1.. BSD LICENSE 2 Copyright(c) 2010-2016 Intel Corporation. All rights reserved. 3 All rights reserved. 4 5 Redistribution and use in source and binary forms, with or without 6 modification, are permitted provided that the following conditions 7 are met: 8 9 * Redistributions of source code must retain the above copyright 10 notice, this list of conditions and the following disclaimer. 11 * Redistributions in binary form must reproduce the above copyright 12 notice, this list of conditions and the following disclaimer in 13 the documentation and/or other materials provided with the 14 distribution. 15 * Neither the name of Intel Corporation nor the names of its 16 contributors may be used to endorse or promote products derived 17 from this software without specific prior written permission. 18 19 THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS 20 "AS IS" AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT 21 LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR 22 A PARTICULAR PURPOSE ARE DISCLAIMED. IN NO EVENT SHALL THE COPYRIGHT 23 OWNER OR CONTRIBUTORS BE LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL, 24 SPECIAL, EXEMPLARY, OR CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT 25 LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES; LOSS OF USE, 26 DATA, OR PROFITS; OR BUSINESS INTERRUPTION) HOWEVER CAUSED AND ON ANY 27 THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT LIABILITY, OR TORT 28 (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY OUT OF THE USE 29 OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE. 30 31.. _testpmd_runtime: 32 33Testpmd Runtime Functions 34========================= 35 36Where the testpmd application is started in interactive mode, (``-i|--interactive``), 37it displays a prompt that can be used to start and stop forwarding, 38configure the application, display statistics (including the extended NIC 39statistics aka xstats) , set the Flow Director and other tasks:: 40 41 testpmd> 42 43The testpmd prompt has some, limited, readline support. 44Common bash command-line functions such as ``Ctrl+a`` and ``Ctrl+e`` to go to the start and end of the prompt line are supported 45as well as access to the command history via the up-arrow. 46 47There is also support for tab completion. 48If you type a partial command and hit ``<TAB>`` you get a list of the available completions: 49 50.. code-block:: console 51 52 testpmd> show port <TAB> 53 54 info [Mul-choice STRING]: show|clear port info|stats|xstats|fdir|stat_qmap|dcb_tc|cap X 55 info [Mul-choice STRING]: show|clear port info|stats|xstats|fdir|stat_qmap|dcb_tc|cap all 56 stats [Mul-choice STRING]: show|clear port info|stats|xstats|fdir|stat_qmap|dcb_tc|cap X 57 stats [Mul-choice STRING]: show|clear port info|stats|xstats|fdir|stat_qmap|dcb_tc|cap all 58 ... 59 60 61.. note:: 62 63 Some examples in this document are too long to fit on one line are are shown wrapped at `"\\"` for display purposes:: 64 65 testpmd> set flow_ctrl rx (on|off) tx (on|off) (high_water) (low_water) \ 66 (pause_time) (send_xon) (port_id) 67 68In the real ``testpmd>`` prompt these commands should be on a single line. 69 70Help Functions 71-------------- 72 73The testpmd has on-line help for the functions that are available at runtime. 74These are divided into sections and can be accessed using help, help section or help all: 75 76.. code-block:: console 77 78 testpmd> help 79 80 help control : Start and stop forwarding. 81 help display : Displaying port, stats and config information. 82 help config : Configuration information. 83 help ports : Configuring ports. 84 help registers : Reading and setting port registers. 85 help filters : Filters configuration help. 86 help all : All of the above sections. 87 88 89Control Functions 90----------------- 91 92start 93~~~~~ 94 95Start packet forwarding with current configuration:: 96 97 testpmd> start 98 99start tx_first 100~~~~~~~~~~~~~~ 101 102Start packet forwarding with current configuration after sending specified number of bursts of packets:: 103 104 testpmd> start tx_first (""|burst_num) 105 106The default burst number is 1 when ``burst_num`` not presented. 107 108stop 109~~~~ 110 111Stop packet forwarding, and display accumulated statistics:: 112 113 testpmd> stop 114 115quit 116~~~~ 117 118Quit to prompt:: 119 120 testpmd> quit 121 122 123Display Functions 124----------------- 125 126The functions in the following sections are used to display information about the 127testpmd configuration or the NIC status. 128 129show port 130~~~~~~~~~ 131 132Display information for a given port or all ports:: 133 134 testpmd> show port (info|stats|xstats|fdir|stat_qmap|dcb_tc|cap) (port_id|all) 135 136The available information categories are: 137 138* ``info``: General port information such as MAC address. 139 140* ``stats``: RX/TX statistics. 141 142* ``xstats``: RX/TX extended NIC statistics. 143 144* ``fdir``: Flow Director information and statistics. 145 146* ``stat_qmap``: Queue statistics mapping. 147 148* ``dcb_tc``: DCB information such as TC mapping. 149 150* ``cap``: Supported offload capabilities. 151 152For example: 153 154.. code-block:: console 155 156 testpmd> show port info 0 157 158 ********************* Infos for port 0 ********************* 159 160 MAC address: XX:XX:XX:XX:XX:XX 161 Connect to socket: 0 162 memory allocation on the socket: 0 163 Link status: up 164 Link speed: 40000 Mbps 165 Link duplex: full-duplex 166 Promiscuous mode: enabled 167 Allmulticast mode: disabled 168 Maximum number of MAC addresses: 64 169 Maximum number of MAC addresses of hash filtering: 0 170 VLAN offload: 171 strip on 172 filter on 173 qinq(extend) off 174 Redirection table size: 512 175 Supported flow types: 176 ipv4-frag 177 ipv4-tcp 178 ipv4-udp 179 ipv4-sctp 180 ipv4-other 181 ipv6-frag 182 ipv6-tcp 183 ipv6-udp 184 ipv6-sctp 185 ipv6-other 186 l2_payload 187 port 188 vxlan 189 geneve 190 nvgre 191 192show port rss reta 193~~~~~~~~~~~~~~~~~~ 194 195Display the rss redirection table entry indicated by masks on port X:: 196 197 testpmd> show port (port_id) rss reta (size) (mask0, mask1...) 198 199size is used to indicate the hardware supported reta size 200 201show port rss-hash 202~~~~~~~~~~~~~~~~~~ 203 204Display the RSS hash functions and RSS hash key of a port:: 205 206 testpmd> show port (port_id) rss-hash ipv4|ipv4-frag|ipv4-tcp|ipv4-udp|ipv4-sctp|ipv4-other|ipv6|ipv6-frag|ipv6-tcp|ipv6-udp|ipv6-sctp|ipv6-other|l2-payload|ipv6-ex|ipv6-tcp-ex|ipv6-udp-ex [key] 207 208clear port 209~~~~~~~~~~ 210 211Clear the port statistics for a given port or for all ports:: 212 213 testpmd> clear port (info|stats|xstats|fdir|stat_qmap) (port_id|all) 214 215For example:: 216 217 testpmd> clear port stats all 218 219show (rxq|txq) 220~~~~~~~~~~~~~~ 221 222Display information for a given port's RX/TX queue:: 223 224 testpmd> show (rxq|txq) info (port_id) (queue_id) 225 226show config 227~~~~~~~~~~~ 228 229Displays the configuration of the application. 230The configuration comes from the command-line, the runtime or the application defaults:: 231 232 testpmd> show config (rxtx|cores|fwd|txpkts) 233 234The available information categories are: 235 236* ``rxtx``: RX/TX configuration items. 237 238* ``cores``: List of forwarding cores. 239 240* ``fwd``: Packet forwarding configuration. 241 242* ``txpkts``: Packets to TX configuration. 243 244For example: 245 246.. code-block:: console 247 248 testpmd> show config rxtx 249 250 io packet forwarding - CRC stripping disabled - packets/burst=16 251 nb forwarding cores=2 - nb forwarding ports=1 252 RX queues=1 - RX desc=128 - RX free threshold=0 253 RX threshold registers: pthresh=8 hthresh=8 wthresh=4 254 TX queues=1 - TX desc=512 - TX free threshold=0 255 TX threshold registers: pthresh=36 hthresh=0 wthresh=0 256 TX RS bit threshold=0 - TXQ flags=0x0 257 258set fwd 259~~~~~~~ 260 261Set the packet forwarding mode:: 262 263 testpmd> set fwd (io|mac|macswap|flowgen| \ 264 rxonly|txonly|csum|icmpecho) (""|retry) 265 266``retry`` can be specified for forwarding engines except ``rx_only``. 267 268The available information categories are: 269 270* ``io``: Forwards packets "as-is" in I/O mode. 271 This is the fastest possible forwarding operation as it does not access packets data. 272 This is the default mode. 273 274* ``mac``: Changes the source and the destination Ethernet addresses of packets before forwarding them. 275 Default application behaviour is to set source Ethernet address to that of the transmitting interface, and destination 276 address to a dummy value (set during init). The user may specify a target destination Ethernet address via the 'eth-peer' or 277 'eth-peer-configfile' command-line options. It is not currently possible to specify a specific source Ethernet address. 278 279* ``macswap``: MAC swap forwarding mode. 280 Swaps the source and the destination Ethernet addresses of packets before forwarding them. 281 282* ``flowgen``: Multi-flow generation mode. 283 Originates a number of flows (with varying destination IP addresses), and terminate receive traffic. 284 285* ``rxonly``: Receives packets but doesn't transmit them. 286 287* ``txonly``: Generates and transmits packets without receiving any. 288 289* ``csum``: Changes the checksum field with hardware or software methods depending on the offload flags on the packet. 290 291* ``icmpecho``: Receives a burst of packets, lookup for IMCP echo requests and, if any, send back ICMP echo replies. 292 293* ``ieee1588``: Demonstrate L2 IEEE1588 V2 PTP timestamping for RX and TX. Requires ``CONFIG_RTE_LIBRTE_IEEE1588=y``. 294 295Note: TX timestamping is only available in the "Full Featured" TX path. To force ``testpmd`` into this mode set ``--txqflags=0``. 296 297Example:: 298 299 testpmd> set fwd rxonly 300 301 Set rxonly packet forwarding mode 302 303 304read rxd 305~~~~~~~~ 306 307Display an RX descriptor for a port RX queue:: 308 309 testpmd> read rxd (port_id) (queue_id) (rxd_id) 310 311For example:: 312 313 testpmd> read rxd 0 0 4 314 0x0000000B - 0x001D0180 / 0x0000000B - 0x001D0180 315 316read txd 317~~~~~~~~ 318 319Display a TX descriptor for a port TX queue:: 320 321 testpmd> read txd (port_id) (queue_id) (txd_id) 322 323For example:: 324 325 testpmd> read txd 0 0 4 326 0x00000001 - 0x24C3C440 / 0x000F0000 - 0x2330003C 327 328 329Configuration Functions 330----------------------- 331 332The testpmd application can be configured from the runtime as well as from the command-line. 333 334This section details the available configuration functions that are available. 335 336.. note:: 337 338 Configuration changes only become active when forwarding is started/restarted. 339 340set default 341~~~~~~~~~~~ 342 343Reset forwarding to the default configuration:: 344 345 testpmd> set default 346 347set verbose 348~~~~~~~~~~~ 349 350Set the debug verbosity level:: 351 352 testpmd> set verbose (level) 353 354Currently the only available levels are 0 (silent except for error) and 1 (fully verbose). 355 356set nbport 357~~~~~~~~~~ 358 359Set the number of ports used by the application: 360 361set nbport (num) 362 363This is equivalent to the ``--nb-ports`` command-line option. 364 365set nbcore 366~~~~~~~~~~ 367 368Set the number of cores used by the application:: 369 370 testpmd> set nbcore (num) 371 372This is equivalent to the ``--nb-cores`` command-line option. 373 374.. note:: 375 376 The number of cores used must not be greater than number of ports used multiplied by the number of queues per port. 377 378set coremask 379~~~~~~~~~~~~ 380 381Set the forwarding cores hexadecimal mask:: 382 383 testpmd> set coremask (mask) 384 385This is equivalent to the ``--coremask`` command-line option. 386 387.. note:: 388 389 The master lcore is reserved for command line parsing only and cannot be masked on for packet forwarding. 390 391set portmask 392~~~~~~~~~~~~ 393 394Set the forwarding ports hexadecimal mask:: 395 396 testpmd> set portmask (mask) 397 398This is equivalent to the ``--portmask`` command-line option. 399 400set burst 401~~~~~~~~~ 402 403Set number of packets per burst:: 404 405 testpmd> set burst (num) 406 407This is equivalent to the ``--burst command-line`` option. 408 409When retry is enabled, the transmit delay time and number of retries can also be set:: 410 411 testpmd> set burst tx delay (microseconds) retry (num) 412 413set txpkts 414~~~~~~~~~~ 415 416Set the length of each segment of the TX-ONLY packets or length of packet for FLOWGEN mode:: 417 418 testpmd> set txpkts (x[,y]*) 419 420Where x[,y]* represents a CSV list of values, without white space. 421 422set txsplit 423~~~~~~~~~~~ 424 425Set the split policy for the TX packets, applicable for TX-ONLY and CSUM forwarding modes:: 426 427 testpmd> set txsplit (off|on|rand) 428 429Where: 430 431* ``off`` disable packet copy & split for CSUM mode. 432 433* ``on`` split outgoing packet into multiple segments. Size of each segment 434 and number of segments per packet is determined by ``set txpkts`` command 435 (see above). 436 437* ``rand`` same as 'on', but number of segments per each packet is a random value between 1 and total number of segments. 438 439set corelist 440~~~~~~~~~~~~ 441 442Set the list of forwarding cores:: 443 444 testpmd> set corelist (x[,y]*) 445 446For example, to change the forwarding cores: 447 448.. code-block:: console 449 450 testpmd> set corelist 3,1 451 testpmd> show config fwd 452 453 io packet forwarding - ports=2 - cores=2 - streams=2 - NUMA support disabled 454 Logical Core 3 (socket 0) forwards packets on 1 streams: 455 RX P=0/Q=0 (socket 0) -> TX P=1/Q=0 (socket 0) peer=02:00:00:00:00:01 456 Logical Core 1 (socket 0) forwards packets on 1 streams: 457 RX P=1/Q=0 (socket 0) -> TX P=0/Q=0 (socket 0) peer=02:00:00:00:00:00 458 459.. note:: 460 461 The cores are used in the same order as specified on the command line. 462 463set portlist 464~~~~~~~~~~~~ 465 466Set the list of forwarding ports:: 467 468 testpmd> set portlist (x[,y]*) 469 470For example, to change the port forwarding: 471 472.. code-block:: console 473 474 testpmd> set portlist 0,2,1,3 475 testpmd> show config fwd 476 477 io packet forwarding - ports=4 - cores=1 - streams=4 478 Logical Core 3 (socket 0) forwards packets on 4 streams: 479 RX P=0/Q=0 (socket 0) -> TX P=2/Q=0 (socket 0) peer=02:00:00:00:00:01 480 RX P=2/Q=0 (socket 0) -> TX P=0/Q=0 (socket 0) peer=02:00:00:00:00:00 481 RX P=1/Q=0 (socket 0) -> TX P=3/Q=0 (socket 0) peer=02:00:00:00:00:03 482 RX P=3/Q=0 (socket 0) -> TX P=1/Q=0 (socket 0) peer=02:00:00:00:00:02 483 484set tx loopback 485~~~~~~~~~~~~~~~ 486 487Enable/disable tx loopback:: 488 489 testpmd> set tx loopback (port_id) (on|off) 490 491set drop enable 492~~~~~~~~~~~~~~~ 493 494set drop enable bit for all queues:: 495 496 testpmd> set all queues drop (port_id) (on|off) 497 498set split drop enable (for VF) 499~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ 500 501set split drop enable bit for VF from PF:: 502 503 testpmd> set vf split drop (port_id) (vf_id) (on|off) 504 505set mac antispoof (for VF) 506~~~~~~~~~~~~~~~~~~~~~~~~~~ 507 508Set mac antispoof for a VF from the PF:: 509 510 testpmd> set vf mac antispoof (port_id) (vf_id) (on|off) 511 512set macsec offload 513~~~~~~~~~~~~~~~~~~ 514 515Enable/disable MACsec offload:: 516 517 testpmd> set macsec offload (port_id) on encrypt (on|off) replay-protect (on|off) 518 testpmd> set macsec offload (port_id) off 519 520set macsec sc 521~~~~~~~~~~~~~ 522 523Configure MACsec secure connection (SC):: 524 525 testpmd> set macsec sc (tx|rx) (port_id) (mac) (pi) 526 527.. note:: 528 529 The pi argument is ignored for tx. 530 Check the NIC Datasheet for hardware limits. 531 532set macsec sa 533~~~~~~~~~~~~~ 534 535Configure MACsec secure association (SA):: 536 537 testpmd> set macsec sa (tx|rx) (port_id) (idx) (an) (pn) (key) 538 539.. note:: 540 541 The IDX value must be 0 or 1. 542 Check the NIC Datasheet for hardware limits. 543 544set broadcast mode (for VF) 545~~~~~~~~~~~~~~~~~~~~~~~~~~~ 546 547Set broadcast mode for a VF from the PF:: 548 549 testpmd> set vf broadcast (port_id) (vf_id) (on|off) 550 551vlan set strip 552~~~~~~~~~~~~~~ 553 554Set the VLAN strip on a port:: 555 556 testpmd> vlan set strip (on|off) (port_id) 557 558vlan set stripq 559~~~~~~~~~~~~~~~ 560 561Set the VLAN strip for a queue on a port:: 562 563 testpmd> vlan set stripq (on|off) (port_id,queue_id) 564 565vlan set stripq (for VF) 566~~~~~~~~~~~~~~~~~~~~~~~~ 567 568Set VLAN strip for all queues in a pool for a VF from the PF:: 569 570 testpmd> set vf vlan stripq (port_id) (vf_id) (on|off) 571 572vlan set insert (for VF) 573~~~~~~~~~~~~~~~~~~~~~~~~ 574 575Set VLAN insert for a VF from the PF:: 576 577 testpmd> set vf vlan insert (port_id) (vf_id) (vlan_id) 578 579vlan set tag (for VF) 580~~~~~~~~~~~~~~~~~~~~~ 581 582Set VLAN tag for a VF from the PF:: 583 584 testpmd> set vf vlan tag (port_id) (vf_id) (on|off) 585 586vlan set antispoof (for VF) 587~~~~~~~~~~~~~~~~~~~~~~~~~~~ 588 589Set VLAN antispoof for a VF from the PF:: 590 591 testpmd> set vf vlan antispoof (port_id) (vf_id) (on|off) 592 593vlan set filter 594~~~~~~~~~~~~~~~ 595 596Set the VLAN filter on a port:: 597 598 testpmd> vlan set filter (on|off) (port_id) 599 600vlan set qinq 601~~~~~~~~~~~~~ 602 603Set the VLAN QinQ (extended queue in queue) on for a port:: 604 605 testpmd> vlan set qinq (on|off) (port_id) 606 607vlan set tpid 608~~~~~~~~~~~~~ 609 610Set the inner or outer VLAN TPID for packet filtering on a port:: 611 612 testpmd> vlan set (inner|outer) tpid (value) (port_id) 613 614.. note:: 615 616 TPID value must be a 16-bit number (value <= 65536). 617 618rx_vlan add 619~~~~~~~~~~~ 620 621Add a VLAN ID, or all identifiers, to the set of VLAN identifiers filtered by port ID:: 622 623 testpmd> rx_vlan add (vlan_id|all) (port_id) 624 625.. note:: 626 627 VLAN filter must be set on that port. VLAN ID < 4096. 628 Depending on the NIC used, number of vlan_ids may be limited to the maximum entries 629 in VFTA table. This is important if enabling all vlan_ids. 630 631rx_vlan rm 632~~~~~~~~~~ 633 634Remove a VLAN ID, or all identifiers, from the set of VLAN identifiers filtered by port ID:: 635 636 testpmd> rx_vlan rm (vlan_id|all) (port_id) 637 638rx_vlan add (for VF) 639~~~~~~~~~~~~~~~~~~~~ 640 641Add a VLAN ID, to the set of VLAN identifiers filtered for VF(s) for port ID:: 642 643 testpmd> rx_vlan add (vlan_id) port (port_id) vf (vf_mask) 644 645rx_vlan rm (for VF) 646~~~~~~~~~~~~~~~~~~~ 647 648Remove a VLAN ID, from the set of VLAN identifiers filtered for VF(s) for port ID:: 649 650 testpmd> rx_vlan rm (vlan_id) port (port_id) vf (vf_mask) 651 652tunnel_filter add 653~~~~~~~~~~~~~~~~~ 654 655Add a tunnel filter on a port:: 656 657 testpmd> tunnel_filter add (port_id) (outer_mac) (inner_mac) (ip_addr) \ 658 (inner_vlan) (vxlan|nvgre|ipingre) (imac-ivlan|imac-ivlan-tenid|\ 659 imac-tenid|imac|omac-imac-tenid|oip|iip) (tenant_id) (queue_id) 660 661The available information categories are: 662 663* ``vxlan``: Set tunnel type as VXLAN. 664 665* ``nvgre``: Set tunnel type as NVGRE. 666 667* ``ipingre``: Set tunnel type as IP-in-GRE. 668 669* ``imac-ivlan``: Set filter type as Inner MAC and VLAN. 670 671* ``imac-ivlan-tenid``: Set filter type as Inner MAC, VLAN and tenant ID. 672 673* ``imac-tenid``: Set filter type as Inner MAC and tenant ID. 674 675* ``imac``: Set filter type as Inner MAC. 676 677* ``omac-imac-tenid``: Set filter type as Outer MAC, Inner MAC and tenant ID. 678 679* ``oip``: Set filter type as Outer IP. 680 681* ``iip``: Set filter type as Inner IP. 682 683Example:: 684 685 testpmd> tunnel_filter add 0 68:05:CA:28:09:82 00:00:00:00:00:00 \ 686 192.168.2.2 0 ipingre oip 1 1 687 688 Set an IP-in-GRE tunnel on port 0, and the filter type is Outer IP. 689 690tunnel_filter remove 691~~~~~~~~~~~~~~~~~~~~ 692 693Remove a tunnel filter on a port:: 694 695 testpmd> tunnel_filter rm (port_id) (outer_mac) (inner_mac) (ip_addr) \ 696 (inner_vlan) (vxlan|nvgre|ipingre) (imac-ivlan|imac-ivlan-tenid|\ 697 imac-tenid|imac|omac-imac-tenid|oip|iip) (tenant_id) (queue_id) 698 699rx_vxlan_port add 700~~~~~~~~~~~~~~~~~ 701 702Add an UDP port for VXLAN packet filter on a port:: 703 704 testpmd> rx_vxlan_port add (udp_port) (port_id) 705 706rx_vxlan_port remove 707~~~~~~~~~~~~~~~~~~~~ 708 709Remove an UDP port for VXLAN packet filter on a port:: 710 711 testpmd> rx_vxlan_port rm (udp_port) (port_id) 712 713tx_vlan set 714~~~~~~~~~~~ 715 716Set hardware insertion of VLAN IDs in packets sent on a port:: 717 718 testpmd> tx_vlan set (port_id) vlan_id[, vlan_id_outer] 719 720For example, set a single VLAN ID (5) insertion on port 0:: 721 722 tx_vlan set 0 5 723 724Or, set double VLAN ID (inner: 2, outer: 3) insertion on port 1:: 725 726 tx_vlan set 1 2 3 727 728 729tx_vlan set pvid 730~~~~~~~~~~~~~~~~ 731 732Set port based hardware insertion of VLAN ID in packets sent on a port:: 733 734 testpmd> tx_vlan set pvid (port_id) (vlan_id) (on|off) 735 736tx_vlan reset 737~~~~~~~~~~~~~ 738 739Disable hardware insertion of a VLAN header in packets sent on a port:: 740 741 testpmd> tx_vlan reset (port_id) 742 743csum set 744~~~~~~~~ 745 746Select hardware or software calculation of the checksum when 747transmitting a packet using the ``csum`` forwarding engine:: 748 749 testpmd> csum set (ip|udp|tcp|sctp|outer-ip) (hw|sw) (port_id) 750 751Where: 752 753* ``ip|udp|tcp|sctp`` always relate to the inner layer. 754 755* ``outer-ip`` relates to the outer IP layer (only for IPv4) in the case where the packet is recognized 756 as a tunnel packet by the forwarding engine (vxlan, gre and ipip are 757 supported). See also the ``csum parse-tunnel`` command. 758 759.. note:: 760 761 Check the NIC Datasheet for hardware limits. 762 763csum parse-tunnel 764~~~~~~~~~~~~~~~~~ 765 766Define how tunneled packets should be handled by the csum forward 767engine:: 768 769 testpmd> csum parse-tunnel (on|off) (tx_port_id) 770 771If enabled, the csum forward engine will try to recognize supported 772tunnel headers (vxlan, gre, ipip). 773 774If disabled, treat tunnel packets as non-tunneled packets (a inner 775header is handled as a packet payload). 776 777.. note:: 778 779 The port argument is the TX port like in the ``csum set`` command. 780 781Example: 782 783Consider a packet in packet like the following:: 784 785 eth_out/ipv4_out/udp_out/vxlan/eth_in/ipv4_in/tcp_in 786 787* If parse-tunnel is enabled, the ``ip|udp|tcp|sctp`` parameters of ``csum set`` 788 command relate to the inner headers (here ``ipv4_in`` and ``tcp_in``), and the 789 ``outer-ip parameter`` relates to the outer headers (here ``ipv4_out``). 790 791* If parse-tunnel is disabled, the ``ip|udp|tcp|sctp`` parameters of ``csum set`` 792 command relate to the outer headers, here ``ipv4_out`` and ``udp_out``. 793 794csum show 795~~~~~~~~~ 796 797Display tx checksum offload configuration:: 798 799 testpmd> csum show (port_id) 800 801tso set 802~~~~~~~ 803 804Enable TCP Segmentation Offload (TSO) in the ``csum`` forwarding engine:: 805 806 testpmd> tso set (segsize) (port_id) 807 808.. note:: 809 810 Check the NIC datasheet for hardware limits. 811 812tso show 813~~~~~~~~ 814 815Display the status of TCP Segmentation Offload:: 816 817 testpmd> tso show (port_id) 818 819mac_addr add 820~~~~~~~~~~~~ 821 822Add an alternative MAC address to a port:: 823 824 testpmd> mac_addr add (port_id) (XX:XX:XX:XX:XX:XX) 825 826mac_addr remove 827~~~~~~~~~~~~~~~ 828 829Remove a MAC address from a port:: 830 831 testpmd> mac_addr remove (port_id) (XX:XX:XX:XX:XX:XX) 832 833mac_addr add (for VF) 834~~~~~~~~~~~~~~~~~~~~~ 835 836Add an alternative MAC address for a VF to a port:: 837 838 testpmd> mac_add add port (port_id) vf (vf_id) (XX:XX:XX:XX:XX:XX) 839 840mac_addr set (for VF) 841~~~~~~~~~~~~~~~~~~~~~ 842 843Set the MAC address for a VF from the PF:: 844 845 testpmd> set vf mac addr (port_id) (vf_id) (XX:XX:XX:XX:XX:XX) 846 847set port-uta 848~~~~~~~~~~~~ 849 850Set the unicast hash filter(s) on/off for a port:: 851 852 testpmd> set port (port_id) uta (XX:XX:XX:XX:XX:XX|all) (on|off) 853 854set promisc 855~~~~~~~~~~~ 856 857Set the promiscuous mode on for a port or for all ports. 858In promiscuous mode packets are not dropped if they aren't for the specified MAC address:: 859 860 testpmd> set promisc (port_id|all) (on|off) 861 862set allmulti 863~~~~~~~~~~~~ 864 865Set the allmulti mode for a port or for all ports:: 866 867 testpmd> set allmulti (port_id|all) (on|off) 868 869Same as the ifconfig (8) option. Controls how multicast packets are handled. 870 871set promisc (for VF) 872~~~~~~~~~~~~~~~~~~~~ 873 874Set the unicast promiscuous mode for a VF from PF. 875It's supported by Intel i40e NICs now. 876In promiscuous mode packets are not dropped if they aren't for the specified MAC address:: 877 878 testpmd> set vf promisc (port_id) (vf_id) (on|off) 879 880set allmulticast (for VF) 881~~~~~~~~~~~~~~~~~~~~~~~~~ 882 883Set the multicast promiscuous mode for a VF from PF. 884It's supported by Intel i40e NICs now. 885In promiscuous mode packets are not dropped if they aren't for the specified MAC address:: 886 887 testpmd> set vf allmulti (port_id) (vf_id) (on|off) 888 889set flow_ctrl rx 890~~~~~~~~~~~~~~~~ 891 892Set the link flow control parameter on a port:: 893 894 testpmd> set flow_ctrl rx (on|off) tx (on|off) (high_water) (low_water) \ 895 (pause_time) (send_xon) mac_ctrl_frame_fwd (on|off) \ 896 autoneg (on|off) (port_id) 897 898Where: 899 900* ``high_water`` (integer): High threshold value to trigger XOFF. 901 902* ``low_water`` (integer): Low threshold value to trigger XON. 903 904* ``pause_time`` (integer): Pause quota in the Pause frame. 905 906* ``send_xon`` (0/1): Send XON frame. 907 908* ``mac_ctrl_frame_fwd``: Enable receiving MAC control frames. 909 910* ``autoneg``: Change the auto-negotiation parameter. 911 912set pfc_ctrl rx 913~~~~~~~~~~~~~~~ 914 915Set the priority flow control parameter on a port:: 916 917 testpmd> set pfc_ctrl rx (on|off) tx (on|off) (high_water) (low_water) \ 918 (pause_time) (priority) (port_id) 919 920Where: 921 922* ``high_water`` (integer): High threshold value. 923 924* ``low_water`` (integer): Low threshold value. 925 926* ``pause_time`` (integer): Pause quota in the Pause frame. 927 928* ``priority`` (0-7): VLAN User Priority. 929 930set stat_qmap 931~~~~~~~~~~~~~ 932 933Set statistics mapping (qmapping 0..15) for RX/TX queue on port:: 934 935 testpmd> set stat_qmap (tx|rx) (port_id) (queue_id) (qmapping) 936 937For example, to set rx queue 2 on port 0 to mapping 5:: 938 939 testpmd>set stat_qmap rx 0 2 5 940 941set port - rx/tx (for VF) 942~~~~~~~~~~~~~~~~~~~~~~~~~ 943 944Set VF receive/transmit from a port:: 945 946 testpmd> set port (port_id) vf (vf_id) (rx|tx) (on|off) 947 948set port - mac address filter (for VF) 949~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ 950 951Add/Remove unicast or multicast MAC addr filter for a VF:: 952 953 testpmd> set port (port_id) vf (vf_id) (mac_addr) \ 954 (exact-mac|exact-mac-vlan|hashmac|hashmac-vlan) (on|off) 955 956set port - rx mode(for VF) 957~~~~~~~~~~~~~~~~~~~~~~~~~~ 958 959Set the VF receive mode of a port:: 960 961 testpmd> set port (port_id) vf (vf_id) \ 962 rxmode (AUPE|ROPE|BAM|MPE) (on|off) 963 964The available receive modes are: 965 966* ``AUPE``: Accepts untagged VLAN. 967 968* ``ROPE``: Accepts unicast hash. 969 970* ``BAM``: Accepts broadcast packets. 971 972* ``MPE``: Accepts all multicast packets. 973 974set port - tx_rate (for Queue) 975~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ 976 977Set TX rate limitation for a queue on a port:: 978 979 testpmd> set port (port_id) queue (queue_id) rate (rate_value) 980 981set port - tx_rate (for VF) 982~~~~~~~~~~~~~~~~~~~~~~~~~~~ 983 984Set TX rate limitation for queues in VF on a port:: 985 986 testpmd> set port (port_id) vf (vf_id) rate (rate_value) queue_mask (queue_mask) 987 988set port - mirror rule 989~~~~~~~~~~~~~~~~~~~~~~ 990 991Set pool or vlan type mirror rule for a port:: 992 993 testpmd> set port (port_id) mirror-rule (rule_id) \ 994 (pool-mirror-up|pool-mirror-down|vlan-mirror) \ 995 (poolmask|vlanid[,vlanid]*) dst-pool (pool_id) (on|off) 996 997Set link mirror rule for a port:: 998 999 testpmd> set port (port_id) mirror-rule (rule_id) \ 1000 (uplink-mirror|downlink-mirror) dst-pool (pool_id) (on|off) 1001 1002For example to enable mirror traffic with vlan 0,1 to pool 0:: 1003 1004 set port 0 mirror-rule 0 vlan-mirror 0,1 dst-pool 0 on 1005 1006reset port - mirror rule 1007~~~~~~~~~~~~~~~~~~~~~~~~ 1008 1009Reset a mirror rule for a port:: 1010 1011 testpmd> reset port (port_id) mirror-rule (rule_id) 1012 1013set flush_rx 1014~~~~~~~~~~~~ 1015 1016Set the flush on RX streams before forwarding. 1017The default is flush ``on``. 1018Mainly used with PCAP drivers to turn off the default behavior of flushing the first 512 packets on RX streams:: 1019 1020 testpmd> set flush_rx off 1021 1022set bypass mode 1023~~~~~~~~~~~~~~~ 1024 1025Set the bypass mode for the lowest port on bypass enabled NIC:: 1026 1027 testpmd> set bypass mode (normal|bypass|isolate) (port_id) 1028 1029set bypass event 1030~~~~~~~~~~~~~~~~ 1031 1032Set the event required to initiate specified bypass mode for the lowest port on a bypass enabled:: 1033 1034 testpmd> set bypass event (timeout|os_on|os_off|power_on|power_off) \ 1035 mode (normal|bypass|isolate) (port_id) 1036 1037Where: 1038 1039* ``timeout``: Enable bypass after watchdog timeout. 1040 1041* ``os_on``: Enable bypass when OS/board is powered on. 1042 1043* ``os_off``: Enable bypass when OS/board is powered off. 1044 1045* ``power_on``: Enable bypass when power supply is turned on. 1046 1047* ``power_off``: Enable bypass when power supply is turned off. 1048 1049 1050set bypass timeout 1051~~~~~~~~~~~~~~~~~~ 1052 1053Set the bypass watchdog timeout to ``n`` seconds where 0 = instant:: 1054 1055 testpmd> set bypass timeout (0|1.5|2|3|4|8|16|32) 1056 1057show bypass config 1058~~~~~~~~~~~~~~~~~~ 1059 1060Show the bypass configuration for a bypass enabled NIC using the lowest port on the NIC:: 1061 1062 testpmd> show bypass config (port_id) 1063 1064set link up 1065~~~~~~~~~~~ 1066 1067Set link up for a port:: 1068 1069 testpmd> set link-up port (port id) 1070 1071set link down 1072~~~~~~~~~~~~~ 1073 1074Set link down for a port:: 1075 1076 testpmd> set link-down port (port id) 1077 1078E-tag set 1079~~~~~~~~~ 1080 1081Enable E-tag insertion for a VF on a port:: 1082 1083 testpmd> E-tag set insertion on port-tag-id (value) port (port_id) vf (vf_id) 1084 1085Disable E-tag insertion for a VF on a port:: 1086 1087 testpmd> E-tag set insertion off port (port_id) vf (vf_id) 1088 1089Enable/disable E-tag stripping on a port:: 1090 1091 testpmd> E-tag set stripping (on|off) port (port_id) 1092 1093Enable/disable E-tag based forwarding on a port:: 1094 1095 testpmd> E-tag set forwarding (on|off) port (port_id) 1096 1097Add an E-tag forwarding filter on a port:: 1098 1099 testpmd> E-tag set filter add e-tag-id (value) dst-pool (pool_id) port (port_id) 1100 1101Delete an E-tag forwarding filter on a port:: 1102 testpmd> E-tag set filter del e-tag-id (value) port (port_id) 1103 1104 1105Port Functions 1106-------------- 1107 1108The following sections show functions for configuring ports. 1109 1110.. note:: 1111 1112 Port configuration changes only become active when forwarding is started/restarted. 1113 1114port attach 1115~~~~~~~~~~~ 1116 1117Attach a port specified by pci address or virtual device args:: 1118 1119 testpmd> port attach (identifier) 1120 1121To attach a new pci device, the device should be recognized by kernel first. 1122Then it should be moved under DPDK management. 1123Finally the port can be attached to testpmd. 1124 1125For example, to move a pci device using ixgbe under DPDK management: 1126 1127.. code-block:: console 1128 1129 # Check the status of the available devices. 1130 ./usertools/dpdk-devbind.py --status 1131 1132 Network devices using DPDK-compatible driver 1133 ============================================ 1134 <none> 1135 1136 Network devices using kernel driver 1137 =================================== 1138 0000:0a:00.0 '82599ES 10-Gigabit' if=eth2 drv=ixgbe unused= 1139 1140 1141 # Bind the device to igb_uio. 1142 sudo ./usertools/dpdk-devbind.py -b igb_uio 0000:0a:00.0 1143 1144 1145 # Recheck the status of the devices. 1146 ./usertools/dpdk-devbind.py --status 1147 Network devices using DPDK-compatible driver 1148 ============================================ 1149 0000:0a:00.0 '82599ES 10-Gigabit' drv=igb_uio unused= 1150 1151To attach a port created by virtual device, above steps are not needed. 1152 1153For example, to attach a port whose pci address is 0000:0a:00.0. 1154 1155.. code-block:: console 1156 1157 testpmd> port attach 0000:0a:00.0 1158 Attaching a new port... 1159 EAL: PCI device 0000:0a:00.0 on NUMA socket -1 1160 EAL: probe driver: 8086:10fb rte_ixgbe_pmd 1161 EAL: PCI memory mapped at 0x7f83bfa00000 1162 EAL: PCI memory mapped at 0x7f83bfa80000 1163 PMD: eth_ixgbe_dev_init(): MAC: 2, PHY: 18, SFP+: 5 1164 PMD: eth_ixgbe_dev_init(): port 0 vendorID=0x8086 deviceID=0x10fb 1165 Port 0 is attached. Now total ports is 1 1166 Done 1167 1168For example, to attach a port created by pcap PMD. 1169 1170.. code-block:: console 1171 1172 testpmd> port attach net_pcap0 1173 Attaching a new port... 1174 PMD: Initializing pmd_pcap for net_pcap0 1175 PMD: Creating pcap-backed ethdev on numa socket 0 1176 Port 0 is attached. Now total ports is 1 1177 Done 1178 1179In this case, identifier is ``net_pcap0``. 1180This identifier format is the same as ``--vdev`` format of DPDK applications. 1181 1182For example, to re-attach a bonded port which has been previously detached, 1183the mode and slave parameters must be given. 1184 1185.. code-block:: console 1186 1187 testpmd> port attach net_bond_0,mode=0,slave=1 1188 Attaching a new port... 1189 EAL: Initializing pmd_bond for net_bond_0 1190 EAL: Create bonded device net_bond_0 on port 0 in mode 0 on socket 0. 1191 Port 0 is attached. Now total ports is 1 1192 Done 1193 1194 1195port detach 1196~~~~~~~~~~~ 1197 1198Detach a specific port:: 1199 1200 testpmd> port detach (port_id) 1201 1202Before detaching a port, the port should be stopped and closed. 1203 1204For example, to detach a pci device port 0. 1205 1206.. code-block:: console 1207 1208 testpmd> port stop 0 1209 Stopping ports... 1210 Done 1211 testpmd> port close 0 1212 Closing ports... 1213 Done 1214 1215 testpmd> port detach 0 1216 Detaching a port... 1217 EAL: PCI device 0000:0a:00.0 on NUMA socket -1 1218 EAL: remove driver: 8086:10fb rte_ixgbe_pmd 1219 EAL: PCI memory unmapped at 0x7f83bfa00000 1220 EAL: PCI memory unmapped at 0x7f83bfa80000 1221 Done 1222 1223 1224For example, to detach a virtual device port 0. 1225 1226.. code-block:: console 1227 1228 testpmd> port stop 0 1229 Stopping ports... 1230 Done 1231 testpmd> port close 0 1232 Closing ports... 1233 Done 1234 1235 testpmd> port detach 0 1236 Detaching a port... 1237 PMD: Closing pcap ethdev on numa socket 0 1238 Port 'net_pcap0' is detached. Now total ports is 0 1239 Done 1240 1241To remove a pci device completely from the system, first detach the port from testpmd. 1242Then the device should be moved under kernel management. 1243Finally the device can be removed using kernel pci hotplug functionality. 1244 1245For example, to move a pci device under kernel management: 1246 1247.. code-block:: console 1248 1249 sudo ./usertools/dpdk-devbind.py -b ixgbe 0000:0a:00.0 1250 1251 ./usertools/dpdk-devbind.py --status 1252 1253 Network devices using DPDK-compatible driver 1254 ============================================ 1255 <none> 1256 1257 Network devices using kernel driver 1258 =================================== 1259 0000:0a:00.0 '82599ES 10-Gigabit' if=eth2 drv=ixgbe unused=igb_uio 1260 1261To remove a port created by a virtual device, above steps are not needed. 1262 1263port start 1264~~~~~~~~~~ 1265 1266Start all ports or a specific port:: 1267 1268 testpmd> port start (port_id|all) 1269 1270port stop 1271~~~~~~~~~ 1272 1273Stop all ports or a specific port:: 1274 1275 testpmd> port stop (port_id|all) 1276 1277port close 1278~~~~~~~~~~ 1279 1280Close all ports or a specific port:: 1281 1282 testpmd> port close (port_id|all) 1283 1284port start/stop queue 1285~~~~~~~~~~~~~~~~~~~~~ 1286 1287Start/stop a rx/tx queue on a specific port:: 1288 1289 testpmd> port (port_id) (rxq|txq) (queue_id) (start|stop) 1290 1291Only take effect when port is started. 1292 1293port config - speed 1294~~~~~~~~~~~~~~~~~~~ 1295 1296Set the speed and duplex mode for all ports or a specific port:: 1297 1298 testpmd> port config (port_id|all) speed (10|100|1000|10000|25000|40000|50000|100000|auto) \ 1299 duplex (half|full|auto) 1300 1301port config - queues/descriptors 1302~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ 1303 1304Set number of queues/descriptors for rxq, txq, rxd and txd:: 1305 1306 testpmd> port config all (rxq|txq|rxd|txd) (value) 1307 1308This is equivalent to the ``--rxq``, ``--txq``, ``--rxd`` and ``--txd`` command-line options. 1309 1310port config - max-pkt-len 1311~~~~~~~~~~~~~~~~~~~~~~~~~ 1312 1313Set the maximum packet length:: 1314 1315 testpmd> port config all max-pkt-len (value) 1316 1317This is equivalent to the ``--max-pkt-len`` command-line option. 1318 1319port config - CRC Strip 1320~~~~~~~~~~~~~~~~~~~~~~~ 1321 1322Set hardware CRC stripping on or off for all ports:: 1323 1324 testpmd> port config all crc-strip (on|off) 1325 1326CRC stripping is off by default. 1327 1328The ``on`` option is equivalent to the ``--crc-strip`` command-line option. 1329 1330port config - scatter 1331~~~~~~~~~~~~~~~~~~~~~~~ 1332 1333Set RX scatter mode on or off for all ports:: 1334 1335 testpmd> port config all scatter (on|off) 1336 1337RX scatter mode is off by default. 1338 1339The ``on`` option is equivalent to the ``--enable-scatter`` command-line option. 1340 1341port config - TX queue flags 1342~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ 1343 1344Set a hexadecimal bitmap of TX queue flags for all ports:: 1345 1346 testpmd> port config all txqflags value 1347 1348This command is equivalent to the ``--txqflags`` command-line option. 1349 1350port config - RX Checksum 1351~~~~~~~~~~~~~~~~~~~~~~~~~ 1352 1353Set hardware RX checksum offload to on or off for all ports:: 1354 1355 testpmd> port config all rx-cksum (on|off) 1356 1357Checksum offload is off by default. 1358 1359The ``on`` option is equivalent to the ``--enable-rx-cksum`` command-line option. 1360 1361port config - VLAN 1362~~~~~~~~~~~~~~~~~~ 1363 1364Set hardware VLAN on or off for all ports:: 1365 1366 testpmd> port config all hw-vlan (on|off) 1367 1368Hardware VLAN is on by default. 1369 1370The ``off`` option is equivalent to the ``--disable-hw-vlan`` command-line option. 1371 1372port config - VLAN filter 1373~~~~~~~~~~~~~~~~~~~~~~~~~ 1374 1375Set hardware VLAN filter on or off for all ports:: 1376 1377 testpmd> port config all hw-vlan-filter (on|off) 1378 1379Hardware VLAN filter is on by default. 1380 1381The ``off`` option is equivalent to the ``--disable-hw-vlan-filter`` command-line option. 1382 1383port config - VLAN strip 1384~~~~~~~~~~~~~~~~~~~~~~~~ 1385 1386Set hardware VLAN strip on or off for all ports:: 1387 1388 testpmd> port config all hw-vlan-strip (on|off) 1389 1390Hardware VLAN strip is on by default. 1391 1392The ``off`` option is equivalent to the ``--disable-hw-vlan-strip`` command-line option. 1393 1394port config - VLAN extend 1395~~~~~~~~~~~~~~~~~~~~~~~~~ 1396 1397Set hardware VLAN extend on or off for all ports:: 1398 1399 testpmd> port config all hw-vlan-extend (on|off) 1400 1401Hardware VLAN extend is off by default. 1402 1403The ``off`` option is equivalent to the ``--disable-hw-vlan-extend`` command-line option. 1404 1405port config - Drop Packets 1406~~~~~~~~~~~~~~~~~~~~~~~~~~ 1407 1408Set packet drop for packets with no descriptors on or off for all ports:: 1409 1410 testpmd> port config all drop-en (on|off) 1411 1412Packet dropping for packets with no descriptors is off by default. 1413 1414The ``on`` option is equivalent to the ``--enable-drop-en`` command-line option. 1415 1416port config - RSS 1417~~~~~~~~~~~~~~~~~ 1418 1419Set the RSS (Receive Side Scaling) mode on or off:: 1420 1421 testpmd> port config all rss (all|ip|tcp|udp|sctp|ether|port|vxlan|geneve|nvgre|none) 1422 1423RSS is on by default. 1424 1425The ``none`` option is equivalent to the ``--disable-rss`` command-line option. 1426 1427port config - RSS Reta 1428~~~~~~~~~~~~~~~~~~~~~~ 1429 1430Set the RSS (Receive Side Scaling) redirection table:: 1431 1432 testpmd> port config all rss reta (hash,queue)[,(hash,queue)] 1433 1434port config - DCB 1435~~~~~~~~~~~~~~~~~ 1436 1437Set the DCB mode for an individual port:: 1438 1439 testpmd> port config (port_id) dcb vt (on|off) (traffic_class) pfc (on|off) 1440 1441The traffic class should be 4 or 8. 1442 1443port config - Burst 1444~~~~~~~~~~~~~~~~~~~ 1445 1446Set the number of packets per burst:: 1447 1448 testpmd> port config all burst (value) 1449 1450This is equivalent to the ``--burst`` command-line option. 1451 1452port config - Threshold 1453~~~~~~~~~~~~~~~~~~~~~~~ 1454 1455Set thresholds for TX/RX queues:: 1456 1457 testpmd> port config all (threshold) (value) 1458 1459Where the threshold type can be: 1460 1461* ``txpt:`` Set the prefetch threshold register of the TX rings, 0 <= value <= 255. 1462 1463* ``txht:`` Set the host threshold register of the TX rings, 0 <= value <= 255. 1464 1465* ``txwt:`` Set the write-back threshold register of the TX rings, 0 <= value <= 255. 1466 1467* ``rxpt:`` Set the prefetch threshold register of the RX rings, 0 <= value <= 255. 1468 1469* ``rxht:`` Set the host threshold register of the RX rings, 0 <= value <= 255. 1470 1471* ``rxwt:`` Set the write-back threshold register of the RX rings, 0 <= value <= 255. 1472 1473* ``txfreet:`` Set the transmit free threshold of the TX rings, 0 <= value <= txd. 1474 1475* ``rxfreet:`` Set the transmit free threshold of the RX rings, 0 <= value <= rxd. 1476 1477* ``txrst:`` Set the transmit RS bit threshold of TX rings, 0 <= value <= txd. 1478 1479These threshold options are also available from the command-line. 1480 1481port config - E-tag 1482~~~~~~~~~~~~~~~~~~~ 1483 1484Set the value of ether-type for E-tag:: 1485 1486 testpmd> port config (port_id|all) l2-tunnel E-tag ether-type (value) 1487 1488Enable/disable the E-tag support:: 1489 1490 testpmd> port config (port_id|all) l2-tunnel E-tag (enable|disable) 1491 1492 1493Link Bonding Functions 1494---------------------- 1495 1496The Link Bonding functions make it possible to dynamically create and 1497manage link bonding devices from within testpmd interactive prompt. 1498 1499create bonded device 1500~~~~~~~~~~~~~~~~~~~~ 1501 1502Create a new bonding device:: 1503 1504 testpmd> create bonded device (mode) (socket) 1505 1506For example, to create a bonded device in mode 1 on socket 0:: 1507 1508 testpmd> create bonded 1 0 1509 created new bonded device (port X) 1510 1511add bonding slave 1512~~~~~~~~~~~~~~~~~ 1513 1514Adds Ethernet device to a Link Bonding device:: 1515 1516 testpmd> add bonding slave (slave id) (port id) 1517 1518For example, to add Ethernet device (port 6) to a Link Bonding device (port 10):: 1519 1520 testpmd> add bonding slave 6 10 1521 1522 1523remove bonding slave 1524~~~~~~~~~~~~~~~~~~~~ 1525 1526Removes an Ethernet slave device from a Link Bonding device:: 1527 1528 testpmd> remove bonding slave (slave id) (port id) 1529 1530For example, to remove Ethernet slave device (port 6) to a Link Bonding device (port 10):: 1531 1532 testpmd> remove bonding slave 6 10 1533 1534set bonding mode 1535~~~~~~~~~~~~~~~~ 1536 1537Set the Link Bonding mode of a Link Bonding device:: 1538 1539 testpmd> set bonding mode (value) (port id) 1540 1541For example, to set the bonding mode of a Link Bonding device (port 10) to broadcast (mode 3):: 1542 1543 testpmd> set bonding mode 3 10 1544 1545set bonding primary 1546~~~~~~~~~~~~~~~~~~~ 1547 1548Set an Ethernet slave device as the primary device on a Link Bonding device:: 1549 1550 testpmd> set bonding primary (slave id) (port id) 1551 1552For example, to set the Ethernet slave device (port 6) as the primary port of a Link Bonding device (port 10):: 1553 1554 testpmd> set bonding primary 6 10 1555 1556set bonding mac 1557~~~~~~~~~~~~~~~ 1558 1559Set the MAC address of a Link Bonding device:: 1560 1561 testpmd> set bonding mac (port id) (mac) 1562 1563For example, to set the MAC address of a Link Bonding device (port 10) to 00:00:00:00:00:01:: 1564 1565 testpmd> set bonding mac 10 00:00:00:00:00:01 1566 1567set bonding xmit_balance_policy 1568~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ 1569 1570Set the transmission policy for a Link Bonding device when it is in Balance XOR mode:: 1571 1572 testpmd> set bonding xmit_balance_policy (port_id) (l2|l23|l34) 1573 1574For example, set a Link Bonding device (port 10) to use a balance policy of layer 3+4 (IP addresses & UDP ports):: 1575 1576 testpmd> set bonding xmit_balance_policy 10 l34 1577 1578 1579set bonding mon_period 1580~~~~~~~~~~~~~~~~~~~~~~ 1581 1582Set the link status monitoring polling period in milliseconds for a bonding device. 1583 1584This adds support for PMD slave devices which do not support link status interrupts. 1585When the mon_period is set to a value greater than 0 then all PMD's which do not support 1586link status ISR will be queried every polling interval to check if their link status has changed:: 1587 1588 testpmd> set bonding mon_period (port_id) (value) 1589 1590For example, to set the link status monitoring polling period of bonded device (port 5) to 150ms:: 1591 1592 testpmd> set bonding mon_period 5 150 1593 1594 1595show bonding config 1596~~~~~~~~~~~~~~~~~~~ 1597 1598Show the current configuration of a Link Bonding device:: 1599 1600 testpmd> show bonding config (port id) 1601 1602For example, 1603to show the configuration a Link Bonding device (port 9) with 3 slave devices (1, 3, 4) 1604in balance mode with a transmission policy of layer 2+3:: 1605 1606 testpmd> show bonding config 9 1607 Bonding mode: 2 1608 Balance Xmit Policy: BALANCE_XMIT_POLICY_LAYER23 1609 Slaves (3): [1 3 4] 1610 Active Slaves (3): [1 3 4] 1611 Primary: [3] 1612 1613 1614Register Functions 1615------------------ 1616 1617The Register Functions can be used to read from and write to registers on the network card referenced by a port number. 1618This is mainly useful for debugging purposes. 1619Reference should be made to the appropriate datasheet for the network card for details on the register addresses 1620and fields that can be accessed. 1621 1622read reg 1623~~~~~~~~ 1624 1625Display the value of a port register:: 1626 1627 testpmd> read reg (port_id) (address) 1628 1629For example, to examine the Flow Director control register (FDIRCTL, 0x0000EE000) on an Intel 82599 10 GbE Controller:: 1630 1631 testpmd> read reg 0 0xEE00 1632 port 0 PCI register at offset 0xEE00: 0x4A060029 (1241907241) 1633 1634read regfield 1635~~~~~~~~~~~~~ 1636 1637Display a port register bit field:: 1638 1639 testpmd> read regfield (port_id) (address) (bit_x) (bit_y) 1640 1641For example, reading the lowest two bits from the register in the example above:: 1642 1643 testpmd> read regfield 0 0xEE00 0 1 1644 port 0 PCI register at offset 0xEE00: bits[0, 1]=0x1 (1) 1645 1646read regbit 1647~~~~~~~~~~~ 1648 1649Display a single port register bit:: 1650 1651 testpmd> read regbit (port_id) (address) (bit_x) 1652 1653For example, reading the lowest bit from the register in the example above:: 1654 1655 testpmd> read regbit 0 0xEE00 0 1656 port 0 PCI register at offset 0xEE00: bit 0=1 1657 1658write reg 1659~~~~~~~~~ 1660 1661Set the value of a port register:: 1662 1663 testpmd> write reg (port_id) (address) (value) 1664 1665For example, to clear a register:: 1666 1667 testpmd> write reg 0 0xEE00 0x0 1668 port 0 PCI register at offset 0xEE00: 0x00000000 (0) 1669 1670write regfield 1671~~~~~~~~~~~~~~ 1672 1673Set bit field of a port register:: 1674 1675 testpmd> write regfield (port_id) (address) (bit_x) (bit_y) (value) 1676 1677For example, writing to the register cleared in the example above:: 1678 1679 testpmd> write regfield 0 0xEE00 0 1 2 1680 port 0 PCI register at offset 0xEE00: 0x00000002 (2) 1681 1682write regbit 1683~~~~~~~~~~~~ 1684 1685Set single bit value of a port register:: 1686 1687 testpmd> write regbit (port_id) (address) (bit_x) (value) 1688 1689For example, to set the high bit in the register from the example above:: 1690 1691 testpmd> write regbit 0 0xEE00 31 1 1692 port 0 PCI register at offset 0xEE00: 0x8000000A (2147483658) 1693 1694 1695Filter Functions 1696---------------- 1697 1698This section details the available filter functions that are available. 1699 1700Note these functions interface the deprecated legacy filtering framework, 1701superseded by *rte_flow*. See `Flow rules management`_. 1702 1703ethertype_filter 1704~~~~~~~~~~~~~~~~~~~~ 1705 1706Add or delete a L2 Ethertype filter, which identify packets by their L2 Ethertype mainly assign them to a receive queue:: 1707 1708 ethertype_filter (port_id) (add|del) (mac_addr|mac_ignr) (mac_address) \ 1709 ethertype (ether_type) (drop|fwd) queue (queue_id) 1710 1711The available information parameters are: 1712 1713* ``port_id``: The port which the Ethertype filter assigned on. 1714 1715* ``mac_addr``: Compare destination mac address. 1716 1717* ``mac_ignr``: Ignore destination mac address match. 1718 1719* ``mac_address``: Destination mac address to match. 1720 1721* ``ether_type``: The EtherType value want to match, 1722 for example 0x0806 for ARP packet. 0x0800 (IPv4) and 0x86DD (IPv6) are invalid. 1723 1724* ``queue_id``: The receive queue associated with this EtherType filter. 1725 It is meaningless when deleting or dropping. 1726 1727Example, to add/remove an ethertype filter rule:: 1728 1729 testpmd> ethertype_filter 0 add mac_ignr 00:11:22:33:44:55 \ 1730 ethertype 0x0806 fwd queue 3 1731 1732 testpmd> ethertype_filter 0 del mac_ignr 00:11:22:33:44:55 \ 1733 ethertype 0x0806 fwd queue 3 1734 17352tuple_filter 1736~~~~~~~~~~~~~~~~~ 1737 1738Add or delete a 2-tuple filter, 1739which identifies packets by specific protocol and destination TCP/UDP port 1740and forwards packets into one of the receive queues:: 1741 1742 2tuple_filter (port_id) (add|del) dst_port (dst_port_value) \ 1743 protocol (protocol_value) mask (mask_value) \ 1744 tcp_flags (tcp_flags_value) priority (prio_value) \ 1745 queue (queue_id) 1746 1747The available information parameters are: 1748 1749* ``port_id``: The port which the 2-tuple filter assigned on. 1750 1751* ``dst_port_value``: Destination port in L4. 1752 1753* ``protocol_value``: IP L4 protocol. 1754 1755* ``mask_value``: Participates in the match or not by bit for field above, 1b means participate. 1756 1757* ``tcp_flags_value``: TCP control bits. The non-zero value is invalid, when the pro_value is not set to 0x06 (TCP). 1758 1759* ``prio_value``: Priority of this filter. 1760 1761* ``queue_id``: The receive queue associated with this 2-tuple filter. 1762 1763Example, to add/remove an 2tuple filter rule:: 1764 1765 testpmd> 2tuple_filter 0 add dst_port 32 protocol 0x06 mask 0x03 \ 1766 tcp_flags 0x02 priority 3 queue 3 1767 1768 testpmd> 2tuple_filter 0 del dst_port 32 protocol 0x06 mask 0x03 \ 1769 tcp_flags 0x02 priority 3 queue 3 1770 17715tuple_filter 1772~~~~~~~~~~~~~~~~~ 1773 1774Add or delete a 5-tuple filter, 1775which consists of a 5-tuple (protocol, source and destination IP addresses, source and destination TCP/UDP/SCTP port) 1776and routes packets into one of the receive queues:: 1777 1778 5tuple_filter (port_id) (add|del) dst_ip (dst_address) src_ip \ 1779 (src_address) dst_port (dst_port_value) \ 1780 src_port (src_port_value) protocol (protocol_value) \ 1781 mask (mask_value) tcp_flags (tcp_flags_value) \ 1782 priority (prio_value) queue (queue_id) 1783 1784The available information parameters are: 1785 1786* ``port_id``: The port which the 5-tuple filter assigned on. 1787 1788* ``dst_address``: Destination IP address. 1789 1790* ``src_address``: Source IP address. 1791 1792* ``dst_port_value``: TCP/UDP destination port. 1793 1794* ``src_port_value``: TCP/UDP source port. 1795 1796* ``protocol_value``: L4 protocol. 1797 1798* ``mask_value``: Participates in the match or not by bit for field above, 1b means participate 1799 1800* ``tcp_flags_value``: TCP control bits. The non-zero value is invalid, when the protocol_value is not set to 0x06 (TCP). 1801 1802* ``prio_value``: The priority of this filter. 1803 1804* ``queue_id``: The receive queue associated with this 5-tuple filter. 1805 1806Example, to add/remove an 5tuple filter rule:: 1807 1808 testpmd> 5tuple_filter 0 add dst_ip 2.2.2.5 src_ip 2.2.2.4 \ 1809 dst_port 64 src_port 32 protocol 0x06 mask 0x1F \ 1810 flags 0x0 priority 3 queue 3 1811 1812 testpmd> 5tuple_filter 0 del dst_ip 2.2.2.5 src_ip 2.2.2.4 \ 1813 dst_port 64 src_port 32 protocol 0x06 mask 0x1F \ 1814 flags 0x0 priority 3 queue 3 1815 1816syn_filter 1817~~~~~~~~~~ 1818 1819Using the SYN filter, TCP packets whose *SYN* flag is set can be forwarded to a separate queue:: 1820 1821 syn_filter (port_id) (add|del) priority (high|low) queue (queue_id) 1822 1823The available information parameters are: 1824 1825* ``port_id``: The port which the SYN filter assigned on. 1826 1827* ``high``: This SYN filter has higher priority than other filters. 1828 1829* ``low``: This SYN filter has lower priority than other filters. 1830 1831* ``queue_id``: The receive queue associated with this SYN filter 1832 1833Example:: 1834 1835 testpmd> syn_filter 0 add priority high queue 3 1836 1837flex_filter 1838~~~~~~~~~~~ 1839 1840With flex filter, packets can be recognized by any arbitrary pattern within the first 128 bytes of the packet 1841and routed into one of the receive queues:: 1842 1843 flex_filter (port_id) (add|del) len (len_value) bytes (bytes_value) \ 1844 mask (mask_value) priority (prio_value) queue (queue_id) 1845 1846The available information parameters are: 1847 1848* ``port_id``: The port which the Flex filter is assigned on. 1849 1850* ``len_value``: Filter length in bytes, no greater than 128. 1851 1852* ``bytes_value``: A string in hexadecimal, means the value the flex filter needs to match. 1853 1854* ``mask_value``: A string in hexadecimal, bit 1 means corresponding byte participates in the match. 1855 1856* ``prio_value``: The priority of this filter. 1857 1858* ``queue_id``: The receive queue associated with this Flex filter. 1859 1860Example:: 1861 1862 testpmd> flex_filter 0 add len 16 bytes 0x00000000000000000000000008060000 \ 1863 mask 000C priority 3 queue 3 1864 1865 testpmd> flex_filter 0 del len 16 bytes 0x00000000000000000000000008060000 \ 1866 mask 000C priority 3 queue 3 1867 1868 1869.. _testpmd_flow_director: 1870 1871flow_director_filter 1872~~~~~~~~~~~~~~~~~~~~ 1873 1874The Flow Director works in receive mode to identify specific flows or sets of flows and route them to specific queues. 1875 1876Four types of filtering are supported which are referred to as Perfect Match, Signature, Perfect-mac-vlan and 1877Perfect-tunnel filters, the match mode is set by the ``--pkt-filter-mode`` command-line parameter: 1878 1879* Perfect match filters. 1880 The hardware checks a match between the masked fields of the received packets and the programmed filters. 1881 The masked fields are for IP flow. 1882 1883* Signature filters. 1884 The hardware checks a match between a hash-based signature of the masked fields of the received packet. 1885 1886* Perfect-mac-vlan match filters. 1887 The hardware checks a match between the masked fields of the received packets and the programmed filters. 1888 The masked fields are for MAC VLAN flow. 1889 1890* Perfect-tunnel match filters. 1891 The hardware checks a match between the masked fields of the received packets and the programmed filters. 1892 The masked fields are for tunnel flow. 1893 1894The Flow Director filters can match the different fields for different type of packet: flow type, specific input set 1895per flow type and the flexible payload. 1896 1897The Flow Director can also mask out parts of all of these fields so that filters 1898are only applied to certain fields or parts of the fields. 1899 1900Different NICs may have different capabilities, command show port fdir (port_id) can be used to acquire the information. 1901 1902# Commands to add flow director filters of different flow types:: 1903 1904 flow_director_filter (port_id) mode IP (add|del|update) \ 1905 flow (ipv4-other|ipv4-frag|ipv6-other|ipv6-frag) \ 1906 src (src_ip_address) dst (dst_ip_address) \ 1907 tos (tos_value) proto (proto_value) ttl (ttl_value) \ 1908 vlan (vlan_value) flexbytes (flexbytes_value) \ 1909 (drop|fwd) pf|vf(vf_id) queue (queue_id) \ 1910 fd_id (fd_id_value) 1911 1912 flow_director_filter (port_id) mode IP (add|del|update) \ 1913 flow (ipv4-tcp|ipv4-udp|ipv6-tcp|ipv6-udp) \ 1914 src (src_ip_address) (src_port) \ 1915 dst (dst_ip_address) (dst_port) \ 1916 tos (tos_value) ttl (ttl_value) \ 1917 vlan (vlan_value) flexbytes (flexbytes_value) \ 1918 (drop|fwd) queue pf|vf(vf_id) (queue_id) \ 1919 fd_id (fd_id_value) 1920 1921 flow_director_filter (port_id) mode IP (add|del|update) \ 1922 flow (ipv4-sctp|ipv6-sctp) \ 1923 src (src_ip_address) (src_port) \ 1924 dst (dst_ip_address) (dst_port) \ 1925 tos (tos_value) ttl (ttl_value) \ 1926 tag (verification_tag) vlan (vlan_value) \ 1927 flexbytes (flexbytes_value) (drop|fwd) \ 1928 pf|vf(vf_id) queue (queue_id) fd_id (fd_id_value) 1929 1930 flow_director_filter (port_id) mode IP (add|del|update) flow l2_payload \ 1931 ether (ethertype) flexbytes (flexbytes_value) \ 1932 (drop|fwd) pf|vf(vf_id) queue (queue_id) 1933 fd_id (fd_id_value) 1934 1935 flow_director_filter (port_id) mode MAC-VLAN (add|del|update) \ 1936 mac (mac_address) vlan (vlan_value) \ 1937 flexbytes (flexbytes_value) (drop|fwd) \ 1938 queue (queue_id) fd_id (fd_id_value) 1939 1940 flow_director_filter (port_id) mode Tunnel (add|del|update) \ 1941 mac (mac_address) vlan (vlan_value) \ 1942 tunnel (NVGRE|VxLAN) tunnel-id (tunnel_id_value) \ 1943 flexbytes (flexbytes_value) (drop|fwd) \ 1944 queue (queue_id) fd_id (fd_id_value) 1945 1946For example, to add an ipv4-udp flow type filter:: 1947 1948 testpmd> flow_director_filter 0 mode IP add flow ipv4-udp src 2.2.2.3 32 \ 1949 dst 2.2.2.5 33 tos 2 ttl 40 vlan 0x1 flexbytes (0x88,0x48) \ 1950 fwd pf queue 1 fd_id 1 1951 1952For example, add an ipv4-other flow type filter:: 1953 1954 testpmd> flow_director_filter 0 mode IP add flow ipv4-other src 2.2.2.3 \ 1955 dst 2.2.2.5 tos 2 proto 20 ttl 40 vlan 0x1 \ 1956 flexbytes (0x88,0x48) fwd pf queue 1 fd_id 1 1957 1958flush_flow_director 1959~~~~~~~~~~~~~~~~~~~ 1960 1961Flush all flow director filters on a device:: 1962 1963 testpmd> flush_flow_director (port_id) 1964 1965Example, to flush all flow director filter on port 0:: 1966 1967 testpmd> flush_flow_director 0 1968 1969flow_director_mask 1970~~~~~~~~~~~~~~~~~~ 1971 1972Set flow director's input masks:: 1973 1974 flow_director_mask (port_id) mode IP vlan (vlan_value) \ 1975 src_mask (ipv4_src) (ipv6_src) (src_port) \ 1976 dst_mask (ipv4_dst) (ipv6_dst) (dst_port) 1977 1978 flow_director_mask (port_id) mode MAC-VLAN vlan (vlan_value) 1979 1980 flow_director_mask (port_id) mode Tunnel vlan (vlan_value) \ 1981 mac (mac_value) tunnel-type (tunnel_type_value) \ 1982 tunnel-id (tunnel_id_value) 1983 1984Example, to set flow director mask on port 0:: 1985 1986 testpmd> flow_director_mask 0 mode IP vlan 0xefff \ 1987 src_mask 255.255.255.255 \ 1988 FFFF:FFFF:FFFF:FFFF:FFFF:FFFF:FFFF:FFFF 0xFFFF \ 1989 dst_mask 255.255.255.255 \ 1990 FFFF:FFFF:FFFF:FFFF:FFFF:FFFF:FFFF:FFFF 0xFFFF 1991 1992flow_director_flex_mask 1993~~~~~~~~~~~~~~~~~~~~~~~ 1994 1995set masks of flow director's flexible payload based on certain flow type:: 1996 1997 testpmd> flow_director_flex_mask (port_id) \ 1998 flow (none|ipv4-other|ipv4-frag|ipv4-tcp|ipv4-udp|ipv4-sctp| \ 1999 ipv6-other|ipv6-frag|ipv6-tcp|ipv6-udp|ipv6-sctp| \ 2000 l2_payload|all) (mask) 2001 2002Example, to set flow director's flex mask for all flow type on port 0:: 2003 2004 testpmd> flow_director_flex_mask 0 flow all \ 2005 (0xff,0xff,0,0,0,0,0,0,0,0,0,0,0,0,0,0) 2006 2007 2008flow_director_flex_payload 2009~~~~~~~~~~~~~~~~~~~~~~~~~~ 2010 2011Configure flexible payload selection:: 2012 2013 flow_director_flex_payload (port_id) (raw|l2|l3|l4) (config) 2014 2015For example, to select the first 16 bytes from the offset 4 (bytes) of packet's payload as flexible payload:: 2016 2017 testpmd> flow_director_flex_payload 0 l4 \ 2018 (4,5,6,7,8,9,10,11,12,13,14,15,16,17,18,19) 2019 2020get_sym_hash_ena_per_port 2021~~~~~~~~~~~~~~~~~~~~~~~~~ 2022 2023Get symmetric hash enable configuration per port:: 2024 2025 get_sym_hash_ena_per_port (port_id) 2026 2027For example, to get symmetric hash enable configuration of port 1:: 2028 2029 testpmd> get_sym_hash_ena_per_port 1 2030 2031set_sym_hash_ena_per_port 2032~~~~~~~~~~~~~~~~~~~~~~~~~ 2033 2034Set symmetric hash enable configuration per port to enable or disable:: 2035 2036 set_sym_hash_ena_per_port (port_id) (enable|disable) 2037 2038For example, to set symmetric hash enable configuration of port 1 to enable:: 2039 2040 testpmd> set_sym_hash_ena_per_port 1 enable 2041 2042get_hash_global_config 2043~~~~~~~~~~~~~~~~~~~~~~ 2044 2045Get the global configurations of hash filters:: 2046 2047 get_hash_global_config (port_id) 2048 2049For example, to get the global configurations of hash filters of port 1:: 2050 2051 testpmd> get_hash_global_config 1 2052 2053set_hash_global_config 2054~~~~~~~~~~~~~~~~~~~~~~ 2055 2056Set the global configurations of hash filters:: 2057 2058 set_hash_global_config (port_id) (toeplitz|simple_xor|default) \ 2059 (ipv4|ipv4-frag|ipv4-tcp|ipv4-udp|ipv4-sctp|ipv4-other|ipv6|ipv6-frag| \ 2060 ipv6-tcp|ipv6-udp|ipv6-sctp|ipv6-other|l2_payload) \ 2061 (enable|disable) 2062 2063For example, to enable simple_xor for flow type of ipv6 on port 2:: 2064 2065 testpmd> set_hash_global_config 2 simple_xor ipv6 enable 2066 2067set_hash_input_set 2068~~~~~~~~~~~~~~~~~~ 2069 2070Set the input set for hash:: 2071 2072 set_hash_input_set (port_id) (ipv4-frag|ipv4-tcp|ipv4-udp|ipv4-sctp| \ 2073 ipv4-other|ipv6-frag|ipv6-tcp|ipv6-udp|ipv6-sctp|ipv6-other| \ 2074 l2_payload) (ovlan|ivlan|src-ipv4|dst-ipv4|src-ipv6|dst-ipv6|ipv4-tos| \ 2075 ipv4-proto|ipv6-tc|ipv6-next-header|udp-src-port|udp-dst-port| \ 2076 tcp-src-port|tcp-dst-port|sctp-src-port|sctp-dst-port|sctp-veri-tag| \ 2077 udp-key|gre-key|fld-1st|fld-2nd|fld-3rd|fld-4th|fld-5th|fld-6th|fld-7th| \ 2078 fld-8th|none) (select|add) 2079 2080For example, to add source IP to hash input set for flow type of ipv4-udp on port 0:: 2081 2082 testpmd> set_hash_input_set 0 ipv4-udp src-ipv4 add 2083 2084set_fdir_input_set 2085~~~~~~~~~~~~~~~~~~ 2086 2087The Flow Director filters can match the different fields for different type of packet, i.e. specific input set 2088on per flow type and the flexible payload. This command can be used to change input set for each flow type. 2089 2090Set the input set for flow director:: 2091 2092 set_fdir_input_set (port_id) (ipv4-frag|ipv4-tcp|ipv4-udp|ipv4-sctp| \ 2093 ipv4-other|ipv6|ipv6-frag|ipv6-tcp|ipv6-udp|ipv6-sctp|ipv6-other| \ 2094 l2_payload) (ivlan|ethertype|src-ipv4|dst-ipv4|src-ipv6|dst-ipv6|ipv4-tos| \ 2095 ipv4-proto|ipv4-ttl|ipv6-tc|ipv6-next-header|ipv6-hop-limits| \ 2096 tudp-src-port|udp-dst-port|cp-src-port|tcp-dst-port|sctp-src-port| \ 2097 sctp-dst-port|sctp-veri-tag|none) (select|add) 2098 2099For example to add source IP to FD input set for flow type of ipv4-udp on port 0:: 2100 2101 testpmd> set_fdir_input_set 0 ipv4-udp src-ipv4 add 2102 2103global_config 2104~~~~~~~~~~~~~ 2105 2106Set different GRE key length for input set:: 2107 2108 global_config (port_id) gre-key-len (number in bytes) 2109 2110For example to set GRE key length for input set to 4 bytes on port 0:: 2111 2112 testpmd> global_config 0 gre-key-len 4 2113 2114 2115.. _testpmd_rte_flow: 2116 2117Flow rules management 2118--------------------- 2119 2120Control of the generic flow API (*rte_flow*) is fully exposed through the 2121``flow`` command (validation, creation, destruction and queries). 2122 2123Considering *rte_flow* overlaps with all `Filter Functions`_, using both 2124features simultaneously may cause undefined side-effects and is therefore 2125not recommended. 2126 2127``flow`` syntax 2128~~~~~~~~~~~~~~~ 2129 2130Because the ``flow`` command uses dynamic tokens to handle the large number 2131of possible flow rules combinations, its behavior differs slightly from 2132other commands, in particular: 2133 2134- Pressing *?* or the *<tab>* key displays contextual help for the current 2135 token, not that of the entire command. 2136 2137- Optional and repeated parameters are supported (provided they are listed 2138 in the contextual help). 2139 2140The first parameter stands for the operation mode. Possible operations and 2141their general syntax are described below. They are covered in detail in the 2142following sections. 2143 2144- Check whether a flow rule can be created:: 2145 2146 flow validate {port_id} 2147 [group {group_id}] [priority {level}] [ingress] [egress] 2148 pattern {item} [/ {item} [...]] / end 2149 actions {action} [/ {action} [...]] / end 2150 2151- Create a flow rule:: 2152 2153 flow create {port_id} 2154 [group {group_id}] [priority {level}] [ingress] [egress] 2155 pattern {item} [/ {item} [...]] / end 2156 actions {action} [/ {action} [...]] / end 2157 2158- Destroy specific flow rules:: 2159 2160 flow destroy {port_id} rule {rule_id} [...] 2161 2162- Destroy all flow rules:: 2163 2164 flow flush {port_id} 2165 2166- Query an existing flow rule:: 2167 2168 flow query {port_id} {rule_id} {action} 2169 2170- List existing flow rules sorted by priority, filtered by group 2171 identifiers:: 2172 2173 flow list {port_id} [group {group_id}] [...] 2174 2175Validating flow rules 2176~~~~~~~~~~~~~~~~~~~~~ 2177 2178``flow validate`` reports whether a flow rule would be accepted by the 2179underlying device in its current state but stops short of creating it. It is 2180bound to ``rte_flow_validate()``:: 2181 2182 flow validate {port_id} 2183 [group {group_id}] [priority {level}] [ingress] [egress] 2184 pattern {item} [/ {item} [...]] / end 2185 actions {action} [/ {action} [...]] / end 2186 2187If successful, it will show:: 2188 2189 Flow rule validated 2190 2191Otherwise it will show an error message of the form:: 2192 2193 Caught error type [...] ([...]): [...] 2194 2195This command uses the same parameters as ``flow create``, their format is 2196described in `Creating flow rules`_. 2197 2198Check whether redirecting any Ethernet packet received on port 0 to RX queue 2199index 6 is supported:: 2200 2201 testpmd> flow validate 0 ingress pattern eth / end 2202 actions queue index 6 / end 2203 Flow rule validated 2204 testpmd> 2205 2206Port 0 does not support TCPv6 rules:: 2207 2208 testpmd> flow validate 0 ingress pattern eth / ipv6 / tcp / end 2209 actions drop / end 2210 Caught error type 9 (specific pattern item): Invalid argument 2211 testpmd> 2212 2213Creating flow rules 2214~~~~~~~~~~~~~~~~~~~ 2215 2216``flow create`` validates and creates the specified flow rule. It is bound 2217to ``rte_flow_create()``:: 2218 2219 flow create {port_id} 2220 [group {group_id}] [priority {level}] [ingress] [egress] 2221 pattern {item} [/ {item} [...]] / end 2222 actions {action} [/ {action} [...]] / end 2223 2224If successful, it will return a flow rule ID usable with other commands:: 2225 2226 Flow rule #[...] created 2227 2228Otherwise it will show an error message of the form:: 2229 2230 Caught error type [...] ([...]): [...] 2231 2232Parameters describe in the following order: 2233 2234- Attributes (*group*, *priority*, *ingress*, *egress* tokens). 2235- A matching pattern, starting with the *pattern* token and terminated by an 2236 *end* pattern item. 2237- Actions, starting with the *actions* token and terminated by an *end* 2238 action. 2239 2240These translate directly to *rte_flow* objects provided as-is to the 2241underlying functions. 2242 2243The shortest valid definition only comprises mandatory tokens:: 2244 2245 testpmd> flow create 0 pattern end actions end 2246 2247Note that PMDs may refuse rules that essentially do nothing such as this 2248one. 2249 2250**All unspecified object values are automatically initialized to 0.** 2251 2252Attributes 2253^^^^^^^^^^ 2254 2255These tokens affect flow rule attributes (``struct rte_flow_attr``) and are 2256specified before the ``pattern`` token. 2257 2258- ``group {group id}``: priority group. 2259- ``priority {level}``: priority level within group. 2260- ``ingress``: rule applies to ingress traffic. 2261- ``egress``: rule applies to egress traffic. 2262 2263Each instance of an attribute specified several times overrides the previous 2264value as shown below (group 4 is used):: 2265 2266 testpmd> flow create 0 group 42 group 24 group 4 [...] 2267 2268Note that once enabled, ``ingress`` and ``egress`` cannot be disabled. 2269 2270While not specifying a direction is an error, some rules may allow both 2271simultaneously. 2272 2273Most rules affect RX therefore contain the ``ingress`` token:: 2274 2275 testpmd> flow create 0 ingress pattern [...] 2276 2277Matching pattern 2278^^^^^^^^^^^^^^^^ 2279 2280A matching pattern starts after the ``pattern`` token. It is made of pattern 2281items and is terminated by a mandatory ``end`` item. 2282 2283Items are named after their type (*RTE_FLOW_ITEM_TYPE_* from ``enum 2284rte_flow_item_type``). 2285 2286The ``/`` token is used as a separator between pattern items as shown 2287below:: 2288 2289 testpmd> flow create 0 ingress pattern eth / ipv4 / udp / end [...] 2290 2291Note that protocol items like these must be stacked from lowest to highest 2292layer to make sense. For instance, the following rule is either invalid or 2293unlikely to match any packet:: 2294 2295 testpmd> flow create 0 ingress pattern eth / udp / ipv4 / end [...] 2296 2297More information on these restrictions can be found in the *rte_flow* 2298documentation. 2299 2300Several items support additional specification structures, for example 2301``ipv4`` allows specifying source and destination addresses as follows:: 2302 2303 testpmd> flow create 0 ingress pattern eth / ipv4 src is 10.1.1.1 2304 dst is 10.2.0.0 / end [...] 2305 2306This rule matches all IPv4 traffic with the specified properties. 2307 2308In this example, ``src`` and ``dst`` are field names of the underlying 2309``struct rte_flow_item_ipv4`` object. All item properties can be specified 2310in a similar fashion. 2311 2312The ``is`` token means that the subsequent value must be matched exactly, 2313and assigns ``spec`` and ``mask`` fields in ``struct rte_flow_item`` 2314accordingly. Possible assignment tokens are: 2315 2316- ``is``: match value perfectly (with full bit-mask). 2317- ``spec``: match value according to configured bit-mask. 2318- ``last``: specify upper bound to establish a range. 2319- ``mask``: specify bit-mask with relevant bits set to one. 2320- ``prefix``: generate bit-mask from a prefix length. 2321 2322These yield identical results:: 2323 2324 ipv4 src is 10.1.1.1 2325 2326:: 2327 2328 ipv4 src spec 10.1.1.1 src mask 255.255.255.255 2329 2330:: 2331 2332 ipv4 src spec 10.1.1.1 src prefix 32 2333 2334:: 2335 2336 ipv4 src is 10.1.1.1 src last 10.1.1.1 # range with a single value 2337 2338:: 2339 2340 ipv4 src is 10.1.1.1 src last 0 # 0 disables range 2341 2342Inclusive ranges can be defined with ``last``:: 2343 2344 ipv4 src is 10.1.1.1 src last 10.2.3.4 # 10.1.1.1 to 10.2.3.4 2345 2346Note that ``mask`` affects both ``spec`` and ``last``:: 2347 2348 ipv4 src is 10.1.1.1 src last 10.2.3.4 src mask 255.255.0.0 2349 # matches 10.1.0.0 to 10.2.255.255 2350 2351Properties can be modified multiple times:: 2352 2353 ipv4 src is 10.1.1.1 src is 10.1.2.3 src is 10.2.3.4 # matches 10.2.3.4 2354 2355:: 2356 2357 ipv4 src is 10.1.1.1 src prefix 24 src prefix 16 # matches 10.1.0.0/16 2358 2359Pattern items 2360^^^^^^^^^^^^^ 2361 2362This section lists supported pattern items and their attributes, if any. 2363 2364- ``end``: end list of pattern items. 2365 2366- ``void``: no-op pattern item. 2367 2368- ``invert``: perform actions when pattern does not match. 2369 2370- ``any``: match any protocol for the current layer. 2371 2372 - ``num {unsigned}``: number of layers covered. 2373 2374- ``pf``: match packets addressed to the physical function. 2375 2376- ``vf``: match packets addressed to a virtual function ID. 2377 2378 - ``id {unsigned}``: destination VF ID. 2379 2380- ``port``: device-specific physical port index to use. 2381 2382 - ``index {unsigned}``: physical port index. 2383 2384- ``raw``: match an arbitrary byte string. 2385 2386 - ``relative {boolean}``: look for pattern after the previous item. 2387 - ``search {boolean}``: search pattern from offset (see also limit). 2388 - ``offset {integer}``: absolute or relative offset for pattern. 2389 - ``limit {unsigned}``: search area limit for start of pattern. 2390 - ``pattern {string}``: byte string to look for. 2391 2392- ``eth``: match Ethernet header. 2393 2394 - ``dst {MAC-48}``: destination MAC. 2395 - ``src {MAC-48}``: source MAC. 2396 - ``type {unsigned}``: EtherType. 2397 2398- ``vlan``: match 802.1Q/ad VLAN tag. 2399 2400 - ``tpid {unsigned}``: tag protocol identifier. 2401 - ``tci {unsigned}``: tag control information. 2402 - ``pcp {unsigned}``: priority code point. 2403 - ``dei {unsigned}``: drop eligible indicator. 2404 - ``vid {unsigned}``: VLAN identifier. 2405 2406- ``ipv4``: match IPv4 header. 2407 2408 - ``tos {unsigned}``: type of service. 2409 - ``ttl {unsigned}``: time to live. 2410 - ``proto {unsigned}``: next protocol ID. 2411 - ``src {ipv4 address}``: source address. 2412 - ``dst {ipv4 address}``: destination address. 2413 2414- ``ipv6``: match IPv6 header. 2415 2416 - ``tc {unsigned}``: traffic class. 2417 - ``flow {unsigned}``: flow label. 2418 - ``proto {unsigned}``: protocol (next header). 2419 - ``hop {unsigned}``: hop limit. 2420 - ``src {ipv6 address}``: source address. 2421 - ``dst {ipv6 address}``: destination address. 2422 2423- ``icmp``: match ICMP header. 2424 2425 - ``type {unsigned}``: ICMP packet type. 2426 - ``code {unsigned}``: ICMP packet code. 2427 2428- ``udp``: match UDP header. 2429 2430 - ``src {unsigned}``: UDP source port. 2431 - ``dst {unsigned}``: UDP destination port. 2432 2433- ``tcp``: match TCP header. 2434 2435 - ``src {unsigned}``: TCP source port. 2436 - ``dst {unsigned}``: TCP destination port. 2437 2438- ``sctp``: match SCTP header. 2439 2440 - ``src {unsigned}``: SCTP source port. 2441 - ``dst {unsigned}``: SCTP destination port. 2442 - ``tag {unsigned}``: validation tag. 2443 - ``cksum {unsigned}``: checksum. 2444 2445- ``vxlan``: match VXLAN header. 2446 2447 - ``vni {unsigned}``: VXLAN identifier. 2448 2449Actions list 2450^^^^^^^^^^^^ 2451 2452A list of actions starts after the ``actions`` token in the same fashion as 2453`Matching pattern`_; actions are separated by ``/`` tokens and the list is 2454terminated by a mandatory ``end`` action. 2455 2456Actions are named after their type (*RTE_FLOW_ACTION_TYPE_* from ``enum 2457rte_flow_action_type``). 2458 2459Dropping all incoming UDPv4 packets can be expressed as follows:: 2460 2461 testpmd> flow create 0 ingress pattern eth / ipv4 / udp / end 2462 actions drop / end 2463 2464Several actions have configurable properties which must be specified when 2465there is no valid default value. For example, ``queue`` requires a target 2466queue index. 2467 2468This rule redirects incoming UDPv4 traffic to queue index 6:: 2469 2470 testpmd> flow create 0 ingress pattern eth / ipv4 / udp / end 2471 actions queue index 6 / end 2472 2473While this one could be rejected by PMDs (unspecified queue index):: 2474 2475 testpmd> flow create 0 ingress pattern eth / ipv4 / udp / end 2476 actions queue / end 2477 2478As defined by *rte_flow*, the list is not ordered, all actions of a given 2479rule are performed simultaneously. These are equivalent:: 2480 2481 queue index 6 / void / mark id 42 / end 2482 2483:: 2484 2485 void / mark id 42 / queue index 6 / end 2486 2487All actions in a list should have different types, otherwise only the last 2488action of a given type is taken into account:: 2489 2490 queue index 4 / queue index 5 / queue index 6 / end # will use queue 6 2491 2492:: 2493 2494 drop / drop / drop / end # drop is performed only once 2495 2496:: 2497 2498 mark id 42 / queue index 3 / mark id 24 / end # mark will be 24 2499 2500Considering they are performed simultaneously, opposite and overlapping 2501actions can sometimes be combined when the end result is unambiguous:: 2502 2503 drop / queue index 6 / end # drop has no effect 2504 2505:: 2506 2507 drop / dup index 6 / end # same as above 2508 2509:: 2510 2511 queue index 6 / rss queues 6 7 8 / end # queue has no effect 2512 2513:: 2514 2515 drop / passthru / end # drop has no effect 2516 2517Note that PMDs may still refuse such combinations. 2518 2519Actions 2520^^^^^^^ 2521 2522This section lists supported actions and their attributes, if any. 2523 2524- ``end``: end list of actions. 2525 2526- ``void``: no-op action. 2527 2528- ``passthru``: let subsequent rule process matched packets. 2529 2530- ``mark``: attach 32 bit value to packets. 2531 2532 - ``id {unsigned}``: 32 bit value to return with packets. 2533 2534- ``flag``: flag packets. 2535 2536- ``queue``: assign packets to a given queue index. 2537 2538 - ``index {unsigned}``: queue index to use. 2539 2540- ``drop``: drop packets (note: passthru has priority). 2541 2542- ``count``: enable counters for this rule. 2543 2544- ``dup``: duplicate packets to a given queue index. 2545 2546 - ``index {unsigned}``: queue index to duplicate packets to. 2547 2548- ``rss``: spread packets among several queues. 2549 2550 - ``queues [{unsigned} [...]] end``: queue indices to use. 2551 2552- ``pf``: redirect packets to physical device function. 2553 2554- ``vf``: redirect packets to virtual device function. 2555 2556 - ``original {boolean}``: use original VF ID if possible. 2557 - ``id {unsigned}``: VF ID to redirect packets to. 2558 2559Destroying flow rules 2560~~~~~~~~~~~~~~~~~~~~~ 2561 2562``flow destroy`` destroys one or more rules from their rule ID (as returned 2563by ``flow create``), this command calls ``rte_flow_destroy()`` as many 2564times as necessary:: 2565 2566 flow destroy {port_id} rule {rule_id} [...] 2567 2568If successful, it will show:: 2569 2570 Flow rule #[...] destroyed 2571 2572It does not report anything for rule IDs that do not exist. The usual error 2573message is shown when a rule cannot be destroyed:: 2574 2575 Caught error type [...] ([...]): [...] 2576 2577``flow flush`` destroys all rules on a device and does not take extra 2578arguments. It is bound to ``rte_flow_flush()``:: 2579 2580 flow flush {port_id} 2581 2582Any errors are reported as above. 2583 2584Creating several rules and destroying them:: 2585 2586 testpmd> flow create 0 ingress pattern eth / ipv6 / end 2587 actions queue index 2 / end 2588 Flow rule #0 created 2589 testpmd> flow create 0 ingress pattern eth / ipv4 / end 2590 actions queue index 3 / end 2591 Flow rule #1 created 2592 testpmd> flow destroy 0 rule 0 rule 1 2593 Flow rule #1 destroyed 2594 Flow rule #0 destroyed 2595 testpmd> 2596 2597The same result can be achieved using ``flow flush``:: 2598 2599 testpmd> flow create 0 ingress pattern eth / ipv6 / end 2600 actions queue index 2 / end 2601 Flow rule #0 created 2602 testpmd> flow create 0 ingress pattern eth / ipv4 / end 2603 actions queue index 3 / end 2604 Flow rule #1 created 2605 testpmd> flow flush 0 2606 testpmd> 2607 2608Non-existent rule IDs are ignored:: 2609 2610 testpmd> flow create 0 ingress pattern eth / ipv6 / end 2611 actions queue index 2 / end 2612 Flow rule #0 created 2613 testpmd> flow create 0 ingress pattern eth / ipv4 / end 2614 actions queue index 3 / end 2615 Flow rule #1 created 2616 testpmd> flow destroy 0 rule 42 rule 10 rule 2 2617 testpmd> 2618 testpmd> flow destroy 0 rule 0 2619 Flow rule #0 destroyed 2620 testpmd> 2621 2622Querying flow rules 2623~~~~~~~~~~~~~~~~~~~ 2624 2625``flow query`` queries a specific action of a flow rule having that 2626ability. Such actions collect information that can be reported using this 2627command. It is bound to ``rte_flow_query()``:: 2628 2629 flow query {port_id} {rule_id} {action} 2630 2631If successful, it will display either the retrieved data for known actions 2632or the following message:: 2633 2634 Cannot display result for action type [...] ([...]) 2635 2636Otherwise, it will complain either that the rule does not exist or that some 2637error occurred:: 2638 2639 Flow rule #[...] not found 2640 2641:: 2642 2643 Caught error type [...] ([...]): [...] 2644 2645Currently only the ``count`` action is supported. This action reports the 2646number of packets that hit the flow rule and the total number of bytes. Its 2647output has the following format:: 2648 2649 count: 2650 hits_set: [...] # whether "hits" contains a valid value 2651 bytes_set: [...] # whether "bytes" contains a valid value 2652 hits: [...] # number of packets 2653 bytes: [...] # number of bytes 2654 2655Querying counters for TCPv6 packets redirected to queue 6:: 2656 2657 testpmd> flow create 0 ingress pattern eth / ipv6 / tcp / end 2658 actions queue index 6 / count / end 2659 Flow rule #4 created 2660 testpmd> flow query 0 4 count 2661 count: 2662 hits_set: 1 2663 bytes_set: 0 2664 hits: 386446 2665 bytes: 0 2666 testpmd> 2667 2668Listing flow rules 2669~~~~~~~~~~~~~~~~~~ 2670 2671``flow list`` lists existing flow rules sorted by priority and optionally 2672filtered by group identifiers:: 2673 2674 flow list {port_id} [group {group_id}] [...] 2675 2676This command only fails with the following message if the device does not 2677exist:: 2678 2679 Invalid port [...] 2680 2681Output consists of a header line followed by a short description of each 2682flow rule, one per line. There is no output at all when no flow rules are 2683configured on the device:: 2684 2685 ID Group Prio Attr Rule 2686 [...] [...] [...] [...] [...] 2687 2688``Attr`` column flags: 2689 2690- ``i`` for ``ingress``. 2691- ``e`` for ``egress``. 2692 2693Creating several flow rules and listing them:: 2694 2695 testpmd> flow create 0 ingress pattern eth / ipv4 / end 2696 actions queue index 6 / end 2697 Flow rule #0 created 2698 testpmd> flow create 0 ingress pattern eth / ipv6 / end 2699 actions queue index 2 / end 2700 Flow rule #1 created 2701 testpmd> flow create 0 priority 5 ingress pattern eth / ipv4 / udp / end 2702 actions rss queues 6 7 8 end / end 2703 Flow rule #2 created 2704 testpmd> flow list 0 2705 ID Group Prio Attr Rule 2706 0 0 0 i- ETH IPV4 => QUEUE 2707 1 0 0 i- ETH IPV6 => QUEUE 2708 2 0 5 i- ETH IPV4 UDP => RSS 2709 testpmd> 2710 2711Rules are sorted by priority (i.e. group ID first, then priority level):: 2712 2713 testpmd> flow list 1 2714 ID Group Prio Attr Rule 2715 0 0 0 i- ETH => COUNT 2716 6 0 500 i- ETH IPV6 TCP => DROP COUNT 2717 5 0 1000 i- ETH IPV6 ICMP => QUEUE 2718 1 24 0 i- ETH IPV4 UDP => QUEUE 2719 4 24 10 i- ETH IPV4 TCP => DROP 2720 3 24 20 i- ETH IPV4 => DROP 2721 2 24 42 i- ETH IPV4 UDP => QUEUE 2722 7 63 0 i- ETH IPV6 UDP VXLAN => MARK QUEUE 2723 testpmd> 2724 2725Output can be limited to specific groups:: 2726 2727 testpmd> flow list 1 group 0 group 63 2728 ID Group Prio Attr Rule 2729 0 0 0 i- ETH => COUNT 2730 6 0 500 i- ETH IPV6 TCP => DROP COUNT 2731 5 0 1000 i- ETH IPV6 ICMP => QUEUE 2732 7 63 0 i- ETH IPV6 UDP VXLAN => MARK QUEUE 2733 testpmd> 2734