1396 1397set link up 1398~~~~~~~~~~~ 1399 1400Set link up for a port:: 1401 1402 testpmd> set link-up port (port id) 1403 1404set link down 1405~~~~~~~~~~~~~ 1406 1407Set link down for a port:: 1408 1409 testpmd> set link-down port (port id) 1410 1411E-tag set 1412~~~~~~~~~ 1413 1414Enable E-tag insertion for a VF on a port:: 1415 1416 testpmd> E-tag set insertion on port-tag-id (value) port (port_id) vf (vf_id) 1417 1418Disable E-tag insertion for a VF on a port:: 1419 1420 testpmd> E-tag set insertion off port (port_id) vf (vf_id) 1421 1422Enable/disable E-tag stripping on a port:: 1423 1424 testpmd> E-tag set stripping (on|off) port (port_id) 1425 1426Enable/disable E-tag based forwarding on a port:: 1427 1428 testpmd> E-tag set forwarding (on|off) port (port_id) 1429 1430Add an E-tag forwarding filter on a port:: 1431 1432 testpmd> E-tag set filter add e-tag-id (value) dst-pool (pool_id) port (port_id) 1433 1434Delete an E-tag forwarding filter on a port:: 1435 testpmd> E-tag set filter del e-tag-id (value) port (port_id) 1436 1437ddp add 1438~~~~~~~ 1439 1440Load a dynamic device personalization (DDP) profile and store backup profile:: 1441 1442 testpmd> ddp add (port_id) (profile_path[,backup_profile_path]) 1443 1444ddp del 1445~~~~~~~ 1446 1447Delete a dynamic device personalization profile and restore backup profile:: 1448 1449 testpmd> ddp del (port_id) (backup_profile_path) 1450 1451ptype mapping 1452~~~~~~~~~~~~~ 1453 1454List all items from the ptype mapping table:: 1455 1456 testpmd> ptype mapping get (port_id) (valid_only) 1457 1458Where: 1459 1460* ``valid_only``: A flag indicates if only list valid items(=1) or all itemss(=0). 1461 1462Replace a specific or a group of software defined ptype with a new one:: 1463 1464 testpmd> ptype mapping replace (port_id) (target) (mask) (pkt_type) 1465 1466where: 1467 1468* ``target``: A specific software ptype or a mask to represent a group of software ptypes. 1469 1470* ``mask``: A flag indicate if "target" is a specific software ptype(=0) or a ptype mask(=1). 1471 1472* ``pkt_type``: The new software ptype to replace the old ones. 1473 1474Update hardware defined ptype to software defined packet type mapping table:: 1475 1476 testpmd> ptype mapping update (port_id) (hw_ptype) (sw_ptype) 1477 1478where: 1479 1480* ``hw_ptype``: hardware ptype as the index of the ptype mapping table. 1481 1482* ``sw_ptype``: software ptype as the value of the ptype mapping table. 1483 1484Reset ptype mapping table:: 1485 1486 testpmd> ptype mapping reset (port_id) 1487 1488config per port Rx offloading 1489~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ 1490 1491Enable or disable a per port Rx offloading on all Rx queues of a port:: 1492 1493 testpmd> port config (port_id) rx_offload (offloading) on|off 1494 1495* ``offloading``: can be any of these offloading capability: 1496 vlan_strip, ipv4_cksum, udp_cksum, tcp_cksum, tcp_lro, 1497 qinq_strip, outer_ipv4_cksum, macsec_strip, 1498 header_split, vlan_filter, vlan_extend, jumbo_frame, 1499 crc_strip, scatter, timestamp, security, keep_crc 1500 1501This command should be run when the port is stopped, or else it will fail. 1502 1503config per queue Rx offloading 1504~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ 1505 1506Enable or disable a per queue Rx offloading only on a specific Rx queue:: 1507 1508 testpmd> port (port_id) rxq (queue_id) rx_offload (offloading) on|off 1509 1510* ``offloading``: can be any of these offloading capability: 1511 vlan_strip, ipv4_cksum, udp_cksum, tcp_cksum, tcp_lro, 1512 qinq_strip, outer_ipv4_cksum, macsec_strip, 1513 header_split, vlan_filter, vlan_extend, jumbo_frame, 1514 crc_strip, scatter, timestamp, security, keep_crc 1515 1516This command should be run when the port is stopped, or else it will fail. 1517 1518config per port Tx offloading 1519~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ 1520 1521Enable or disable a per port Tx offloading on all Tx queues of a port:: 1522 1523 testpmd> port config (port_id) tx_offload (offloading) on|off 1524 1525* ``offloading``: can be any of these offloading capability: 1526 vlan_insert, ipv4_cksum, udp_cksum, tcp_cksum, 1527 sctp_cksum, tcp_tso, udp_tso, outer_ipv4_cksum, 1528 qinq_insert, vxlan_tnl_tso, gre_tnl_tso, 1529 ipip_tnl_tso, geneve_tnl_tso, macsec_insert, 1530 mt_lockfree, multi_segs, mbuf_fast_free, security 1531 1532This command should be run when the port is stopped, or else it will fail. 1533 1534config per queue Tx offloading 1535~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ 1536 1537Enable or disable a per queue Tx offloading only on a specific Tx queue:: 1538 1539 testpmd> port (port_id) txq (queue_id) tx_offload (offloading) on|off 1540 1541* ``offloading``: can be any of these offloading capability: 1542 vlan_insert, ipv4_cksum, udp_cksum, tcp_cksum, 1543 sctp_cksum, tcp_tso, udp_tso, outer_ipv4_cksum, 1544 qinq_insert, vxlan_tnl_tso, gre_tnl_tso, 1545 ipip_tnl_tso, geneve_tnl_tso, macsec_insert, 1546 mt_lockfree, multi_segs, mbuf_fast_free, security 1547 1548This command should be run when the port is stopped, or else it will fail. 1549 1550Config VXLAN Encap outer layers 1551~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ 1552 1553Configure the outer layer to encapsulate a packet inside a VXLAN tunnel:: 1554 1555 set vxlan ip-version (ipv4|ipv6) vni (vni) udp-src (udp-src) \ 1556 udp-dst (udp-dst) ip-src (ip-src) ip-dst (ip-dst) eth-src (eth-src) \ 1557 eth-dst (eth-dst) 1558 1559 set vxlan-with-vlan ip-version (ipv4|ipv6) vni (vni) udp-src (udp-src) \ 1560 udp-dst (udp-dst) ip-src (ip-src) ip-dst (ip-dst) vlan-tci (vlan-tci) \ 1561 eth-src (eth-src) eth-dst (eth-dst) 1562 1563Those command will set an internal configuration inside testpmd, any following 1564flow rule using the action vxlan_encap will use the last configuration set. 1565To have a different encapsulation header, one of those commands must be called 1566before the flow rule creation. 1567 1568Config NVGRE Encap outer layers 1569~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ 1570 1571Configure the outer layer to encapsulate a packet inside a NVGRE tunnel:: 1572 1573 set nvgre ip-version (ipv4|ipv6) tni (tni) ip-src (ip-src) ip-dst (ip-dst) \ 1574 eth-src (eth-src) eth-dst (eth-dst) 1575 set nvgre-with-vlan ip-version (ipv4|ipv6) tni (tni) ip-src (ip-src) \ 1576 ip-dst (ip-dst) vlan-tci (vlan-tci) eth-src (eth-src) eth-dst (eth-dst) 1577 1578Those command will set an internal configuration inside testpmd, any following 1579flow rule using the action nvgre_encap will use the last configuration set. 1580To have a different encapsulation header, one of those commands must be called 1581before the flow rule creation. 1582 1583Port Functions 1584-------------- 1585 1586The following sections show functions for configuring ports. 1587 1588.. note:: 1589 1590 Port configuration changes only become active when forwarding is started/restarted. 1591 1592port attach 1593~~~~~~~~~~~ 1594 1595Attach a port specified by pci address or virtual device args:: 1596 1597 testpmd> port attach (identifier) 1598 1599To attach a new pci device, the device should be recognized by kernel first. 1600Then it should be moved under DPDK management. 1601Finally the port can be attached to testpmd. 1602 1603For example, to move a pci device using ixgbe under DPDK management: 1604 1605.. code-block:: console 1606 1607 # Check the status of the available devices. 1608 ./usertools/dpdk-devbind.py --status 1609 1610 Network devices using DPDK-compatible driver 1611 ============================================ 1612 <none> 1613 1614 Network devices using kernel driver 1615 =================================== 1616 0000:0a:00.0 '82599ES 10-Gigabit' if=eth2 drv=ixgbe unused= 1617 1618 1619 # Bind the device to igb_uio. 1620 sudo ./usertools/dpdk-devbind.py -b igb_uio 0000:0a:00.0 1621 1622 1623 # Recheck the status of the devices. 1624 ./usertools/dpdk-devbind.py --status 1625 Network devices using DPDK-compatible driver 1626 ============================================ 1627 0000:0a:00.0 '82599ES 10-Gigabit' drv=igb_uio unused= 1628 1629To attach a port created by virtual device, above steps are not needed. 1630 1631For example, to attach a port whose pci address is 0000:0a:00.0. 1632 1633.. code-block:: console 1634 1635 testpmd> port attach 0000:0a:00.0 1636 Attaching a new port... 1637 EAL: PCI device 0000:0a:00.0 on NUMA socket -1 1638 EAL: probe driver: 8086:10fb rte_ixgbe_pmd 1639 EAL: PCI memory mapped at 0x7f83bfa00000 1640 EAL: PCI memory mapped at 0x7f83bfa80000 1641 PMD: eth_ixgbe_dev_init(): MAC: 2, PHY: 18, SFP+: 5 1642 PMD: eth_ixgbe_dev_init(): port 0 vendorID=0x8086 deviceID=0x10fb 1643 Port 0 is attached. Now total ports is 1 1644 Done 1645 1646For example, to attach a port created by pcap PMD. 1647 1648.. code-block:: console 1649 1650 testpmd> port attach net_pcap0 1651 Attaching a new port... 1652 PMD: Initializing pmd_pcap for net_pcap0 1653 PMD: Creating pcap-backed ethdev on numa socket 0 1654 Port 0 is attached. Now total ports is 1 1655 Done 1656 1657In this case, identifier is ``net_pcap0``. 1658This identifier format is the same as ``--vdev`` format of DPDK applications. 1659 1660For example, to re-attach a bonded port which has been previously detached, 1661the mode and slave parameters must be given. 1662 1663.. code-block:: console 1664 1665 testpmd> port attach net_bond_0,mode=0,slave=1 1666 Attaching a new port... 1667 EAL: Initializing pmd_bond for net_bond_0 1668 EAL: Create bonded device net_bond_0 on port 0 in mode 0 on socket 0. 1669 Port 0 is attached. Now total ports is 1 1670 Done 1671 1672 1673port detach 1674~~~~~~~~~~~ 1675 1676Detach a specific port:: 1677 1678 testpmd> port detach (port_id) 1679 1680Before detaching a port, the port should be stopped and closed. 1681 1682For example, to detach a pci device port 0. 1683 1684.. code-block:: console 1685 1686 testpmd> port stop 0 1687 Stopping ports... 1688 Done 1689 testpmd> port close 0 1690 Closing ports... 1691 Done 1692 1693 testpmd> port detach 0 1694 Detaching a port... 1695 EAL: PCI device 0000:0a:00.0 on NUMA socket -1 1696 EAL: remove driver: 8086:10fb rte_ixgbe_pmd 1697 EAL: PCI memory unmapped at 0x7f83bfa00000 1698 EAL: PCI memory unmapped at 0x7f83bfa80000 1699 Done 1700 1701 1702For example, to detach a virtual device port 0. 1703 1704.. code-block:: console 1705 1706 testpmd> port stop 0 1707 Stopping ports... 1708 Done 1709 testpmd> port close 0 1710 Closing ports... 1711 Done 1712 1713 testpmd> port detach 0 1714 Detaching a port... 1715 PMD: Closing pcap ethdev on numa socket 0 1716 Port 'net_pcap0' is detached. Now total ports is 0 1717 Done 1718 1719To remove a pci device completely from the system, first detach the port from testpmd. 1720Then the device should be moved under kernel management. 1721Finally the device can be removed using kernel pci hotplug functionality. 1722 1723For example, to move a pci device under kernel management: 1724 1725.. code-block:: console 1726 1727 sudo ./usertools/dpdk-devbind.py -b ixgbe 0000:0a:00.0 1728 1729 ./usertools/dpdk-devbind.py --status 1730 1731 Network devices using DPDK-compatible driver 1732 ============================================ 1733 <none> 1734 1735 Network devices using kernel driver 1736 =================================== 1737 0000:0a:00.0 '82599ES 10-Gigabit' if=eth2 drv=ixgbe unused=igb_uio 1738 1739To remove a port created by a virtual device, above steps are not needed. 1740 1741port start 1742~~~~~~~~~~ 1743 1744Start all ports or a specific port:: 1745 1746 testpmd> port start (port_id|all) 1747 1748port stop 1749~~~~~~~~~ 1750 1751Stop all ports or a specific port:: 1752 1753 testpmd> port stop (port_id|all) 1754 1755port close 1756~~~~~~~~~~ 1757 1758Close all ports or a specific port:: 1759 1760 testpmd> port close (port_id|all) 1761 1762port config - queue ring size 1763~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ 1764 1765Configure a rx/tx queue ring size:: 1766 1767 testpmd> port (port_id) (rxq|txq) (queue_id) ring_size (value) 1768 1769Only take effect after command that (re-)start the port or command that setup specific queue. 1770 1771port start/stop queue 1772~~~~~~~~~~~~~~~~~~~~~ 1773 1774Start/stop a rx/tx queue on a specific port:: 1775 1776 testpmd> port (port_id) (rxq|txq) (queue_id) (start|stop) 1777 1778port config - queue deferred start 1779~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ 1780 1781Switch on/off deferred start of a specific port queue:: 1782 1783 testpmd> port (port_id) (rxq|txq) (queue_id) deferred_start (on|off) 1784 1785port setup queue 1786~~~~~~~~~~~~~~~~~~~~~ 1787 1788Setup a rx/tx queue on a specific port:: 1789 1790 testpmd> port (port_id) (rxq|txq) (queue_id) setup 1791 1792Only take effect when port is started. 1793 1794port config - speed 1795~~~~~~~~~~~~~~~~~~~ 1796 1797Set the speed and duplex mode for all ports or a specific port:: 1798 1799 testpmd> port config (port_id|all) speed (10|100|1000|10000|25000|40000|50000|100000|auto) \ 1800 duplex (half|full|auto) 1801 1802port config - queues/descriptors 1803~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ 1804 1805Set number of queues/descriptors for rxq, txq, rxd and txd:: 1806 1807 testpmd> port config all (rxq|txq|rxd|txd) (value) 1808 1809This is equivalent to the ``--rxq``, ``--txq``, ``--rxd`` and ``--txd`` command-line options. 1810 1811port config - max-pkt-len 1812~~~~~~~~~~~~~~~~~~~~~~~~~ 1813 1814Set the maximum packet length:: 1815 1816 testpmd> port config all max-pkt-len (value) 1817 1818This is equivalent to the ``--max-pkt-len`` command-line option. 1819 1820port config - CRC Strip 1821~~~~~~~~~~~~~~~~~~~~~~~ 1822 1823Set hardware CRC stripping on or off for all ports:: 1824 1825 testpmd> port config all crc-strip (on|off) 1826 1827CRC stripping is on by default. 1828 1829The ``off`` option is equivalent to the ``--disable-crc-strip`` command-line option. 1830 1831port config - scatter 1832~~~~~~~~~~~~~~~~~~~~~~~ 1833 1834Set RX scatter mode on or off for all ports:: 1835 1836 testpmd> port config all scatter (on|off) 1837 1838RX scatter mode is off by default. 1839 1840The ``on`` option is equivalent to the ``--enable-scatter`` command-line option. 1841 1842port config - RX Checksum 1843~~~~~~~~~~~~~~~~~~~~~~~~~ 1844 1845Set hardware RX checksum offload to on or off for all ports:: 1846 1847 testpmd> port config all rx-cksum (on|off) 1848 1849Checksum offload is off by default. 1850 1851The ``on`` option is equivalent to the ``--enable-rx-cksum`` command-line option. 1852 1853port config - VLAN 1854~~~~~~~~~~~~~~~~~~ 1855 1856Set hardware VLAN on or off for all ports:: 1857 1858 testpmd> port config all hw-vlan (on|off) 1859 1860Hardware VLAN is off by default. 1861 1862The ``on`` option is equivalent to the ``--enable-hw-vlan`` command-line option. 1863 1864port config - VLAN filter 1865~~~~~~~~~~~~~~~~~~~~~~~~~ 1866 1867Set hardware VLAN filter on or off for all ports:: 1868 1869 testpmd> port config all hw-vlan-filter (on|off) 1870 1871Hardware VLAN filter is off by default. 1872 1873The ``on`` option is equivalent to the ``--enable-hw-vlan-filter`` command-line option. 1874 1875port config - VLAN strip 1876~~~~~~~~~~~~~~~~~~~~~~~~ 1877 1878Set hardware VLAN strip on or off for all ports:: 1879 1880 testpmd> port config all hw-vlan-strip (on|off) 1881 1882Hardware VLAN strip is off by default. 1883 1884The ``on`` option is equivalent to the ``--enable-hw-vlan-strip`` command-line option. 1885 1886port config - VLAN extend 1887~~~~~~~~~~~~~~~~~~~~~~~~~ 1888 1889Set hardware VLAN extend on or off for all ports:: 1890 1891 testpmd> port config all hw-vlan-extend (on|off) 1892 1893Hardware VLAN extend is off by default. 1894 1895The ``on`` option is equivalent to the ``--enable-hw-vlan-extend`` command-line option. 1896 1897port config - Drop Packets 1898~~~~~~~~~~~~~~~~~~~~~~~~~~ 1899 1900Set packet drop for packets with no descriptors on or off for all ports:: 1901 1902 testpmd> port config all drop-en (on|off) 1903 1904Packet dropping for packets with no descriptors is off by default. 1905 1906The ``on`` option is equivalent to the ``--enable-drop-en`` command-line option. 1907 1908port config - RSS 1909~~~~~~~~~~~~~~~~~ 1910 1911Set the RSS (Receive Side Scaling) mode on or off:: 1912 1913 testpmd> port config all rss (all|default|ip|tcp|udp|sctp|ether|port|vxlan|geneve|nvgre|none) 1914 1915RSS is on by default. 1916 1917The ``all`` option is equivalent to ip|tcp|udp|sctp|ether. 1918The ``default`` option enables all supported RSS types reported by device info. 1919The ``none`` option is equivalent to the ``--disable-rss`` command-line option. 1920 1921port config - RSS Reta 1922~~~~~~~~~~~~~~~~~~~~~~ 1923 1924Set the RSS (Receive Side Scaling) redirection table:: 1925 1926 testpmd> port config all rss reta (hash,queue)[,(hash,queue)] 1927 1928port config - DCB 1929~~~~~~~~~~~~~~~~~ 1930 1931Set the DCB mode for an individual port:: 1932 1933 testpmd> port config (port_id) dcb vt (on|off) (traffic_class) pfc (on|off) 1934 1935The traffic class should be 4 or 8. 1936 1937port config - Burst 1938~~~~~~~~~~~~~~~~~~~ 1939 1940Set the number of packets per burst:: 1941 1942 testpmd> port config all burst (value) 1943 1944This is equivalent to the ``--burst`` command-line option. 1945 1946port config - Threshold 1947~~~~~~~~~~~~~~~~~~~~~~~ 1948 1949Set thresholds for TX/RX queues:: 1950 1951 testpmd> port config all (threshold) (value) 1952 1953Where the threshold type can be: 1954 1955* ``txpt:`` Set the prefetch threshold register of the TX rings, 0 <= value <= 255. 1956 1957* ``txht:`` Set the host threshold register of the TX rings, 0 <= value <= 255. 1958 1959* ``txwt:`` Set the write-back threshold register of the TX rings, 0 <= value <= 255. 1960 1961* ``rxpt:`` Set the prefetch threshold register of the RX rings, 0 <= value <= 255. 1962 1963* ``rxht:`` Set the host threshold register of the RX rings, 0 <= value <= 255. 1964 1965* ``rxwt:`` Set the write-back threshold register of the RX rings, 0 <= value <= 255. 1966 1967* ``txfreet:`` Set the transmit free threshold of the TX rings, 0 <= value <= txd. 1968 1969* ``rxfreet:`` Set the transmit free threshold of the RX rings, 0 <= value <= rxd. 1970 1971* ``txrst:`` Set the transmit RS bit threshold of TX rings, 0 <= value <= txd. 1972 1973These threshold options are also available from the command-line. 1974 1975port config - E-tag 1976~~~~~~~~~~~~~~~~~~~ 1977 1978Set the value of ether-type for E-tag:: 1979 1980 testpmd> port config (port_id|all) l2-tunnel E-tag ether-type (value) 1981 1982Enable/disable the E-tag support:: 1983 1984 testpmd> port config (port_id|all) l2-tunnel E-tag (enable|disable) 1985 1986port config pctype mapping 1987~~~~~~~~~~~~~~~~~~~~~~~~~~ 1988 1989Reset pctype mapping table:: 1990 1991 testpmd> port config (port_id) pctype mapping reset 1992 1993Update hardware defined pctype to software defined flow type mapping table:: 1994 1995 testpmd> port config (port_id) pctype mapping update (pctype_id_0[,pctype_id_1]*) (flow_type_id) 1996 1997where: 1998 1999* ``pctype_id_x``: hardware pctype id as index of bit in bitmask value of the pctype mapping table. 2000 2001* ``flow_type_id``: software flow type id as the index of the pctype mapping table. 2002 2003port config input set 2004~~~~~~~~~~~~~~~~~~~~~ 2005 2006Config RSS/FDIR/FDIR flexible payload input set for some pctype:: 2007 testpmd> port config (port_id) pctype (pctype_id) \ 2008 (hash_inset|fdir_inset|fdir_flx_inset) \ 2009 (get|set|clear) field (field_idx) 2010 2011Clear 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) clear all 2014 2015where: 2016 2017* ``pctype_id``: hardware packet classification types. 2018* ``field_idx``: hardware field index. 2019 2020port config udp_tunnel_port 2021~~~~~~~~~~~~~~~~~~~~~~~~~~~ 2022 2023Add/remove UDP tunnel port for VXLAN/GENEVE tunneling protocols:: 2024 testpmd> port config (port_id) udp_tunnel_port add|rm vxlan|geneve (udp_port) 2025 2026Link Bonding Functions 2027---------------------- 2028 2029The Link Bonding functions make it possible to dynamically create and 2030manage link bonding devices from within testpmd interactive prompt. 2031 2032create bonded device 2033~~~~~~~~~~~~~~~~~~~~ 2034 2035Create a new bonding device:: 2036 2037 testpmd> create bonded device (mode) (socket) 2038 2039For example, to create a bonded device in mode 1 on socket 0:: 2040 2041 testpmd> create bonded device 1 0 2042 created new bonded device (port X) 2043 2044add bonding slave 2045~~~~~~~~~~~~~~~~~ 2046 2047Adds Ethernet device to a Link Bonding device:: 2048 2049 testpmd> add bonding slave (slave id) (port id) 2050 2051For example, to add Ethernet device (port 6) to a Link Bonding device (port 10):: 2052 2053 testpmd> add bonding slave 6 10 2054 2055 2056remove bonding slave 2057~~~~~~~~~~~~~~~~~~~~ 2058 2059Removes an Ethernet slave device from a Link Bonding device:: 2060 2061 testpmd> remove bonding slave (slave id) (port id) 2062 2063For example, to remove Ethernet slave device (port 6) to a Link Bonding device (port 10):: 2064 2065 testpmd> remove bonding slave 6 10 2066 2067set bonding mode 2068~~~~~~~~~~~~~~~~ 2069 2070Set the Link Bonding mode of a Link Bonding device:: 2071 2072 testpmd> set bonding mode (value) (port id) 2073 2074For example, to set the bonding mode of a Link Bonding device (port 10) to broadcast (mode 3):: 2075 2076 testpmd> set bonding mode 3 10 2077 2078set bonding primary 2079~~~~~~~~~~~~~~~~~~~ 2080 2081Set an Ethernet slave device as the primary device on a Link Bonding device:: 2082 2083 testpmd> set bonding primary (slave id) (port id) 2084 2085For example, to set the Ethernet slave device (port 6) as the primary port of a Link Bonding device (port 10):: 2086 2087 testpmd> set bonding primary 6 10 2088 2089set bonding mac 2090~~~~~~~~~~~~~~~ 2091 2092Set the MAC address of a Link Bonding device:: 2093 2094 testpmd> set bonding mac (port id) (mac) 2095 2096For example, to set the MAC address of a Link Bonding device (port 10) to 00:00:00:00:00:01:: 2097 2098 testpmd> set bonding mac 10 00:00:00:00:00:01 2099 2100set bonding xmit_balance_policy 2101~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ 2102 2103Set the transmission policy for a Link Bonding device when it is in Balance XOR mode:: 2104 2105 testpmd> set bonding xmit_balance_policy (port_id) (l2|l23|l34) 2106 2107For example, set a Link Bonding device (port 10) to use a balance policy of layer 3+4 (IP addresses & UDP ports):: 2108 2109 testpmd> set bonding xmit_balance_policy 10 l34 2110 2111 2112set bonding mon_period 2113~~~~~~~~~~~~~~~~~~~~~~ 2114 2115Set the link status monitoring polling period in milliseconds for a bonding device. 2116 2117This adds support for PMD slave devices which do not support link status interrupts. 2118When the mon_period is set to a value greater than 0 then all PMD's which do not support 2119link status ISR will be queried every polling interval to check if their link status has changed:: 2120 2121 testpmd> set bonding mon_period (port_id) (value) 2122 2123For example, to set the link status monitoring polling period of bonded device (port 5) to 150ms:: 2124 2125 testpmd> set bonding mon_period 5 150 2126 2127 2128set bonding lacp dedicated_queue 2129~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ 2130 2131Enable dedicated tx/rx queues on bonding devices slaves to handle LACP control plane traffic 2132when in mode 4 (link-aggregration-802.3ad):: 2133 2134 testpmd> set bonding lacp dedicated_queues (port_id) (enable|disable) 2135 2136 2137set bonding agg_mode 2138~~~~~~~~~~~~~~~~~~~~ 2139 2140Enable one of the specific aggregators mode when in mode 4 (link-aggregration-802.3ad):: 2141 2142 testpmd> set bonding agg_mode (port_id) (bandwidth|count|stable) 2143 2144 2145show bonding config 2146~~~~~~~~~~~~~~~~~~~ 2147 2148Show the current configuration of a Link Bonding device:: 2149 2150 testpmd> show bonding config (port id) 2151 2152For example, 2153to show the configuration a Link Bonding device (port 9) with 3 slave devices (1, 3, 4) 2154in balance mode with a transmission policy of layer 2+3:: 2155 2156 testpmd> show bonding config 9 2157 Bonding mode: 2 2158 Balance Xmit Policy: BALANCE_XMIT_POLICY_LAYER23 2159 Slaves (3): [1 3 4] 2160 Active Slaves (3): [1 3 4] 2161 Primary: [3] 2162 2163 2164Register Functions 2165------------------ 2166 2167The Register Functions can be used to read from and write to registers on the network card referenced by a port number. 2168This is mainly useful for debugging purposes. 2169Reference should be made to the appropriate datasheet for the network card for details on the register addresses 2170and fields that can be accessed. 2171 2172read reg 2173~~~~~~~~ 2174 2175Display the value of a port register:: 2176 2177 testpmd> read reg (port_id) (address) 2178 2179For example, to examine the Flow Director control register (FDIRCTL, 0x0000EE000) on an Intel 82599 10 GbE Controller:: 2180 2181 testpmd> read reg 0 0xEE00 2182 port 0 PCI register at offset 0xEE00: 0x4A060029 (1241907241) 2183 2184read regfield 2185~~~~~~~~~~~~~ 2186 2187Display a port register bit field:: 2188 2189 testpmd> read regfield (port_id) (address) (bit_x) (bit_y) 2190 2191For example, reading the lowest two bits from the register in the example above:: 2192 2193 testpmd> read regfield 0 0xEE00 0 1 2194 port 0 PCI register at offset 0xEE00: bits[0, 1]=0x1 (1) 2195 2196read regbit 2197~~~~~~~~~~~ 2198 2199Display a single port register bit:: 2200 2201 testpmd> read regbit (port_id) (address) (bit_x) 2202 2203For example, reading the lowest bit from the register in the example above:: 2204 2205 testpmd> read regbit 0 0xEE00 0 2206 port 0 PCI register at offset 0xEE00: bit 0=1 2207 2208write reg 2209~~~~~~~~~ 2210 2211Set the value of a port register:: 2212 2213 testpmd> write reg (port_id) (address) (value) 2214 2215For example, to clear a register:: 2216 2217 testpmd> write reg 0 0xEE00 0x0 2218 port 0 PCI register at offset 0xEE00: 0x00000000 (0) 2219 2220write regfield 2221~~~~~~~~~~~~~~ 2222 2223Set bit field of a port register:: 2224 2225 testpmd> write regfield (port_id) (address) (bit_x) (bit_y) (value) 2226 2227For example, writing to the register cleared in the example above:: 2228 2229 testpmd> write regfield 0 0xEE00 0 1 2 2230 port 0 PCI register at offset 0xEE00: 0x00000002 (2) 2231 2232write regbit 2233~~~~~~~~~~~~ 2234 2235Set single bit value of a port register:: 2236 2237 testpmd> write regbit (port_id) (address) (bit_x) (value) 2238 2239For example, to set the high bit in the register from the example above:: 2240 2241 testpmd> write regbit 0 0xEE00 31 1 2242 port 0 PCI register at offset 0xEE00: 0x8000000A (2147483658) 2243 2244Traffic Metering and Policing 2245----------------------------- 2246 2247The following section shows functions for configuring traffic metering and 2248policing on the ethernet device through the use of generic ethdev API. 2249 2250show port traffic management capability 2251~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ 2252 2253Show traffic metering and policing capability of the port:: 2254 2255 testpmd> show port meter cap (port_id) 2256 2257add port meter profile (srTCM rfc2967) 2258~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ 2259 2260Add meter profile (srTCM rfc2697) to the ethernet device:: 2261 2262 testpmd> add port meter profile srtcm_rfc2697 (port_id) (profile_id) \ 2263 (cir) (cbs) (ebs) 2264 2265where: 2266 2267* ``profile_id``: ID for the meter profile. 2268* ``cir``: Committed Information Rate (CIR) (bytes/second). 2269* ``cbs``: Committed Burst Size (CBS) (bytes). 2270* ``ebs``: Excess Burst Size (EBS) (bytes). 2271 2272add port meter profile (trTCM rfc2968) 2273~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ 2274 2275Add meter profile (srTCM rfc2698) to the ethernet device:: 2276 2277 testpmd> add port meter profile trtcm_rfc2698 (port_id) (profile_id) \ 2278 (cir) (pir) (cbs) (pbs) 2279 2280where: 2281 2282* ``profile_id``: ID for the meter profile. 2283* ``cir``: Committed information rate (bytes/second). 2284* ``pir``: Peak information rate (bytes/second). 2285* ``cbs``: Committed burst size (bytes). 2286* ``pbs``: Peak burst size (bytes). 2287 2288add port meter profile (trTCM rfc4115) 2289~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ 2290 2291Add meter profile (trTCM rfc4115) to the ethernet device:: 2292 2293 testpmd> add port meter profile trtcm_rfc4115 (port_id) (profile_id) \ 2294 (cir) (eir) (cbs) (ebs) 2295 2296where: 2297 2298* ``profile_id``: ID for the meter profile. 2299* ``cir``: Committed information rate (bytes/second). 2300* ``eir``: Excess information rate (bytes/second). 2301* ``cbs``: Committed burst size (bytes). 2302* ``ebs``: Excess burst size (bytes). 2303 2304delete port meter profile 2305~~~~~~~~~~~~~~~~~~~~~~~~~ 2306 2307Delete meter profile from the ethernet device:: 2308 2309 testpmd> del port meter profile (port_id) (profile_id) 2310 2311create port meter 2312~~~~~~~~~~~~~~~~~ 2313 2314Create new meter object for the ethernet device:: 2315 2316 testpmd> create port meter (port_id) (mtr_id) (profile_id) \ 2317 (meter_enable) (g_action) (y_action) (r_action) (stats_mask) (shared) \ 2318 (use_pre_meter_color) [(dscp_tbl_entry0) (dscp_tbl_entry1)...\ 2319 (dscp_tbl_entry63)] 2320 2321where: 2322 2323* ``mtr_id``: meter object ID. 2324* ``profile_id``: ID for the meter profile. 2325* ``meter_enable``: When this parameter has a non-zero value, the meter object 2326 gets enabled at the time of creation, otherwise remains disabled. 2327* ``g_action``: Policer action for the packet with green color. 2328* ``y_action``: Policer action for the packet with yellow color. 2329* ``r_action``: Policer action for the packet with red color. 2330* ``stats_mask``: Mask of statistics counter types to be enabled for the 2331 meter object. 2332* ``shared``: When this parameter has a non-zero value, the meter object is 2333 shared by multiple flows. Otherwise, meter object is used by single flow. 2334* ``use_pre_meter_color``: When this parameter has a non-zero value, the 2335 input color for the current meter object is determined by the latest meter 2336 object in the same flow. Otherwise, the current meter object uses the 2337 *dscp_table* to determine the input color. 2338* ``dscp_tbl_entryx``: DSCP table entry x providing meter providing input 2339 color, 0 <= x <= 63. 2340 2341enable port meter 2342~~~~~~~~~~~~~~~~~ 2343 2344Enable meter for the ethernet device:: 2345 2346 testpmd> enable port meter (port_id) (mtr_id) 2347 2348disable port meter 2349~~~~~~~~~~~~~~~~~~ 2350 2351Disable meter for the ethernet device:: 2352 2353 testpmd> disable port meter (port_id) (mtr_id) 2354 2355delete port meter 2356~~~~~~~~~~~~~~~~~ 2357 2358Delete meter for the ethernet device:: 2359 2360 testpmd> del port meter (port_id) (mtr_id) 2361 2362Set port meter profile 2363~~~~~~~~~~~~~~~~~~~~~~ 2364 2365Set meter profile for the ethernet device:: 2366 2367 testpmd> set port meter profile (port_id) (mtr_id) (profile_id) 2368 2369set port meter dscp table 2370~~~~~~~~~~~~~~~~~~~~~~~~~ 2371 2372Set meter dscp table for the ethernet device:: 2373 2374 testpmd> set port meter dscp table (port_id) (mtr_id) [(dscp_tbl_entry0) \ 2375 (dscp_tbl_entry1)...(dscp_tbl_entry63)] 2376 2377set port meter policer action 2378~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ 2379 2380Set meter policer action for the ethernet device:: 2381 2382 testpmd> set port meter policer action (port_id) (mtr_id) (action_mask) \ 2383 (action0) [(action1) (action1)] 2384 2385where: 2386 2387* ``action_mask``: Bit mask indicating which policer actions need to be 2388 updated. One or more policer actions can be updated in a single function 2389 invocation. To update the policer action associated with color C, bit 2390 (1 << C) needs to be set in *action_mask* and element at position C 2391 in the *actions* array needs to be valid. 2392* ``actionx``: Policer action for the color x, 2393 RTE_MTR_GREEN <= x < RTE_MTR_COLORS 2394 2395set port meter stats mask 2396~~~~~~~~~~~~~~~~~~~~~~~~~ 2397 2398Set meter stats mask for the ethernet device:: 2399 2400 testpmd> set port meter stats mask (port_id) (mtr_id) (stats_mask) 2401 2402where: 2403 2404* ``stats_mask``: Bit mask indicating statistics counter types to be enabled. 2405 2406show port meter stats 2407~~~~~~~~~~~~~~~~~~~~~ 2408 2409Show meter stats of the ethernet device:: 2410 2411 testpmd> show port meter stats (port_id) (mtr_id) (clear) 2412 2413where: 2414 2415* ``clear``: Flag that indicates whether the statistics counters should 2416 be cleared (i.e. set to zero) immediately after they have been read or not. 2417 2418Traffic Management 2419------------------ 2420 2421The following section shows functions for configuring traffic management on 2422on the ethernet device through the use of generic TM API. 2423 2424show port traffic management capability 2425~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ 2426 2427Show traffic management capability of the port:: 2428 2429 testpmd> show port tm cap (port_id) 2430 2431show port traffic management capability (hierarchy level) 2432~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ 2433 2434Show traffic management hierarchy level capability of the port:: 2435 2436 testpmd> show port tm level cap (port_id) (level_id) 2437 2438show port traffic management capability (hierarchy node level) 2439~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ 2440 2441Show the traffic management hierarchy node capability of the port:: 2442 2443 testpmd> show port tm node cap (port_id) (node_id) 2444 2445show port traffic management hierarchy node type 2446~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ 2447 2448Show the port traffic management hierarchy node type:: 2449 2450 testpmd> show port tm node type (port_id) (node_id) 2451 2452show port traffic management hierarchy node stats 2453~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ 2454 2455Show the port traffic management hierarchy node statistics:: 2456 2457 testpmd> show port tm node stats (port_id) (node_id) (clear) 2458 2459where: 2460 2461* ``clear``: When this parameter has a non-zero value, the statistics counters 2462 are cleared (i.e. set to zero) immediately after they have been read, 2463 otherwise the statistics counters are left untouched. 2464 2465Add port traffic management private shaper profile 2466~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ 2467 2468Add the port traffic management private shaper profile:: 2469 2470 testpmd> add port tm node shaper profile (port_id) (shaper_profile_id) \ 2471 (tb_rate) (tb_size) (packet_length_adjust) 2472 2473where: 2474 2475* ``shaper_profile id``: Shaper profile ID for the new profile. 2476* ``tb_rate``: Token bucket rate (bytes per second). 2477* ``tb_size``: Token bucket size (bytes). 2478* ``packet_length_adjust``: The value (bytes) to be added to the length of 2479 each packet for the purpose of shaping. This parameter value can be used to 2480 correct the packet length with the framing overhead bytes that are consumed 2481 on the wire. 2482 2483Delete port traffic management private shaper profile 2484~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ 2485 2486Delete the port traffic management private shaper:: 2487 2488 testpmd> del port tm node shaper profile (port_id) (shaper_profile_id) 2489 2490where: 2491 2492* ``shaper_profile id``: Shaper profile ID that needs to be deleted. 2493 2494Add port traffic management shared shaper 2495~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ 2496 2497Create the port traffic management shared shaper:: 2498 2499 testpmd> add port tm node shared shaper (port_id) (shared_shaper_id) \ 2500 (shaper_profile_id) 2501 2502where: 2503 2504* ``shared_shaper_id``: Shared shaper ID to be created. 2505* ``shaper_profile id``: Shaper profile ID for shared shaper. 2506 2507Set port traffic management shared shaper 2508~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ 2509 2510Update the port traffic management shared shaper:: 2511 2512 testpmd> set port tm node shared shaper (port_id) (shared_shaper_id) \ 2513 (shaper_profile_id) 2514 2515where: 2516 2517* ``shared_shaper_id``: Shared shaper ID to be update. 2518* ``shaper_profile id``: Shaper profile ID for shared shaper. 2519 2520Delete port traffic management shared shaper 2521~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ 2522 2523Delete the port traffic management shared shaper:: 2524 2525 testpmd> del port tm node shared shaper (port_id) (shared_shaper_id) 2526 2527where: 2528 2529* ``shared_shaper_id``: Shared shaper ID to be deleted. 2530 2531Set port traffic management hiearchy node private shaper 2532~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ 2533 2534set the port traffic management hierarchy node private shaper:: 2535 2536 testpmd> set port tm node shaper profile (port_id) (node_id) \ 2537 (shaper_profile_id) 2538 2539where: 2540 2541* ``shaper_profile id``: Private shaper profile ID to be enabled on the 2542 hierarchy node. 2543 2544Add port traffic management WRED profile 2545~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ 2546 2547Create a new WRED profile:: 2548 2549 testpmd> add port tm node wred profile (port_id) (wred_profile_id) \ 2550 (color_g) (min_th_g) (max_th_g) (maxp_inv_g) (wq_log2_g) \ 2551 (color_y) (min_th_y) (max_th_y) (maxp_inv_y) (wq_log2_y) \ 2552 (color_r) (min_th_r) (max_th_r) (maxp_inv_r) (wq_log2_r) 2553 2554where: 2555 2556* ``wred_profile id``: Identifier for the newly create WRED profile 2557* ``color_g``: Packet color (green) 2558* ``min_th_g``: Minimum queue threshold for packet with green color 2559* ``max_th_g``: Minimum queue threshold for packet with green color 2560* ``maxp_inv_g``: Inverse of packet marking probability maximum value (maxp) 2561* ``wq_log2_g``: Negated log2 of queue weight (wq) 2562* ``color_y``: Packet color (yellow) 2563* ``min_th_y``: Minimum queue threshold for packet with yellow color 2564* ``max_th_y``: Minimum queue threshold for packet with yellow color 2565* ``maxp_inv_y``: Inverse of packet marking probability maximum value (maxp) 2566* ``wq_log2_y``: Negated log2 of queue weight (wq) 2567* ``color_r``: Packet color (red) 2568* ``min_th_r``: Minimum queue threshold for packet with yellow color 2569* ``max_th_r``: Minimum queue threshold for packet with yellow color 2570* ``maxp_inv_r``: Inverse of packet marking probability maximum value (maxp) 2571* ``wq_log2_r``: Negated log2 of queue weight (wq) 2572 2573Delete port traffic management WRED profile 2574~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ 2575 2576Delete the WRED profile:: 2577 2578 testpmd> del port tm node wred profile (port_id) (wred_profile_id) 2579 2580Add port traffic management hierarchy nonleaf node 2581~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ 2582 2583Add nonleaf node to port traffic management hiearchy:: 2584 2585 testpmd> add port tm nonleaf node (port_id) (node_id) (parent_node_id) \ 2586 (priority) (weight) (level_id) (shaper_profile_id) \ 2587 (n_sp_priorities) (stats_mask) (n_shared_shapers) \ 2588 [(shared_shaper_0) (shared_shaper_1) ...] \ 2589 2590where: 2591 2592* ``parent_node_id``: Node ID of the parent. 2593* ``priority``: Node priority (highest node priority is zero). This is used by 2594 the SP algorithm running on the parent node for scheduling this node. 2595* ``weight``: Node weight (lowest weight is one). The node weight is relative 2596 to the weight sum of all siblings that have the same priority. It is used by 2597 the WFQ algorithm running on the parent node for scheduling this node. 2598* ``level_id``: Hiearchy level of the node. 2599* ``shaper_profile_id``: Shaper profile ID of the private shaper to be used by 2600 the node. 2601* ``n_sp_priorities``: Number of strict priorities. 2602* ``stats_mask``: Mask of statistics counter types to be enabled for this node. 2603* ``n_shared_shapers``: Number of shared shapers. 2604* ``shared_shaper_id``: Shared shaper id. 2605 2606Add port traffic management hierarchy leaf node 2607~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ 2608 2609Add leaf node to port traffic management hiearchy:: 2610 2611 testpmd> add port tm leaf node (port_id) (node_id) (parent_node_id) \ 2612 (priority) (weight) (level_id) (shaper_profile_id) \ 2613 (cman_mode) (wred_profile_id) (stats_mask) (n_shared_shapers) \ 2614 [(shared_shaper_id) (shared_shaper_id) ...] \ 2615 2616where: 2617 2618* ``parent_node_id``: Node ID of the parent. 2619* ``priority``: Node priority (highest node priority is zero). This is used by 2620 the SP algorithm running on the parent node for scheduling this node. 2621* ``weight``: Node weight (lowest weight is one). The node weight is relative 2622 to the weight sum of all siblings that have the same priority. It is used by 2623 the WFQ algorithm running on the parent node for scheduling this node. 2624* ``level_id``: Hiearchy level of the node. 2625* ``shaper_profile_id``: Shaper profile ID of the private shaper to be used by 2626 the node. 2627* ``cman_mode``: Congestion management mode to be enabled for this node. 2628* ``wred_profile_id``: WRED profile id to be enabled for this node. 2629* ``stats_mask``: Mask of statistics counter types to be enabled for this node. 2630* ``n_shared_shapers``: Number of shared shapers. 2631* ``shared_shaper_id``: Shared shaper id. 2632 2633Delete port traffic management hierarchy node 2634~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ 2635 2636Delete node from port traffic management hiearchy:: 2637 2638 testpmd> del port tm node (port_id) (node_id) 2639 2640Update port traffic management hierarchy parent node 2641~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ 2642 2643Update port traffic management hierarchy parent node:: 2644 2645 testpmd> set port tm node parent (port_id) (node_id) (parent_node_id) \ 2646 (priority) (weight) 2647 2648This function can only be called after the hierarchy commit invocation. Its 2649success depends on the port support for this operation, as advertised through 2650the port capability set. This function is valid for all nodes of the traffic 2651management hierarchy except root node. 2652 2653Suspend port traffic management hierarchy node 2654~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ 2655 2656 testpmd> suspend port tm node (port_id) (node_id) 2657 2658Resume port traffic management hierarchy node 2659~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ 2660 2661 testpmd> resume port tm node (port_id) (node_id) 2662 2663Commit port traffic management hierarchy 2664~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ 2665 2666Commit the traffic management hierarchy on the port:: 2667 2668 testpmd> port tm hierarchy commit (port_id) (clean_on_fail) 2669 2670where: 2671 2672* ``clean_on_fail``: When set to non-zero, hierarchy is cleared on function 2673 call failure. On the other hand, hierarchy is preserved when this parameter 2674 is equal to zero. 2675 2676Set port traffic management mark VLAN dei 2677~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ 2678 2679Enables/Disables the traffic management marking on the port for VLAN packets:: 2680 2681 testpmd> set port tm mark vlan_dei <port_id> <green> <yellow> <red> 2682 2683where: 2684 2685* ``port_id``: The port which on which VLAN packets marked as ``green`` or 2686 ``yellow`` or ``red`` will have dei bit enabled 2687 2688* ``green`` enable 1, disable 0 marking for dei bit of VLAN packets marked as green 2689 2690* ``yellow`` enable 1, disable 0 marking for dei bit of VLAN packets marked as yellow 2691 2692* ``red`` enable 1, disable 0 marking for dei bit of VLAN packets marked as red 2693 2694Set port traffic management mark IP dscp 2695~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ 2696 2697Enables/Disables the traffic management marking on the port for IP dscp packets:: 2698 2699 testpmd> set port tm mark ip_dscp <port_id> <green> <yellow> <red> 2700 2701where: 2702 2703* ``port_id``: The port which on which IP packets marked as ``green`` or 2704 ``yellow`` or ``red`` will have IP dscp bits updated 2705 2706* ``green`` enable 1, disable 0 marking IP dscp to low drop precedence for green packets 2707 2708* ``yellow`` enable 1, disable 0 marking IP dscp to medium drop precedence for yellow packets 2709 2710* ``red`` enable 1, disable 0 marking IP dscp to high drop precedence for red packets 2711 2712Set port traffic management mark IP ecn 2713~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ 2714 2715Enables/Disables the traffic management marking on the port for IP ecn packets:: 2716 2717 testpmd> set port tm mark ip_ecn <port_id> <green> <yellow> <red> 2718 2719where: 2720 2721* ``port_id``: The port which on which IP packets marked as ``green`` or 2722 ``yellow`` or ``red`` will have IP ecn bits updated 2723 2724* ``green`` enable 1, disable 0 marking IP ecn for green marked packets with ecn of 2'b01 or 2'b10 2725 to ecn of 2'b11 when IP is caring TCP or SCTP 2726 2727* ``yellow`` enable 1, disable 0 marking IP ecn for yellow marked packets with ecn of 2'b01 or 2'b10 2728 to ecn of 2'b11 when IP is caring TCP or SCTP 2729 2730* ``red`` enable 1, disable 0 marking IP ecn for yellow marked packets with ecn of 2'b01 or 2'b10 2731 to ecn of 2'b11 when IP is caring TCP or SCTP 2732 2733Set port traffic management default hierarchy (softnic forwarding mode) 2734~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ 2735 2736set the traffic management default hierarchy on the port:: 2737 2738 testpmd> set port tm hierarchy default (port_id) 2739 2740Filter Functions 2741---------------- 2742 2743This section details the available filter functions that are available. 2744 2745Note these functions interface the deprecated legacy filtering framework, 2746superseded by *rte_flow*. See `Flow rules management`_. 2747 2748ethertype_filter 2749~~~~~~~~~~~~~~~~~~~~ 2750 2751Add or delete a L2 Ethertype filter, which identify packets by their L2 Ethertype mainly assign them to a receive queue:: 2752 2753 ethertype_filter (port_id) (add|del) (mac_addr|mac_ignr) (mac_address) \ 2754 ethertype (ether_type) (drop|fwd) queue (queue_id) 2755 2756The available information parameters are: 2757 2758* ``port_id``: The port which the Ethertype filter assigned on. 2759 2760* ``mac_addr``: Compare destination mac address. 2761 2762* ``mac_ignr``: Ignore destination mac address match. 2763 2764* ``mac_address``: Destination mac address to match. 2765 2766* ``ether_type``: The EtherType value want to match, 2767 for example 0x0806 for ARP packet. 0x0800 (IPv4) and 0x86DD (IPv6) are invalid. 2768 2769* ``queue_id``: The receive queue associated with this EtherType filter. 2770 It is meaningless when deleting or dropping. 2771 2772Example, to add/remove an ethertype filter rule:: 2773 2774 testpmd> ethertype_filter 0 add mac_ignr 00:11:22:33:44:55 \ 2775 ethertype 0x0806 fwd queue 3 2776 2777 testpmd> ethertype_filter 0 del mac_ignr 00:11:22:33:44:55 \ 2778 ethertype 0x0806 fwd queue 3 2779 27802tuple_filter 2781~~~~~~~~~~~~~~~~~ 2782 2783Add or delete a 2-tuple filter, 2784which identifies packets by specific protocol and destination TCP/UDP port 2785and forwards packets into one of the receive queues:: 2786 2787 2tuple_filter (port_id) (add|del) dst_port (dst_port_value) \ 2788 protocol (protocol_value) mask (mask_value) \ 2789 tcp_flags (tcp_flags_value) priority (prio_value) \ 2790 queue (queue_id) 2791 2792The available information parameters are: 2793 2794* ``port_id``: The port which the 2-tuple filter assigned on. 2795 2796* ``dst_port_value``: Destination port in L4. 2797 2798* ``protocol_value``: IP L4 protocol. 2799 2800* ``mask_value``: Participates in the match or not by bit for field above, 1b means participate. 2801 2802* ``tcp_flags_value``: TCP control bits. The non-zero value is invalid, when the pro_value is not set to 0x06 (TCP). 2803 2804* ``prio_value``: Priority of this filter. 2805 2806* ``queue_id``: The receive queue associated with this 2-tuple filter. 2807 2808Example, to add/remove an 2tuple filter rule:: 2809 2810 testpmd> 2tuple_filter 0 add dst_port 32 protocol 0x06 mask 0x03 \ 2811 tcp_flags 0x02 priority 3 queue 3 2812 2813 testpmd> 2tuple_filter 0 del dst_port 32 protocol 0x06 mask 0x03 \ 2814 tcp_flags 0x02 priority 3 queue 3 2815 28165tuple_filter 2817~~~~~~~~~~~~~~~~~ 2818 2819Add or delete a 5-tuple filter, 2820which consists of a 5-tuple (protocol, source and destination IP addresses, source and destination TCP/UDP/SCTP port) 2821and routes packets into one of the receive queues:: 2822 2823 5tuple_filter (port_id) (add|del) dst_ip (dst_address) src_ip \ 2824 (src_address) dst_port (dst_port_value) \ 2825 src_port (src_port_value) protocol (protocol_value) \ 2826 mask (mask_value) tcp_flags (tcp_flags_value) \ 2827 priority (prio_value) queue (queue_id) 2828 2829The available information parameters are: 2830 2831* ``port_id``: The port which the 5-tuple filter assigned on. 2832 2833* ``dst_address``: Destination IP address. 2834 2835* ``src_address``: Source IP address. 2836 2837* ``dst_port_value``: TCP/UDP destination port. 2838 2839* ``src_port_value``: TCP/UDP source port. 2840 2841* ``protocol_value``: L4 protocol. 2842 2843* ``mask_value``: Participates in the match or not by bit for field above, 1b means participate 2844 2845* ``tcp_flags_value``: TCP control bits. The non-zero value is invalid, when the protocol_value is not set to 0x06 (TCP). 2846 2847* ``prio_value``: The priority of this filter. 2848 2849* ``queue_id``: The receive queue associated with this 5-tuple filter. 2850 2851Example, to add/remove an 5tuple filter rule:: 2852 2853 testpmd> 5tuple_filter 0 add dst_ip 2.2.2.5 src_ip 2.2.2.4 \ 2854 dst_port 64 src_port 32 protocol 0x06 mask 0x1F \ 2855 flags 0x0 priority 3 queue 3 2856 2857 testpmd> 5tuple_filter 0 del dst_ip 2.2.2.5 src_ip 2.2.2.4 \ 2858 dst_port 64 src_port 32 protocol 0x06 mask 0x1F \ 2859 flags 0x0 priority 3 queue 3 2860 2861syn_filter 2862~~~~~~~~~~ 2863 2864Using the SYN filter, TCP packets whose *SYN* flag is set can be forwarded to a separate queue:: 2865 2866 syn_filter (port_id) (add|del) priority (high|low) queue (queue_id) 2867 2868The available information parameters are: 2869 2870* ``port_id``: The port which the SYN filter assigned on. 2871 2872* ``high``: This SYN filter has higher priority than other filters. 2873 2874* ``low``: This SYN filter has lower priority than other filters. 2875 2876* ``queue_id``: The receive queue associated with this SYN filter 2877 2878Example:: 2879 2880 testpmd> syn_filter 0 add priority high queue 3 2881 2882flex_filter 2883~~~~~~~~~~~ 2884 2885With flex filter, packets can be recognized by any arbitrary pattern within the first 128 bytes of the packet 2886and routed into one of the receive queues:: 2887 2888 flex_filter (port_id) (add|del) len (len_value) bytes (bytes_value) \ 2889 mask (mask_value) priority (prio_value) queue (queue_id) 2890 2891The available information parameters are: 2892 2893* ``port_id``: The port which the Flex filter is assigned on. 2894 2895* ``len_value``: Filter length in bytes, no greater than 128. 2896 2897* ``bytes_value``: A string in hexadecimal, means the value the flex filter needs to match. 2898 2899* ``mask_value``: A string in hexadecimal, bit 1 means corresponding byte participates in the match. 2900 2901* ``prio_value``: The priority of this filter. 2902 2903* ``queue_id``: The receive queue associated with this Flex filter. 2904 2905Example:: 2906 2907 testpmd> flex_filter 0 add len 16 bytes 0x00000000000000000000000008060000 \ 2908 mask 000C priority 3 queue 3 2909 2910 testpmd> flex_filter 0 del len 16 bytes 0x00000000000000000000000008060000 \ 2911 mask 000C priority 3 queue 3 2912 2913 2914.. _testpmd_flow_director: 2915 2916flow_director_filter 2917~~~~~~~~~~~~~~~~~~~~ 2918 2919The Flow Director works in receive mode to identify specific flows or sets of flows and route them to specific queues. 2920 2921Four types of filtering are supported which are referred to as Perfect Match, Signature, Perfect-mac-vlan and 2922Perfect-tunnel filters, the match mode is set by the ``--pkt-filter-mode`` command-line parameter: 2923 2924* Perfect match filters. 2925 The hardware checks a match between the masked fields of the received packets and the programmed filters. 2926 The masked fields are for IP flow. 2927 2928* Signature filters. 2929 The hardware checks a match between a hash-based signature of the masked fields of the received packet. 2930 2931* Perfect-mac-vlan match filters. 2932 The hardware checks a match between the masked fields of the received packets and the programmed filters. 2933 The masked fields are for MAC VLAN flow. 2934 2935* Perfect-tunnel match filters. 2936 The hardware checks a match between the masked fields of the received packets and the programmed filters. 2937 The masked fields are for tunnel flow. 2938 2939* Perfect-raw-flow-type match filters. 2940 The hardware checks a match between the masked fields of the received packets and pre-loaded raw (template) packet. 2941 The masked fields are specified by input sets. 2942 2943The Flow Director filters can match the different fields for different type of packet: flow type, specific input set 2944per flow type and the flexible payload. 2945 2946The Flow Director can also mask out parts of all of these fields so that filters 2947are only applied to certain fields or parts of the fields. 2948 2949Note that for raw flow type mode the source and destination fields in the 2950raw packet buffer need to be presented in a reversed order with respect 2951to the expected received packets. 2952For example: IP source and destination addresses or TCP/UDP/SCTP 2953source and destination ports 2954 2955Different NICs may have different capabilities, command show port fdir (port_id) can be used to acquire the information. 2956 2957# Commands to add flow director filters of different flow types:: 2958 2959 flow_director_filter (port_id) mode IP (add|del|update) \ 2960 flow (ipv4-other|ipv4-frag|ipv6-other|ipv6-frag) \ 2961 src (src_ip_address) dst (dst_ip_address) \ 2962 tos (tos_value) proto (proto_value) ttl (ttl_value) \ 2963 vlan (vlan_value) flexbytes (flexbytes_value) \ 2964 (drop|fwd) pf|vf(vf_id) queue (queue_id) \ 2965 fd_id (fd_id_value) 2966 2967 flow_director_filter (port_id) mode IP (add|del|update) \ 2968 flow (ipv4-tcp|ipv4-udp|ipv6-tcp|ipv6-udp) \ 2969 src (src_ip_address) (src_port) \ 2970 dst (dst_ip_address) (dst_port) \ 2971 tos (tos_value) ttl (ttl_value) \ 2972 vlan (vlan_value) flexbytes (flexbytes_value) \ 2973 (drop|fwd) queue pf|vf(vf_id) (queue_id) \ 2974 fd_id (fd_id_value) 2975 2976 flow_director_filter (port_id) mode IP (add|del|update) \ 2977 flow (ipv4-sctp|ipv6-sctp) \ 2978 src (src_ip_address) (src_port) \ 2979 dst (dst_ip_address) (dst_port) \ 2980 tos (tos_value) ttl (ttl_value) \ 2981 tag (verification_tag) vlan (vlan_value) \ 2982 flexbytes (flexbytes_value) (drop|fwd) \ 2983 pf|vf(vf_id) queue (queue_id) fd_id (fd_id_value) 2984 2985 flow_director_filter (port_id) mode IP (add|del|update) flow l2_payload \ 2986 ether (ethertype) flexbytes (flexbytes_value) \ 2987 (drop|fwd) pf|vf(vf_id) queue (queue_id) 2988 fd_id (fd_id_value) 2989 2990 flow_director_filter (port_id) mode MAC-VLAN (add|del|update) \ 2991 mac (mac_address) vlan (vlan_value) \ 2992 flexbytes (flexbytes_value) (drop|fwd) \ 2993 queue (queue_id) fd_id (fd_id_value) 2994 2995 flow_director_filter (port_id) mode Tunnel (add|del|update) \ 2996 mac (mac_address) vlan (vlan_value) \ 2997 tunnel (NVGRE|VxLAN) tunnel-id (tunnel_id_value) \ 2998 flexbytes (flexbytes_value) (drop|fwd) \ 2999 queue (queue_id) fd_id (fd_id_value) 3000 3001 flow_director_filter (port_id) mode raw (add|del|update) flow (flow_id) \ 3002 (drop|fwd) queue (queue_id) fd_id (fd_id_value) \ 3003 packet (packet file name) 3004 3005For example, to add an ipv4-udp flow type filter:: 3006 3007 testpmd> flow_director_filter 0 mode IP add flow ipv4-udp src 2.2.2.3 32 \ 3008 dst 2.2.2.5 33 tos 2 ttl 40 vlan 0x1 flexbytes (0x88,0x48) \ 3009 fwd pf queue 1 fd_id 1 3010 3011For example, add an ipv4-other flow type filter:: 3012 3013 testpmd> flow_director_filter 0 mode IP add flow ipv4-other src 2.2.2.3 \ 3014 dst 2.2.2.5 tos 2 proto 20 ttl 40 vlan 0x1 \ 3015 flexbytes (0x88,0x48) fwd pf queue 1 fd_id 1 3016 3017flush_flow_director 3018~~~~~~~~~~~~~~~~~~~ 3019 3020Flush all flow director filters on a device:: 3021 3022 testpmd> flush_flow_director (port_id) 3023 3024Example, to flush all flow director filter on port 0:: 3025 3026 testpmd> flush_flow_director 0 3027 3028flow_director_mask 3029~~~~~~~~~~~~~~~~~~ 3030 3031Set flow director's input masks:: 3032 3033 flow_director_mask (port_id) mode IP vlan (vlan_value) \ 3034 src_mask (ipv4_src) (ipv6_src) (src_port) \ 3035 dst_mask (ipv4_dst) (ipv6_dst) (dst_port) 3036 3037 flow_director_mask (port_id) mode MAC-VLAN vlan (vlan_value) 3038 3039 flow_director_mask (port_id) mode Tunnel vlan (vlan_value) \ 3040 mac (mac_value) tunnel-type (tunnel_type_value) \ 3041 tunnel-id (tunnel_id_value) 3042 3043Example, to set flow director mask on port 0:: 3044 3045 testpmd> flow_director_mask 0 mode IP vlan 0xefff \ 3046 src_mask 255.255.255.255 \ 3047 FFFF:FFFF:FFFF:FFFF:FFFF:FFFF:FFFF:FFFF 0xFFFF \ 3048 dst_mask 255.255.255.255 \ 3049 FFFF:FFFF:FFFF:FFFF:FFFF:FFFF:FFFF:FFFF 0xFFFF 3050 3051flow_director_flex_mask 3052~~~~~~~~~~~~~~~~~~~~~~~ 3053 3054set masks of flow director's flexible payload based on certain flow type:: 3055 3056 testpmd> flow_director_flex_mask (port_id) \ 3057 flow (none|ipv4-other|ipv4-frag|ipv4-tcp|ipv4-udp|ipv4-sctp| \ 3058 ipv6-other|ipv6-frag|ipv6-tcp|ipv6-udp|ipv6-sctp| \ 3059 l2_payload|all) (mask) 3060 3061Example, to set flow director's flex mask for all flow type on port 0:: 3062 3063 testpmd> flow_director_flex_mask 0 flow all \ 3064 (0xff,0xff,0,0,0,0,0,0,0,0,0,0,0,0,0,0) 3065 3066 3067flow_director_flex_payload 3068~~~~~~~~~~~~~~~~~~~~~~~~~~ 3069 3070Configure flexible payload selection:: 3071 3072 flow_director_flex_payload (port_id) (raw|l2|l3|l4) (config) 3073 3074For example, to select the first 16 bytes from the offset 4 (bytes) of packet's payload as flexible payload:: 3075 3076 testpmd> flow_director_flex_payload 0 l4 \ 3077 (4,5,6,7,8,9,10,11,12,13,14,15,16,17,18,19) 3078 3079get_sym_hash_ena_per_port 3080~~~~~~~~~~~~~~~~~~~~~~~~~ 3081 3082Get symmetric hash enable configuration per port:: 3083 3084 get_sym_hash_ena_per_port (port_id) 3085 3086For example, to get symmetric hash enable configuration of port 1:: 3087 3088 testpmd> get_sym_hash_ena_per_port 1 3089 3090set_sym_hash_ena_per_port 3091~~~~~~~~~~~~~~~~~~~~~~~~~ 3092 3093Set symmetric hash enable configuration per port to enable or disable:: 3094 3095 set_sym_hash_ena_per_port (port_id) (enable|disable) 3096 3097For example, to set symmetric hash enable configuration of port 1 to enable:: 3098 3099 testpmd> set_sym_hash_ena_per_port 1 enable 3100 3101get_hash_global_config 3102~~~~~~~~~~~~~~~~~~~~~~ 3103 3104Get the global configurations of hash filters:: 3105 3106 get_hash_global_config (port_id) 3107 3108For example, to get the global configurations of hash filters of port 1:: 3109 3110 testpmd> get_hash_global_config 1 3111 3112set_hash_global_config 3113~~~~~~~~~~~~~~~~~~~~~~ 3114 3115Set the global configurations of hash filters:: 3116 3117 set_hash_global_config (port_id) (toeplitz|simple_xor|default) \ 3118 (ipv4|ipv4-frag|ipv4-tcp|ipv4-udp|ipv4-sctp|ipv4-other|ipv6|ipv6-frag| \ 3119 ipv6-tcp|ipv6-udp|ipv6-sctp|ipv6-other|l2_payload|<flow_id>) \ 3120 (enable|disable) 3121 3122For example, to enable simple_xor for flow type of ipv6 on port 2:: 3123 3124 testpmd> set_hash_global_config 2 simple_xor ipv6 enable 3125 3126set_hash_input_set 3127~~~~~~~~~~~~~~~~~~ 3128 3129Set the input set for hash:: 3130 3131 set_hash_input_set (port_id) (ipv4-frag|ipv4-tcp|ipv4-udp|ipv4-sctp| \ 3132 ipv4-other|ipv6-frag|ipv6-tcp|ipv6-udp|ipv6-sctp|ipv6-other| \ 3133 l2_payload|<flow_id>) (ovlan|ivlan|src-ipv4|dst-ipv4|src-ipv6|dst-ipv6| \ 3134 ipv4-tos|ipv4-proto|ipv6-tc|ipv6-next-header|udp-src-port|udp-dst-port| \ 3135 tcp-src-port|tcp-dst-port|sctp-src-port|sctp-dst-port|sctp-veri-tag| \ 3136 udp-key|gre-key|fld-1st|fld-2nd|fld-3rd|fld-4th|fld-5th|fld-6th|fld-7th| \ 3137 fld-8th|none) (select|add) 3138 3139For example, to add source IP to hash input set for flow type of ipv4-udp on port 0:: 3140 3141 testpmd> set_hash_input_set 0 ipv4-udp src-ipv4 add 3142 3143set_fdir_input_set 3144~~~~~~~~~~~~~~~~~~ 3145 3146The Flow Director filters can match the different fields for different type of packet, i.e. specific input set 3147on per flow type and the flexible payload. This command can be used to change input set for each flow type. 3148 3149Set the input set for flow director:: 3150 3151 set_fdir_input_set (port_id) (ipv4-frag|ipv4-tcp|ipv4-udp|ipv4-sctp| \ 3152 ipv4-other|ipv6|ipv6-frag|ipv6-tcp|ipv6-udp|ipv6-sctp|ipv6-other| \ 3153 l2_payload|<flow_id>) (ivlan|ethertype|src-ipv4|dst-ipv4|src-ipv6|dst-ipv6| \ 3154 ipv4-tos|ipv4-proto|ipv4-ttl|ipv6-tc|ipv6-next-header|ipv6-hop-limits| \ 3155 tudp-src-port|udp-dst-port|cp-src-port|tcp-dst-port|sctp-src-port| \ 3156 sctp-dst-port|sctp-veri-tag|none) (select|add) 3157 3158For example to add source IP to FD input set for flow type of ipv4-udp on port 0:: 3159 3160 testpmd> set_fdir_input_set 0 ipv4-udp src-ipv4 add 3161 3162global_config 3163~~~~~~~~~~~~~ 3164 3165Set different GRE key length for input set:: 3166 3167 global_config (port_id) gre-key-len (number in bytes) 3168 3169For example to set GRE key length for input set to 4 bytes on port 0:: 3170 3171 testpmd> global_config 0 gre-key-len 4 3172 3173 3174.. _testpmd_rte_flow: 3175 3176Flow rules management 3177--------------------- 3178 3179Control of the generic flow API (*rte_flow*) is fully exposed through the 3180``flow`` command (validation, creation, destruction, queries and operation 3181modes). 3182 3183Considering *rte_flow* overlaps with all `Filter Functions`_, using both 3184features simultaneously may cause undefined side-effects and is therefore 3185not recommended. 3186 3187``flow`` syntax 3188~~~~~~~~~~~~~~~ 3189 3190Because the ``flow`` command uses dynamic tokens to handle the large number 3191of possible flow rules combinations, its behavior differs slightly from 3192other commands, in particular: 3193 3194- Pressing *?* or the *<tab>* key displays contextual help for the current 3195 token, not that of the entire command. 3196 3197- Optional and repeated parameters are supported (provided they are listed 3198 in the contextual help). 3199 3200The first parameter stands for the operation mode. Possible operations and 3201their general syntax are described below. They are covered in detail in the 3202following sections. 3203 3204- Check whether a flow rule can be created:: 3205 3206 flow validate {port_id} 3207 [group {group_id}] [priority {level}] [ingress] [egress] [transfer] 3208 pattern {item} [/ {item} [...]] / end 3209 actions {action} [/ {action} [...]] / end 3210 3211- Create a flow rule:: 3212 3213 flow create {port_id} 3214 [group {group_id}] [priority {level}] [ingress] [egress] [transfer] 3215 pattern {item} [/ {item} [...]] / end 3216 actions {action} [/ {action} [...]] / end 3217 3218- Destroy specific flow rules:: 3219 3220 flow destroy {port_id} rule {rule_id} [...] 3221 3222- Destroy all flow rules:: 3223 3224 flow flush {port_id} 3225 3226- Query an existing flow rule:: 3227 3228 flow query {port_id} {rule_id} {action} 3229 3230- List existing flow rules sorted by priority, filtered by group 3231 identifiers:: 3232 3233 flow list {port_id} [group {group_id}] [...] 3234 3235- Restrict ingress traffic to the defined flow rules:: 3236 3237 flow isolate {port_id} {boolean} 3238 3239Validating flow rules 3240~~~~~~~~~~~~~~~~~~~~~ 3241 3242``flow validate`` reports whether a flow rule would be accepted by the 3243underlying device in its current state but stops short of creating it. It is 3244bound to ``rte_flow_validate()``:: 3245 3246 flow validate {port_id} 3247 [group {group_id}] [priority {level}] [ingress] [egress] [transfer] 3248 pattern {item} [/ {item} [...]] / end 3249 actions {action} [/ {action} [...]] / end 3250 3251If successful, it will show:: 3252 3253 Flow rule validated 3254 3255Otherwise it will show an error message of the form:: 3256 3257 Caught error type [...] ([...]): [...] 3258 3259This command uses the same parameters as ``flow create``, their format is 3260described in `Creating flow rules`_. 3261 3262Check whether redirecting any Ethernet packet received on port 0 to RX queue 3263index 6 is supported:: 3264 3265 testpmd> flow validate 0 ingress pattern eth / end 3266 actions queue index 6 / end 3267 Flow rule validated 3268 testpmd> 3269 3270Port 0 does not support TCPv6 rules:: 3271 3272 testpmd> flow validate 0 ingress pattern eth / ipv6 / tcp / end 3273 actions drop / end 3274 Caught error type 9 (specific pattern item): Invalid argument 3275 testpmd> 3276 3277Creating flow rules 3278~~~~~~~~~~~~~~~~~~~ 3279 3280``flow create`` validates and creates the specified flow rule. It is bound 3281to ``rte_flow_create()``:: 3282 3283 flow create {port_id} 3284 [group {group_id}] [priority {level}] [ingress] [egress] [transfer] 3285 pattern {item} [/ {item} [...]] / end 3286 actions {action} [/ {action} [...]] / end 3287 3288If successful, it will return a flow rule ID usable with other commands:: 3289 3290 Flow rule #[...] created 3291 3292Otherwise it will show an error message of the form:: 3293 3294 Caught error type [...] ([...]): [...] 3295 3296Parameters describe in the following order: 3297 3298- Attributes (*group*, *priority*, *ingress*, *egress*, *transfer* tokens). 3299- A matching pattern, starting with the *pattern* token and terminated by an 3300 *end* pattern item. 3301- Actions, starting with the *actions* token and terminated by an *end* 3302 action. 3303 3304These translate directly to *rte_flow* objects provided as-is to the 3305underlying functions. 3306 3307The shortest valid definition only comprises mandatory tokens:: 3308 3309 testpmd> flow create 0 pattern end actions end 3310 3311Note that PMDs may refuse rules that essentially do nothing such as this 3312one. 3313 3314**All unspecified object values are automatically initialized to 0.** 3315 3316Attributes 3317^^^^^^^^^^ 3318 3319These tokens affect flow rule attributes (``struct rte_flow_attr``) and are 3320specified before the ``pattern`` token. 3321 3322- ``group {group id}``: priority group. 3323- ``priority {level}``: priority level within group. 3324- ``ingress``: rule applies to ingress traffic. 3325- ``egress``: rule applies to egress traffic. 3326- ``transfer``: apply rule directly to endpoints found in pattern. 3327 3328Each instance of an attribute specified several times overrides the previous 3329value as shown below (group 4 is used):: 3330 3331 testpmd> flow create 0 group 42 group 24 group 4 [...] 3332 3333Note that once enabled, ``ingress`` and ``egress`` cannot be disabled. 3334 3335While not specifying a direction is an error, some rules may allow both 3336simultaneously. 3337 3338Most rules affect RX therefore contain the ``ingress`` token:: 3339 3340 testpmd> flow create 0 ingress pattern [...] 3341 3342Matching pattern 3343^^^^^^^^^^^^^^^^ 3344 3345A matching pattern starts after the ``pattern`` token. It is made of pattern 3346items and is terminated by a mandatory ``end`` item. 3347 3348Items are named after their type (*RTE_FLOW_ITEM_TYPE_* from ``enum 3349rte_flow_item_type``). 3350 3351The ``/`` token is used as a separator between pattern items as shown 3352below:: 3353 3354 testpmd> flow create 0 ingress pattern eth / ipv4 / udp / end [...] 3355 3356Note that protocol items like these must be stacked from lowest to highest 3357layer to make sense. For instance, the following rule is either invalid or 3358unlikely to match any packet:: 3359 3360 testpmd> flow create 0 ingress pattern eth / udp / ipv4 / end [...] 3361 3362More information on these restrictions can be found in the *rte_flow* 3363documentation. 3364 3365Several items support additional specification structures, for example 3366``ipv4`` allows specifying source and destination addresses as follows:: 3367 3368 testpmd> flow create 0 ingress pattern eth / ipv4 src is 10.1.1.1 3369 dst is 10.2.0.0 / end [...] 3370 3371This rule matches all IPv4 traffic with the specified properties. 3372 3373In this example, ``src`` and ``dst`` are field names of the underlying 3374``struct rte_flow_item_ipv4`` object. All item properties can be specified 3375in a similar fashion. 3376 3377The ``is`` token means that the subsequent value must be matched exactly, 3378and assigns ``spec`` and ``mask`` fields in ``struct rte_flow_item`` 3379accordingly. Possible assignment tokens are: 3380 3381- ``is``: match value perfectly (with full bit-mask). 3382- ``spec``: match value according to configured bit-mask. 3383- ``last``: specify upper bound to establish a range. 3384- ``mask``: specify bit-mask with relevant bits set to one. 3385- ``prefix``: generate bit-mask from a prefix length. 3386 3387These yield identical results:: 3388 3389 ipv4 src is 10.1.1.1 3390 3391:: 3392 3393 ipv4 src spec 10.1.1.1 src mask 255.255.255.255 3394 3395:: 3396 3397 ipv4 src spec 10.1.1.1 src prefix 32 3398 3399:: 3400 3401 ipv4 src is 10.1.1.1 src last 10.1.1.1 # range with a single value 3402 3403:: 3404 3405 ipv4 src is 10.1.1.1 src last 0 # 0 disables range 3406 3407Inclusive ranges can be defined with ``last``:: 3408 3409 ipv4 src is 10.1.1.1 src last 10.2.3.4 # 10.1.1.1 to 10.2.3.4 3410 3411Note that ``mask`` affects both ``spec`` and ``last``:: 3412 3413 ipv4 src is 10.1.1.1 src last 10.2.3.4 src mask 255.255.0.0 3414 # matches 10.1.0.0 to 10.2.255.255 3415 3416Properties can be modified multiple times:: 3417 3418 ipv4 src is 10.1.1.1 src is 10.1.2.3 src is 10.2.3.4 # matches 10.2.3.4 3419 3420:: 3421 3422 ipv4 src is 10.1.1.1 src prefix 24 src prefix 16 # matches 10.1.0.0/16 3423 3424Pattern items 3425^^^^^^^^^^^^^ 3426 3427This section lists supported pattern items and their attributes, if any. 3428 3429- ``end``: end list of pattern items. 3430 3431- ``void``: no-op pattern item. 3432 3433- ``invert``: perform actions when pattern does not match. 3434 3435- ``any``: match any protocol for the current layer. 3436 3437 - ``num {unsigned}``: number of layers covered. 3438 3439- ``pf``: match traffic from/to the physical function. 3440 3441- ``vf``: match traffic from/to a virtual function ID. 3442 3443 - ``id {unsigned}``: VF ID. 3444 3445- ``phy_port``: match traffic from/to a specific physical port. 3446 3447 - ``index {unsigned}``: physical port index. 3448 3449- ``port_id``: match traffic from/to a given DPDK port ID. 3450 3451 - ``id {unsigned}``: DPDK port ID. 3452 3453- ``mark``: match value set in previously matched flow rule using the mark action. 3454 3455 - ``id {unsigned}``: arbitrary integer value. 3456 3457- ``raw``: match an arbitrary byte string. 3458 3459 - ``relative {boolean}``: look for pattern after the previous item. 3460 - ``search {boolean}``: search pattern from offset (see also limit). 3461 - ``offset {integer}``: absolute or relative offset for pattern. 3462 - ``limit {unsigned}``: search area limit for start of pattern. 3463 - ``pattern {string}``: byte string to look for. 3464 3465- ``eth``: match Ethernet header. 3466 3467 - ``dst {MAC-48}``: destination MAC. 3468 - ``src {MAC-48}``: source MAC. 3469 - ``type {unsigned}``: EtherType or TPID. 3470 3471- ``vlan``: match 802.1Q/ad VLAN tag. 3472 3473 - ``tci {unsigned}``: tag control information. 3474 - ``pcp {unsigned}``: priority code point. 3475 - ``dei {unsigned}``: drop eligible indicator. 3476 - ``vid {unsigned}``: VLAN identifier. 3477 - ``inner_type {unsigned}``: inner EtherType or TPID. 3478 3479- ``ipv4``: match IPv4 header. 3480 3481 - ``tos {unsigned}``: type of service. 3482 - ``ttl {unsigned}``: time to live. 3483 - ``proto {unsigned}``: next protocol ID. 3484 - ``src {ipv4 address}``: source address. 3485 - ``dst {ipv4 address}``: destination address. 3486 3487- ``ipv6``: match IPv6 header. 3488 3489 - ``tc {unsigned}``: traffic class. 3490 - ``flow {unsigned}``: flow label. 3491 - ``proto {unsigned}``: protocol (next header). 3492 - ``hop {unsigned}``: hop limit. 3493 - ``src {ipv6 address}``: source address. 3494 - ``dst {ipv6 address}``: destination address. 3495 3496- ``icmp``: match ICMP header. 3497 3498 - ``type {unsigned}``: ICMP packet type. 3499 - ``code {unsigned}``: ICMP packet code. 3500 3501- ``udp``: match UDP header. 3502 3503 - ``src {unsigned}``: UDP source port. 3504 - ``dst {unsigned}``: UDP destination port. 3505 3506- ``tcp``: match TCP header. 3507 3508 - ``src {unsigned}``: TCP source port. 3509 - ``dst {unsigned}``: TCP destination port. 3510 3511- ``sctp``: match SCTP header. 3512 3513 - ``src {unsigned}``: SCTP source port. 3514 - ``dst {unsigned}``: SCTP destination port. 3515 - ``tag {unsigned}``: validation tag. 3516 - ``cksum {unsigned}``: checksum. 3517 3518- ``vxlan``: match VXLAN header. 3519 3520 - ``vni {unsigned}``: VXLAN identifier. 3521 3522- ``e_tag``: match IEEE 802.1BR E-Tag header. 3523 3524 - ``grp_ecid_b {unsigned}``: GRP and E-CID base. 3525 3526- ``nvgre``: match NVGRE header. 3527 3528 - ``tni {unsigned}``: virtual subnet ID. 3529 3530- ``mpls``: match MPLS header. 3531 3532 - ``label {unsigned}``: MPLS label. 3533 3534- ``gre``: match GRE header. 3535 3536 - ``protocol {unsigned}``: protocol type. 3537 3538- ``fuzzy``: fuzzy pattern match, expect faster than default. 3539 3540 - ``thresh {unsigned}``: accuracy threshold. 3541 3542- ``gtp``, ``gtpc``, ``gtpu``: match GTPv1 header. 3543 3544 - ``teid {unsigned}``: tunnel endpoint identifier. 3545 3546- ``geneve``: match GENEVE header. 3547 3548 - ``vni {unsigned}``: virtual network identifier. 3549 - ``protocol {unsigned}``: protocol type. 3550 3551- ``vxlan-gpe``: match VXLAN-GPE header. 3552 3553 - ``vni {unsigned}``: VXLAN-GPE identifier. 3554 3555- ``arp_eth_ipv4``: match ARP header for Ethernet/IPv4. 3556 3557 - ``sha {MAC-48}``: sender hardware address. 3558 - ``spa {ipv4 address}``: sender IPv4 address. 3559 - ``tha {MAC-48}``: target hardware address. 3560 - ``tpa {ipv4 address}``: target IPv4 address. 3561 3562- ``ipv6_ext``: match presence of any IPv6 extension header. 3563 3564 - ``next_hdr {unsigned}``: next header. 3565 3566- ``icmp6``: match any ICMPv6 header. 3567 3568 - ``type {unsigned}``: ICMPv6 type. 3569 - ``code {unsigned}``: ICMPv6 code. 3570 3571- ``icmp6_nd_ns``: match ICMPv6 neighbor discovery solicitation. 3572 3573 - ``target_addr {ipv6 address}``: target address. 3574 3575- ``icmp6_nd_na``: match ICMPv6 neighbor discovery advertisement. 3576 3577 - ``target_addr {ipv6 address}``: target address. 3578 3579- ``icmp6_nd_opt``: match presence of any ICMPv6 neighbor discovery option. 3580 3581 - ``type {unsigned}``: ND option type. 3582 3583- ``icmp6_nd_opt_sla_eth``: match ICMPv6 neighbor discovery source Ethernet 3584 link-layer address option. 3585 3586 - ``sla {MAC-48}``: source Ethernet LLA. 3587 3588- ``icmp6_nd_opt_sla_eth``: match ICMPv6 neighbor discovery target Ethernet 3589 link-layer address option. 3590 3591 - ``tla {MAC-48}``: target Ethernet LLA. 3592 3593Actions list 3594^^^^^^^^^^^^ 3595 3596A list of actions starts after the ``actions`` token in the same fashion as 3597`Matching pattern`_; actions are separated by ``/`` tokens and the list is 3598terminated by a mandatory ``end`` action. 3599 3600Actions are named after their type (*RTE_FLOW_ACTION_TYPE_* from ``enum 3601rte_flow_action_type``). 3602 3603Dropping all incoming UDPv4 packets can be expressed as follows:: 3604 3605 testpmd> flow create 0 ingress pattern eth / ipv4 / udp / end 3606 actions drop / end 3607 3608Several actions have configurable properties which must be specified when 3609there is no valid default value. For example, ``queue`` requires a target 3610queue index. 3611 3612This rule redirects incoming UDPv4 traffic to queue index 6:: 3613 3614 testpmd> flow create 0 ingress pattern eth / ipv4 / udp / end 3615 actions queue index 6 / end 3616 3617While this one could be rejected by PMDs (unspecified queue index):: 3618 3619 testpmd> flow create 0 ingress pattern eth / ipv4 / udp / end 3620 actions queue / end 3621 3622As defined by *rte_flow*, the list is not ordered, all actions of a given 3623rule are performed simultaneously. These are equivalent:: 3624 3625 queue index 6 / void / mark id 42 / end 3626 3627:: 3628 3629 void / mark id 42 / queue index 6 / end 3630 3631All actions in a list should have different types, otherwise only the last 3632action of a given type is taken into account:: 3633 3634 queue index 4 / queue index 5 / queue index 6 / end # will use queue 6 3635 3636:: 3637 3638 drop / drop / drop / end # drop is performed only once 3639 3640:: 3641 3642 mark id 42 / queue index 3 / mark id 24 / end # mark will be 24 3643 3644Considering they are performed simultaneously, opposite and overlapping 3645actions can sometimes be combined when the end result is unambiguous:: 3646 3647 drop / queue index 6 / end # drop has no effect 3648 3649:: 3650 3651 queue index 6 / rss queues 6 7 8 / end # queue has no effect 3652 3653:: 3654 3655 drop / passthru / end # drop has no effect 3656 3657Note that PMDs may still refuse such combinations. 3658 3659Actions 3660^^^^^^^ 3661 3662This section lists supported actions and their attributes, if any. 3663 3664- ``end``: end list of actions. 3665 3666- ``void``: no-op action. 3667 3668- ``passthru``: let subsequent rule process matched packets. 3669 3670- ``jump``: redirect traffic to group on device. 3671 3672 - ``group {unsigned}``: group to redirect to. 3673 3674- ``mark``: attach 32 bit value to packets. 3675 3676 - ``id {unsigned}``: 32 bit value to return with packets. 3677 3678- ``flag``: flag packets. 3679 3680- ``queue``: assign packets to a given queue index. 3681 3682 - ``index {unsigned}``: queue index to use. 3683 3684- ``drop``: drop packets (note: passthru has priority). 3685 3686- ``count``: enable counters for this rule. 3687 3688- ``rss``: spread packets among several queues. 3689 3690 - ``func {hash function}``: RSS hash function to apply, allowed tokens are 3691 the same as `set_hash_global_config`_. 3692 3693 - ``level {unsigned}``: encapsulation level for ``types``. 3694 3695 - ``types [{RSS hash type} [...]] end``: specific RSS hash types, allowed 3696 tokens are the same as `set_hash_input_set`_, except that an empty list 3697 does not disable RSS but instead requests unspecified "best-effort" 3698 settings. 3699 3700 - ``key {string}``: RSS hash key, overrides ``key_len``. 3701 3702 - ``key_len {unsigned}``: RSS hash key length in bytes, can be used in 3703 conjunction with ``key`` to pad or truncate it. 3704 3705 - ``queues [{unsigned} [...]] end``: queue indices to use. 3706 3707- ``pf``: direct traffic to physical function. 3708 3709- ``vf``: direct traffic to a virtual function ID. 3710 3711 - ``original {boolean}``: use original VF ID if possible. 3712 - ``id {unsigned}``: VF ID. 3713 3714- ``phy_port``: direct packets to physical port index. 3715 3716 - ``original {boolean}``: use original port index if possible. 3717 - ``index {unsigned}``: physical port index. 3718 3719- ``port_id``: direct matching traffic to a given DPDK port ID. 3720 3721 - ``original {boolean}``: use original DPDK port ID if possible. 3722 - ``id {unsigned}``: DPDK port ID. 3723 3724- ``of_set_mpls_ttl``: OpenFlow's ``OFPAT_SET_MPLS_TTL``. 3725 3726 - ``mpls_ttl``: MPLS TTL. 3727 3728- ``of_dec_mpls_ttl``: OpenFlow's ``OFPAT_DEC_MPLS_TTL``. 3729 3730- ``of_set_nw_ttl``: OpenFlow's ``OFPAT_SET_NW_TTL``. 3731 3732 - ``nw_ttl``: IP TTL. 3733 3734- ``of_dec_nw_ttl``: OpenFlow's ``OFPAT_DEC_NW_TTL``. 3735 3736- ``of_copy_ttl_out``: OpenFlow's ``OFPAT_COPY_TTL_OUT``. 3737 3738- ``of_copy_ttl_in``: OpenFlow's ``OFPAT_COPY_TTL_IN``. 3739 3740- ``of_pop_vlan``: OpenFlow's ``OFPAT_POP_VLAN``. 3741 3742- ``of_push_vlan``: OpenFlow's ``OFPAT_PUSH_VLAN``. 3743 3744 - ``ethertype``: Ethertype. 3745 3746- ``of_set_vlan_vid``: OpenFlow's ``OFPAT_SET_VLAN_VID``. 3747 3748 - ``vlan_vid``: VLAN id. 3749 3750- ``of_set_vlan_pcp``: OpenFlow's ``OFPAT_SET_VLAN_PCP``. 3751 3752 - ``vlan_pcp``: VLAN priority. 3753 3754- ``of_pop_mpls``: OpenFlow's ``OFPAT_POP_MPLS``. 3755 3756 - ``ethertype``: Ethertype. 3757 3758- ``of_push_mpls``: OpenFlow's ``OFPAT_PUSH_MPLS``. 3759 3760 - ``ethertype``: Ethertype. 3761 3762- ``vxlan_encap``: Performs a VXLAN encapsulation, outer layer configuration 3763 is done through `Config VXLAN Encap outer layers`_. 3764 3765- ``vxlan_decap``: Performs a decapsulation action by stripping all headers of 3766 the VXLAN tunnel network overlay from the matched flow. 3767 3768- ``nvgre_encap``: Performs a NVGRE encapsulation, outer layer configuration 3769 is done through `Config NVGRE Encap outer layers`_. 3770 3771- ``nvgre_decap``: Performs a decapsulation action by stripping all headers of 3772 the NVGRE tunnel network overlay from the matched flow. 3773 3774- ``set_ipv4_src``: Set a new IPv4 source address in the outermost IPv4 header. 3775 3776 - ``ipv4_addr``: New IPv4 source address. 3777 3778- ``set_ipv4_dst``: Set a new IPv4 destination address in the outermost IPv4 3779 header. 3780 3781 - ``ipv4_addr``: New IPv4 destination address. 3782 3783- ``set_ipv6_src``: Set a new IPv6 source address in the outermost IPv6 header. 3784 3785 - ``ipv6_addr``: New IPv6 source address. 3786 3787- ``set_ipv6_dst``: Set a new IPv6 destination address in the outermost IPv6 3788 header. 3789 3790 - ``ipv6_addr``: New IPv6 destination address. 3791 3792- ``of_set_tp_src``: Set a new source port number in the outermost TCP/UDP 3793 header. 3794 3795 - ``port``: New TCP/UDP source port number. 3796 3797- ``of_set_tp_dst``: Set a new destination port number in the outermost TCP/UDP 3798 header. 3799 3800 - ``port``: New TCP/UDP destination port number. 3801 3802- ``mac_swap``: Swap the source and destination MAC addresses in the outermost 3803 Ethernet header. 3804 3805Destroying flow rules 3806~~~~~~~~~~~~~~~~~~~~~ 3807 3808``flow destroy`` destroys one or more rules from their rule ID (as returned 3809by ``flow create``), this command calls ``rte_flow_destroy()`` as many 3810times as necessary:: 3811 3812 flow destroy {port_id} rule {rule_id} [...] 3813 3814If successful, it will show:: 3815 3816 Flow rule #[...] destroyed 3817 3818It does not report anything for rule IDs that do not exist. The usual error 3819message is shown when a rule cannot be destroyed:: 3820 3821 Caught error type [...] ([...]): [...] 3822 3823``flow flush`` destroys all rules on a device and does not take extra 3824arguments. It is bound to ``rte_flow_flush()``:: 3825 3826 flow flush {port_id} 3827 3828Any errors are reported as above. 3829 3830Creating several rules and destroying them:: 3831 3832 testpmd> flow create 0 ingress pattern eth / ipv6 / end 3833 actions queue index 2 / end 3834 Flow rule #0 created 3835 testpmd> flow create 0 ingress pattern eth / ipv4 / end 3836 actions queue index 3 / end 3837 Flow rule #1 created 3838 testpmd> flow destroy 0 rule 0 rule 1 3839 Flow rule #1 destroyed 3840 Flow rule #0 destroyed 3841 testpmd> 3842 3843The same result can be achieved using ``flow flush``:: 3844 3845 testpmd> flow create 0 ingress pattern eth / ipv6 / end 3846 actions queue index 2 / end 3847 Flow rule #0 created 3848 testpmd> flow create 0 ingress pattern eth / ipv4 / end 3849 actions queue index 3 / end 3850 Flow rule #1 created 3851 testpmd> flow flush 0 3852 testpmd> 3853 3854Non-existent rule IDs are ignored:: 3855 3856 testpmd> flow create 0 ingress pattern eth / ipv6 / end 3857 actions queue index 2 / end 3858 Flow rule #0 created 3859 testpmd> flow create 0 ingress pattern eth / ipv4 / end 3860 actions queue index 3 / end 3861 Flow rule #1 created 3862 testpmd> flow destroy 0 rule 42 rule 10 rule 2 3863 testpmd> 3864 testpmd> flow destroy 0 rule 0 3865 Flow rule #0 destroyed 3866 testpmd> 3867 3868Querying flow rules 3869~~~~~~~~~~~~~~~~~~~ 3870 3871``flow query`` queries a specific action of a flow rule having that 3872ability. Such actions collect information that can be reported using this 3873command. It is bound to ``rte_flow_query()``:: 3874 3875 flow query {port_id} {rule_id} {action} 3876 3877If successful, it will display either the retrieved data for known actions 3878or the following message:: 3879 3880 Cannot display result for action type [...] ([...]) 3881 3882Otherwise, it will complain either that the rule does not exist or that some 3883error occurred:: 3884 3885 Flow rule #[...] not found 3886 3887:: 3888 3889 Caught error type [...] ([...]): [...] 3890 3891Currently only the ``count`` action is supported. This action reports the 3892number of packets that hit the flow rule and the total number of bytes. Its 3893output has the following format:: 3894 3895 count: 3896 hits_set: [...] # whether "hits" contains a valid value 3897 bytes_set: [...] # whether "bytes" contains a valid value 3898 hits: [...] # number of packets 3899 bytes: [...] # number of bytes 3900 3901Querying counters for TCPv6 packets redirected to queue 6:: 3902 3903 testpmd> flow create 0 ingress pattern eth / ipv6 / tcp / end 3904 actions queue index 6 / count / end 3905 Flow rule #4 created 3906 testpmd> flow query 0 4 count 3907 count: 3908 hits_set: 1 3909 bytes_set: 0 3910 hits: 386446 3911 bytes: 0 3912 testpmd> 3913 3914Listing flow rules 3915~~~~~~~~~~~~~~~~~~ 3916 3917``flow list`` lists existing flow rules sorted by priority and optionally 3918filtered by group identifiers:: 3919 3920 flow list {port_id} [group {group_id}] [...] 3921 3922This command only fails with the following message if the device does not 3923exist:: 3924 3925 Invalid port [...] 3926 3927Output consists of a header line followed by a short description of each 3928flow rule, one per line. There is no output at all when no flow rules are 3929configured on the device:: 3930 3931 ID Group Prio Attr Rule 3932 [...] [...] [...] [...] [...] 3933 3934``Attr`` column flags: 3935 3936- ``i`` for ``ingress``. 3937- ``e`` for ``egress``. 3938 3939Creating several flow rules and listing them:: 3940 3941 testpmd> flow create 0 ingress pattern eth / ipv4 / end 3942 actions queue index 6 / end 3943 Flow rule #0 created 3944 testpmd> flow create 0 ingress pattern eth / ipv6 / end 3945 actions queue index 2 / end 3946 Flow rule #1 created 3947 testpmd> flow create 0 priority 5 ingress pattern eth / ipv4 / udp / end 3948 actions rss queues 6 7 8 end / end 3949 Flow rule #2 created 3950 testpmd> flow list 0 3951 ID Group Prio Attr Rule 3952 0 0 0 i- ETH IPV4 => QUEUE 3953 1 0 0 i- ETH IPV6 => QUEUE 3954 2 0 5 i- ETH IPV4 UDP => RSS 3955 testpmd> 3956 3957Rules are sorted by priority (i.e. group ID first, then priority level):: 3958 3959 testpmd> flow list 1 3960 ID Group Prio Attr Rule 3961 0 0 0 i- ETH => COUNT 3962 6 0 500 i- ETH IPV6 TCP => DROP COUNT 3963 5 0 1000 i- ETH IPV6 ICMP => QUEUE 3964 1 24 0 i- ETH IPV4 UDP => QUEUE 3965 4 24 10 i- ETH IPV4 TCP => DROP 3966 3 24 20 i- ETH IPV4 => DROP 3967 2 24 42 i- ETH IPV4 UDP => QUEUE 3968 7 63 0 i- ETH IPV6 UDP VXLAN => MARK QUEUE 3969 testpmd> 3970 3971Output can be limited to specific groups:: 3972 3973 testpmd> flow list 1 group 0 group 63 3974 ID Group Prio Attr Rule 3975 0 0 0 i- ETH => COUNT 3976 6 0 500 i- ETH IPV6 TCP => DROP COUNT 3977 5 0 1000 i- ETH IPV6 ICMP => QUEUE 3978 7 63 0 i- ETH IPV6 UDP VXLAN => MARK QUEUE 3979 testpmd> 3980 3981Toggling isolated mode 3982~~~~~~~~~~~~~~~~~~~~~~ 3983 3984``flow isolate`` can be used to tell the underlying PMD that ingress traffic 3985must only be injected from the defined flow rules; that no default traffic 3986is expected outside those rules and the driver is free to assign more 3987resources to handle them. It is bound to ``rte_flow_isolate()``:: 3988 3989 flow isolate {port_id} {boolean} 3990 3991If successful, enabling or disabling isolated mode shows either:: 3992 3993 Ingress traffic on port [...] 3994 is now restricted to the defined flow rules 3995 3996Or:: 3997 3998 Ingress traffic on port [...] 3999 is not restricted anymore to the defined flow rules 4000 4001Otherwise, in case of error:: 4002 4003 Caught error type [...] ([...]): [...] 4004 4005Mainly due to its side effects, PMDs supporting this mode may not have the 4006ability to toggle it more than once without reinitializing affected ports 4007first (e.g. by exiting testpmd). 4008 4009Enabling isolated mode:: 4010 4011 testpmd> flow isolate 0 true 4012 Ingress traffic on port 0 is now restricted to the defined flow rules 4013 testpmd> 4014 4015Disabling isolated mode:: 4016 4017 testpmd> flow isolate 0 false 4018 Ingress traffic on port 0 is not restricted anymore to the defined flow rules 4019 testpmd> 4020 4021Sample QinQ flow rules 4022~~~~~~~~~~~~~~~~~~~~~~ 4023 4024Before creating QinQ rule(s) the following commands should be issued to enable QinQ:: 4025 4026 testpmd> port stop 0 4027 testpmd> vlan set qinq on 0 4028 4029The above command sets the inner and outer TPID's to 0x8100. 4030 4031To change the TPID's the following commands should be used:: 4032 4033 testpmd> vlan set outer tpid 0xa100 0 4034 testpmd> vlan set inner tpid 0x9100 0 4035 testpmd> port start 0 4036 4037Validate and create a QinQ rule on port 0 to steer traffic to a VF queue in a VM. 4038 4039:: 4040 4041 testpmd> flow validate 0 ingress pattern eth / vlan tci is 123 / 4042 vlan tci is 456 / end actions vf id 1 / queue index 0 / end 4043 Flow rule #0 validated 4044 4045 testpmd> flow create 0 ingress pattern eth / vlan tci is 4 / 4046 vlan tci is 456 / end actions vf id 123 / queue index 0 / end 4047 Flow rule #0 created 4048 4049 testpmd> flow list 0 4050 ID Group Prio Attr Rule 4051 0 0 0 i- ETH VLAN VLAN=>VF QUEUE 4052 4053Validate and create a QinQ rule on port 0 to steer traffic to a queue on the host. 4054 4055:: 4056 4057 testpmd> flow validate 0 ingress pattern eth / vlan tci is 321 / 4058 vlan tci is 654 / end actions pf / queue index 0 / end 4059 Flow rule #1 validated 4060 4061 testpmd> flow create 0 ingress pattern eth / vlan tci is 321 / 4062 vlan tci is 654 / end actions pf / queue index 1 / end 4063 Flow rule #1 created 4064 4065 testpmd> flow list 0 4066 ID Group Prio Attr Rule 4067 0 0 0 i- ETH VLAN VLAN=>VF QUEUE 4068 1 0 0 i- ETH VLAN VLAN=>PF QUEUE 4069 4070Sample VXLAN encapsulation rule 4071~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ 4072 4073VXLAN encapsulation outer layer has default value pre-configured in testpmd 4074source code, those can be changed by using the following commands 4075 4076IPv4 VXLAN outer header:: 4077 4078 testpmd> set vxlan ip-version ipv4 vni 4 udp-src 4 udp-dst 4 ip-src 127.0.0.1 4079 ip-dst 128.0.0.1 eth-src 11:11:11:11:11:11 eth-dst 22:22:22:22:22:22 4080 testpmd> flow create 0 ingress pattern end actions vxlan_encap / 4081 queue index 0 / end 4082 4083 testpmd> set vxlan-with-vlan ip-version ipv4 vni 4 udp-src 4 udp-dst 4 ip-src 4084 127.0.0.1 ip-dst 128.0.0.1 vlan-tci 34 eth-src 11:11:11:11:11:11 4085 eth-dst 22:22:22:22:22:22 4086 testpmd> flow create 0 ingress pattern end actions vxlan_encap / 4087 queue index 0 / end 4088 4089IPv6 VXLAN outer header:: 4090 4091 testpmd> set vxlan ip-version ipv6 vni 4 udp-src 4 udp-dst 4 ip-src ::1 4092 ip-dst ::2222 eth-src 11:11:11:11:11:11 eth-dst 22:22:22:22:22:22 4093 testpmd> flow create 0 ingress pattern end actions vxlan_encap / 4094 queue index 0 / end 4095 4096 testpmd> set vxlan-with-vlan ip-version ipv6 vni 4 udp-src 4 udp-dst 4 4097 ip-src ::1 ip-dst ::2222 vlan-tci 34 eth-src 11:11:11:11:11:11 4098 eth-dst 22:22:22:22:22:22 4099 testpmd> flow create 0 ingress pattern end actions vxlan_encap / 4100 queue index 0 / end 4101 4102Sample NVGRE encapsulation rule 4103~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ 4104 4105NVGRE encapsulation outer layer has default value pre-configured in testpmd 4106source code, those can be changed by using the following commands 4107 4108IPv4 NVGRE outer header:: 4109 4110 testpmd> set nvgre ip-version ipv4 tni 4 ip-src 127.0.0.1 ip-dst 128.0.0.1 4111 eth-src 11:11:11:11:11:11 eth-dst 22:22:22:22:22:22 4112 testpmd> flow create 0 ingress pattern end actions nvgre_encap / 4113 queue index 0 / end 4114 4115 testpmd> set nvgre-with-vlan ip-version ipv4 tni 4 ip-src 127.0.0.1 4116 ip-dst 128.0.0.1 vlan-tci 34 eth-src 11:11:11:11:11:11 4117 eth-dst 22:22:22:22:22:22 4118 testpmd> flow create 0 ingress pattern end actions nvgre_encap / 4119 queue index 0 / end 4120 4121IPv6 NVGRE outer header:: 4122 4123 testpmd> set nvgre ip-version ipv6 tni 4 ip-src ::1 ip-dst ::2222 4124 eth-src 11:11:11:11:11:11 eth-dst 22:22:22:22:22:22 4125 testpmd> flow create 0 ingress pattern end actions nvgre_encap / 4126 queue index 0 / end 4127 4128 testpmd> set nvgre-with-vlan ip-version ipv6 tni 4 ip-src ::1 ip-dst ::2222 4129 vlan-tci 34 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 4133BPF Functions 4134-------------- 4135 4136The following sections show functions to load/unload eBPF based filters. 4137 4138bpf-load 4139~~~~~~~~ 4140 4141Load an eBPF program as a callback for partciular RX/TX queue:: 4142 4143 testpmd> bpf-load rx|tx (portid) (queueid) (load-flags) (bpf-prog-filename) 4144 4145The available load-flags are: 4146 4147* ``J``: use JIT generated native code, otherwise BPF interpreter will be used. 4148 4149* ``M``: assume input parameter is a pointer to rte_mbuf, otherwise assume it is a pointer to first segment's data. 4150 4151* ``-``: none. 4152 4153.. note:: 4154 4155 You'll need clang v3.7 or above to build bpf program you'd like to load 4156 4157For example: 4158 4159.. code-block:: console 4160 4161 cd test/bpf 4162 clang -O2 -target bpf -c t1.c 4163 4164Then to load (and JIT compile) t1.o at RX queue 0, port 1:: 4165 4166.. code-block:: console 4167 4168 testpmd> bpf-load rx 1 0 J ./dpdk.org/test/bpf/t1.o 4169 4170To load (not JITed) t1.o at TX queue 0, port 0:: 4171 4172.. code-block:: console 4173 4174 testpmd> bpf-load tx 0 0 - ./dpdk.org/test/bpf/t1.o 4175 4176bpf-unload 4177~~~~~~~~~~ 4178 4179Unload previously loaded eBPF program for partciular RX/TX queue:: 4180 4181 testpmd> bpf-unload rx|tx (portid) (queueid) 4182 4183For example to unload BPF filter from TX queue 0, port 0: 4184 4185.. code-block:: console 4186 4187 testpmd> bpf-load tx 0 0 - ./dpdk.org/test/bpf/t1.o 4188