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 841~~~~~~~~~~~~ 842 843Set the default MAC address for a port:: 844 845 testpmd> mac_addr set (port_id) (XX:XX:XX:XX:XX:XX) 846 847mac_addr set (for VF) 848~~~~~~~~~~~~~~~~~~~~~ 849 850Set the MAC address for a VF from the PF:: 851 852 testpmd> set vf mac addr (port_id) (vf_id) (XX:XX:XX:XX:XX:XX) 853 854set port-uta 855~~~~~~~~~~~~ 856 857Set the unicast hash filter(s) on/off for a port:: 858 859 testpmd> set port (port_id) uta (XX:XX:XX:XX:XX:XX|all) (on|off) 860 861set promisc 862~~~~~~~~~~~ 863 864Set the promiscuous mode on for a port or for all ports. 865In promiscuous mode packets are not dropped if they aren't for the specified MAC address:: 866 867 testpmd> set promisc (port_id|all) (on|off) 868 869set allmulti 870~~~~~~~~~~~~ 871 872Set the allmulti mode for a port or for all ports:: 873 874 testpmd> set allmulti (port_id|all) (on|off) 875 876Same as the ifconfig (8) option. Controls how multicast packets are handled. 877 878set promisc (for VF) 879~~~~~~~~~~~~~~~~~~~~ 880 881Set the unicast promiscuous mode for a VF from PF. 882It's supported by Intel i40e NICs now. 883In promiscuous mode packets are not dropped if they aren't for the specified MAC address:: 884 885 testpmd> set vf promisc (port_id) (vf_id) (on|off) 886 887set allmulticast (for VF) 888~~~~~~~~~~~~~~~~~~~~~~~~~ 889 890Set the multicast promiscuous mode for a VF from PF. 891It's supported by Intel i40e NICs now. 892In promiscuous mode packets are not dropped if they aren't for the specified MAC address:: 893 894 testpmd> set vf allmulti (port_id) (vf_id) (on|off) 895 896set flow_ctrl rx 897~~~~~~~~~~~~~~~~ 898 899Set the link flow control parameter on a port:: 900 901 testpmd> set flow_ctrl rx (on|off) tx (on|off) (high_water) (low_water) \ 902 (pause_time) (send_xon) mac_ctrl_frame_fwd (on|off) \ 903 autoneg (on|off) (port_id) 904 905Where: 906 907* ``high_water`` (integer): High threshold value to trigger XOFF. 908 909* ``low_water`` (integer): Low threshold value to trigger XON. 910 911* ``pause_time`` (integer): Pause quota in the Pause frame. 912 913* ``send_xon`` (0/1): Send XON frame. 914 915* ``mac_ctrl_frame_fwd``: Enable receiving MAC control frames. 916 917* ``autoneg``: Change the auto-negotiation parameter. 918 919set pfc_ctrl rx 920~~~~~~~~~~~~~~~ 921 922Set the priority flow control parameter on a port:: 923 924 testpmd> set pfc_ctrl rx (on|off) tx (on|off) (high_water) (low_water) \ 925 (pause_time) (priority) (port_id) 926 927Where: 928 929* ``high_water`` (integer): High threshold value. 930 931* ``low_water`` (integer): Low threshold value. 932 933* ``pause_time`` (integer): Pause quota in the Pause frame. 934 935* ``priority`` (0-7): VLAN User Priority. 936 937set stat_qmap 938~~~~~~~~~~~~~ 939 940Set statistics mapping (qmapping 0..15) for RX/TX queue on port:: 941 942 testpmd> set stat_qmap (tx|rx) (port_id) (queue_id) (qmapping) 943 944For example, to set rx queue 2 on port 0 to mapping 5:: 945 946 testpmd>set stat_qmap rx 0 2 5 947 948set port - rx/tx (for VF) 949~~~~~~~~~~~~~~~~~~~~~~~~~ 950 951Set VF receive/transmit from a port:: 952 953 testpmd> set port (port_id) vf (vf_id) (rx|tx) (on|off) 954 955set port - mac address filter (for VF) 956~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ 957 958Add/Remove unicast or multicast MAC addr filter for a VF:: 959 960 testpmd> set port (port_id) vf (vf_id) (mac_addr) \ 961 (exact-mac|exact-mac-vlan|hashmac|hashmac-vlan) (on|off) 962 963set port - rx mode(for VF) 964~~~~~~~~~~~~~~~~~~~~~~~~~~ 965 966Set the VF receive mode of a port:: 967 968 testpmd> set port (port_id) vf (vf_id) \ 969 rxmode (AUPE|ROPE|BAM|MPE) (on|off) 970 971The available receive modes are: 972 973* ``AUPE``: Accepts untagged VLAN. 974 975* ``ROPE``: Accepts unicast hash. 976 977* ``BAM``: Accepts broadcast packets. 978 979* ``MPE``: Accepts all multicast packets. 980 981set port - tx_rate (for Queue) 982~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ 983 984Set TX rate limitation for a queue on a port:: 985 986 testpmd> set port (port_id) queue (queue_id) rate (rate_value) 987 988set port - tx_rate (for VF) 989~~~~~~~~~~~~~~~~~~~~~~~~~~~ 990 991Set TX rate limitation for queues in VF on a port:: 992 993 testpmd> set port (port_id) vf (vf_id) rate (rate_value) queue_mask (queue_mask) 994 995set port - mirror rule 996~~~~~~~~~~~~~~~~~~~~~~ 997 998Set pool or vlan type mirror rule for a port:: 999 1000 testpmd> set port (port_id) mirror-rule (rule_id) \ 1001 (pool-mirror-up|pool-mirror-down|vlan-mirror) \ 1002 (poolmask|vlanid[,vlanid]*) dst-pool (pool_id) (on|off) 1003 1004Set link mirror rule for a port:: 1005 1006 testpmd> set port (port_id) mirror-rule (rule_id) \ 1007 (uplink-mirror|downlink-mirror) dst-pool (pool_id) (on|off) 1008 1009For example to enable mirror traffic with vlan 0,1 to pool 0:: 1010 1011 set port 0 mirror-rule 0 vlan-mirror 0,1 dst-pool 0 on 1012 1013reset port - mirror rule 1014~~~~~~~~~~~~~~~~~~~~~~~~ 1015 1016Reset a mirror rule for a port:: 1017 1018 testpmd> reset port (port_id) mirror-rule (rule_id) 1019 1020set flush_rx 1021~~~~~~~~~~~~ 1022 1023Set the flush on RX streams before forwarding. 1024The default is flush ``on``. 1025Mainly used with PCAP drivers to turn off the default behavior of flushing the first 512 packets on RX streams:: 1026 1027 testpmd> set flush_rx off 1028 1029set bypass mode 1030~~~~~~~~~~~~~~~ 1031 1032Set the bypass mode for the lowest port on bypass enabled NIC:: 1033 1034 testpmd> set bypass mode (normal|bypass|isolate) (port_id) 1035 1036set bypass event 1037~~~~~~~~~~~~~~~~ 1038 1039Set the event required to initiate specified bypass mode for the lowest port on a bypass enabled:: 1040 1041 testpmd> set bypass event (timeout|os_on|os_off|power_on|power_off) \ 1042 mode (normal|bypass|isolate) (port_id) 1043 1044Where: 1045 1046* ``timeout``: Enable bypass after watchdog timeout. 1047 1048* ``os_on``: Enable bypass when OS/board is powered on. 1049 1050* ``os_off``: Enable bypass when OS/board is powered off. 1051 1052* ``power_on``: Enable bypass when power supply is turned on. 1053 1054* ``power_off``: Enable bypass when power supply is turned off. 1055 1056 1057set bypass timeout 1058~~~~~~~~~~~~~~~~~~ 1059 1060Set the bypass watchdog timeout to ``n`` seconds where 0 = instant:: 1061 1062 testpmd> set bypass timeout (0|1.5|2|3|4|8|16|32) 1063 1064show bypass config 1065~~~~~~~~~~~~~~~~~~ 1066 1067Show the bypass configuration for a bypass enabled NIC using the lowest port on the NIC:: 1068 1069 testpmd> show bypass config (port_id) 1070 1071set link up 1072~~~~~~~~~~~ 1073 1074Set link up for a port:: 1075 1076 testpmd> set link-up port (port id) 1077 1078set link down 1079~~~~~~~~~~~~~ 1080 1081Set link down for a port:: 1082 1083 testpmd> set link-down port (port id) 1084 1085E-tag set 1086~~~~~~~~~ 1087 1088Enable E-tag insertion for a VF on a port:: 1089 1090 testpmd> E-tag set insertion on port-tag-id (value) port (port_id) vf (vf_id) 1091 1092Disable E-tag insertion for a VF on a port:: 1093 1094 testpmd> E-tag set insertion off port (port_id) vf (vf_id) 1095 1096Enable/disable E-tag stripping on a port:: 1097 1098 testpmd> E-tag set stripping (on|off) port (port_id) 1099 1100Enable/disable E-tag based forwarding on a port:: 1101 1102 testpmd> E-tag set forwarding (on|off) port (port_id) 1103 1104Add an E-tag forwarding filter on a port:: 1105 1106 testpmd> E-tag set filter add e-tag-id (value) dst-pool (pool_id) port (port_id) 1107 1108Delete an E-tag forwarding filter on a port:: 1109 testpmd> E-tag set filter del e-tag-id (value) port (port_id) 1110 1111 1112Port Functions 1113-------------- 1114 1115The following sections show functions for configuring ports. 1116 1117.. note:: 1118 1119 Port configuration changes only become active when forwarding is started/restarted. 1120 1121port attach 1122~~~~~~~~~~~ 1123 1124Attach a port specified by pci address or virtual device args:: 1125 1126 testpmd> port attach (identifier) 1127 1128To attach a new pci device, the device should be recognized by kernel first. 1129Then it should be moved under DPDK management. 1130Finally the port can be attached to testpmd. 1131 1132For example, to move a pci device using ixgbe under DPDK management: 1133 1134.. code-block:: console 1135 1136 # Check the status of the available devices. 1137 ./usertools/dpdk-devbind.py --status 1138 1139 Network devices using DPDK-compatible driver 1140 ============================================ 1141 <none> 1142 1143 Network devices using kernel driver 1144 =================================== 1145 0000:0a:00.0 '82599ES 10-Gigabit' if=eth2 drv=ixgbe unused= 1146 1147 1148 # Bind the device to igb_uio. 1149 sudo ./usertools/dpdk-devbind.py -b igb_uio 0000:0a:00.0 1150 1151 1152 # Recheck the status of the devices. 1153 ./usertools/dpdk-devbind.py --status 1154 Network devices using DPDK-compatible driver 1155 ============================================ 1156 0000:0a:00.0 '82599ES 10-Gigabit' drv=igb_uio unused= 1157 1158To attach a port created by virtual device, above steps are not needed. 1159 1160For example, to attach a port whose pci address is 0000:0a:00.0. 1161 1162.. code-block:: console 1163 1164 testpmd> port attach 0000:0a:00.0 1165 Attaching a new port... 1166 EAL: PCI device 0000:0a:00.0 on NUMA socket -1 1167 EAL: probe driver: 8086:10fb rte_ixgbe_pmd 1168 EAL: PCI memory mapped at 0x7f83bfa00000 1169 EAL: PCI memory mapped at 0x7f83bfa80000 1170 PMD: eth_ixgbe_dev_init(): MAC: 2, PHY: 18, SFP+: 5 1171 PMD: eth_ixgbe_dev_init(): port 0 vendorID=0x8086 deviceID=0x10fb 1172 Port 0 is attached. Now total ports is 1 1173 Done 1174 1175For example, to attach a port created by pcap PMD. 1176 1177.. code-block:: console 1178 1179 testpmd> port attach net_pcap0 1180 Attaching a new port... 1181 PMD: Initializing pmd_pcap for net_pcap0 1182 PMD: Creating pcap-backed ethdev on numa socket 0 1183 Port 0 is attached. Now total ports is 1 1184 Done 1185 1186In this case, identifier is ``net_pcap0``. 1187This identifier format is the same as ``--vdev`` format of DPDK applications. 1188 1189For example, to re-attach a bonded port which has been previously detached, 1190the mode and slave parameters must be given. 1191 1192.. code-block:: console 1193 1194 testpmd> port attach net_bond_0,mode=0,slave=1 1195 Attaching a new port... 1196 EAL: Initializing pmd_bond for net_bond_0 1197 EAL: Create bonded device net_bond_0 on port 0 in mode 0 on socket 0. 1198 Port 0 is attached. Now total ports is 1 1199 Done 1200 1201 1202port detach 1203~~~~~~~~~~~ 1204 1205Detach a specific port:: 1206 1207 testpmd> port detach (port_id) 1208 1209Before detaching a port, the port should be stopped and closed. 1210 1211For example, to detach a pci device port 0. 1212 1213.. code-block:: console 1214 1215 testpmd> port stop 0 1216 Stopping ports... 1217 Done 1218 testpmd> port close 0 1219 Closing ports... 1220 Done 1221 1222 testpmd> port detach 0 1223 Detaching a port... 1224 EAL: PCI device 0000:0a:00.0 on NUMA socket -1 1225 EAL: remove driver: 8086:10fb rte_ixgbe_pmd 1226 EAL: PCI memory unmapped at 0x7f83bfa00000 1227 EAL: PCI memory unmapped at 0x7f83bfa80000 1228 Done 1229 1230 1231For example, to detach a virtual device port 0. 1232 1233.. code-block:: console 1234 1235 testpmd> port stop 0 1236 Stopping ports... 1237 Done 1238 testpmd> port close 0 1239 Closing ports... 1240 Done 1241 1242 testpmd> port detach 0 1243 Detaching a port... 1244 PMD: Closing pcap ethdev on numa socket 0 1245 Port 'net_pcap0' is detached. Now total ports is 0 1246 Done 1247 1248To remove a pci device completely from the system, first detach the port from testpmd. 1249Then the device should be moved under kernel management. 1250Finally the device can be removed using kernel pci hotplug functionality. 1251 1252For example, to move a pci device under kernel management: 1253 1254.. code-block:: console 1255 1256 sudo ./usertools/dpdk-devbind.py -b ixgbe 0000:0a:00.0 1257 1258 ./usertools/dpdk-devbind.py --status 1259 1260 Network devices using DPDK-compatible driver 1261 ============================================ 1262 <none> 1263 1264 Network devices using kernel driver 1265 =================================== 1266 0000:0a:00.0 '82599ES 10-Gigabit' if=eth2 drv=ixgbe unused=igb_uio 1267 1268To remove a port created by a virtual device, above steps are not needed. 1269 1270port start 1271~~~~~~~~~~ 1272 1273Start all ports or a specific port:: 1274 1275 testpmd> port start (port_id|all) 1276 1277port stop 1278~~~~~~~~~ 1279 1280Stop all ports or a specific port:: 1281 1282 testpmd> port stop (port_id|all) 1283 1284port close 1285~~~~~~~~~~ 1286 1287Close all ports or a specific port:: 1288 1289 testpmd> port close (port_id|all) 1290 1291port start/stop queue 1292~~~~~~~~~~~~~~~~~~~~~ 1293 1294Start/stop a rx/tx queue on a specific port:: 1295 1296 testpmd> port (port_id) (rxq|txq) (queue_id) (start|stop) 1297 1298Only take effect when port is started. 1299 1300port config - speed 1301~~~~~~~~~~~~~~~~~~~ 1302 1303Set the speed and duplex mode for all ports or a specific port:: 1304 1305 testpmd> port config (port_id|all) speed (10|100|1000|10000|25000|40000|50000|100000|auto) \ 1306 duplex (half|full|auto) 1307 1308port config - queues/descriptors 1309~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ 1310 1311Set number of queues/descriptors for rxq, txq, rxd and txd:: 1312 1313 testpmd> port config all (rxq|txq|rxd|txd) (value) 1314 1315This is equivalent to the ``--rxq``, ``--txq``, ``--rxd`` and ``--txd`` command-line options. 1316 1317port config - max-pkt-len 1318~~~~~~~~~~~~~~~~~~~~~~~~~ 1319 1320Set the maximum packet length:: 1321 1322 testpmd> port config all max-pkt-len (value) 1323 1324This is equivalent to the ``--max-pkt-len`` command-line option. 1325 1326port config - CRC Strip 1327~~~~~~~~~~~~~~~~~~~~~~~ 1328 1329Set hardware CRC stripping on or off for all ports:: 1330 1331 testpmd> port config all crc-strip (on|off) 1332 1333CRC stripping is off by default. 1334 1335The ``on`` option is equivalent to the ``--crc-strip`` command-line option. 1336 1337port config - scatter 1338~~~~~~~~~~~~~~~~~~~~~~~ 1339 1340Set RX scatter mode on or off for all ports:: 1341 1342 testpmd> port config all scatter (on|off) 1343 1344RX scatter mode is off by default. 1345 1346The ``on`` option is equivalent to the ``--enable-scatter`` command-line option. 1347 1348port config - TX queue flags 1349~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ 1350 1351Set a hexadecimal bitmap of TX queue flags for all ports:: 1352 1353 testpmd> port config all txqflags value 1354 1355This command is equivalent to the ``--txqflags`` command-line option. 1356 1357port config - RX Checksum 1358~~~~~~~~~~~~~~~~~~~~~~~~~ 1359 1360Set hardware RX checksum offload to on or off for all ports:: 1361 1362 testpmd> port config all rx-cksum (on|off) 1363 1364Checksum offload is off by default. 1365 1366The ``on`` option is equivalent to the ``--enable-rx-cksum`` command-line option. 1367 1368port config - VLAN 1369~~~~~~~~~~~~~~~~~~ 1370 1371Set hardware VLAN on or off for all ports:: 1372 1373 testpmd> port config all hw-vlan (on|off) 1374 1375Hardware VLAN is on by default. 1376 1377The ``off`` option is equivalent to the ``--disable-hw-vlan`` command-line option. 1378 1379port config - VLAN filter 1380~~~~~~~~~~~~~~~~~~~~~~~~~ 1381 1382Set hardware VLAN filter on or off for all ports:: 1383 1384 testpmd> port config all hw-vlan-filter (on|off) 1385 1386Hardware VLAN filter is on by default. 1387 1388The ``off`` option is equivalent to the ``--disable-hw-vlan-filter`` command-line option. 1389 1390port config - VLAN strip 1391~~~~~~~~~~~~~~~~~~~~~~~~ 1392 1393Set hardware VLAN strip on or off for all ports:: 1394 1395 testpmd> port config all hw-vlan-strip (on|off) 1396 1397Hardware VLAN strip is on by default. 1398 1399The ``off`` option is equivalent to the ``--disable-hw-vlan-strip`` command-line option. 1400 1401port config - VLAN extend 1402~~~~~~~~~~~~~~~~~~~~~~~~~ 1403 1404Set hardware VLAN extend on or off for all ports:: 1405 1406 testpmd> port config all hw-vlan-extend (on|off) 1407 1408Hardware VLAN extend is off by default. 1409 1410The ``off`` option is equivalent to the ``--disable-hw-vlan-extend`` command-line option. 1411 1412port config - Drop Packets 1413~~~~~~~~~~~~~~~~~~~~~~~~~~ 1414 1415Set packet drop for packets with no descriptors on or off for all ports:: 1416 1417 testpmd> port config all drop-en (on|off) 1418 1419Packet dropping for packets with no descriptors is off by default. 1420 1421The ``on`` option is equivalent to the ``--enable-drop-en`` command-line option. 1422 1423port config - RSS 1424~~~~~~~~~~~~~~~~~ 1425 1426Set the RSS (Receive Side Scaling) mode on or off:: 1427 1428 testpmd> port config all rss (all|ip|tcp|udp|sctp|ether|port|vxlan|geneve|nvgre|none) 1429 1430RSS is on by default. 1431 1432The ``none`` option is equivalent to the ``--disable-rss`` command-line option. 1433 1434port config - RSS Reta 1435~~~~~~~~~~~~~~~~~~~~~~ 1436 1437Set the RSS (Receive Side Scaling) redirection table:: 1438 1439 testpmd> port config all rss reta (hash,queue)[,(hash,queue)] 1440 1441port config - DCB 1442~~~~~~~~~~~~~~~~~ 1443 1444Set the DCB mode for an individual port:: 1445 1446 testpmd> port config (port_id) dcb vt (on|off) (traffic_class) pfc (on|off) 1447 1448The traffic class should be 4 or 8. 1449 1450port config - Burst 1451~~~~~~~~~~~~~~~~~~~ 1452 1453Set the number of packets per burst:: 1454 1455 testpmd> port config all burst (value) 1456 1457This is equivalent to the ``--burst`` command-line option. 1458 1459port config - Threshold 1460~~~~~~~~~~~~~~~~~~~~~~~ 1461 1462Set thresholds for TX/RX queues:: 1463 1464 testpmd> port config all (threshold) (value) 1465 1466Where the threshold type can be: 1467 1468* ``txpt:`` Set the prefetch threshold register of the TX rings, 0 <= value <= 255. 1469 1470* ``txht:`` Set the host threshold register of the TX rings, 0 <= value <= 255. 1471 1472* ``txwt:`` Set the write-back threshold register of the TX rings, 0 <= value <= 255. 1473 1474* ``rxpt:`` Set the prefetch threshold register of the RX rings, 0 <= value <= 255. 1475 1476* ``rxht:`` Set the host threshold register of the RX rings, 0 <= value <= 255. 1477 1478* ``rxwt:`` Set the write-back threshold register of the RX rings, 0 <= value <= 255. 1479 1480* ``txfreet:`` Set the transmit free threshold of the TX rings, 0 <= value <= txd. 1481 1482* ``rxfreet:`` Set the transmit free threshold of the RX rings, 0 <= value <= rxd. 1483 1484* ``txrst:`` Set the transmit RS bit threshold of TX rings, 0 <= value <= txd. 1485 1486These threshold options are also available from the command-line. 1487 1488port config - E-tag 1489~~~~~~~~~~~~~~~~~~~ 1490 1491Set the value of ether-type for E-tag:: 1492 1493 testpmd> port config (port_id|all) l2-tunnel E-tag ether-type (value) 1494 1495Enable/disable the E-tag support:: 1496 1497 testpmd> port config (port_id|all) l2-tunnel E-tag (enable|disable) 1498 1499 1500Link Bonding Functions 1501---------------------- 1502 1503The Link Bonding functions make it possible to dynamically create and 1504manage link bonding devices from within testpmd interactive prompt. 1505 1506create bonded device 1507~~~~~~~~~~~~~~~~~~~~ 1508 1509Create a new bonding device:: 1510 1511 testpmd> create bonded device (mode) (socket) 1512 1513For example, to create a bonded device in mode 1 on socket 0:: 1514 1515 testpmd> create bonded 1 0 1516 created new bonded device (port X) 1517 1518add bonding slave 1519~~~~~~~~~~~~~~~~~ 1520 1521Adds Ethernet device to a Link Bonding device:: 1522 1523 testpmd> add bonding slave (slave id) (port id) 1524 1525For example, to add Ethernet device (port 6) to a Link Bonding device (port 10):: 1526 1527 testpmd> add bonding slave 6 10 1528 1529 1530remove bonding slave 1531~~~~~~~~~~~~~~~~~~~~ 1532 1533Removes an Ethernet slave device from a Link Bonding device:: 1534 1535 testpmd> remove bonding slave (slave id) (port id) 1536 1537For example, to remove Ethernet slave device (port 6) to a Link Bonding device (port 10):: 1538 1539 testpmd> remove bonding slave 6 10 1540 1541set bonding mode 1542~~~~~~~~~~~~~~~~ 1543 1544Set the Link Bonding mode of a Link Bonding device:: 1545 1546 testpmd> set bonding mode (value) (port id) 1547 1548For example, to set the bonding mode of a Link Bonding device (port 10) to broadcast (mode 3):: 1549 1550 testpmd> set bonding mode 3 10 1551 1552set bonding primary 1553~~~~~~~~~~~~~~~~~~~ 1554 1555Set an Ethernet slave device as the primary device on a Link Bonding device:: 1556 1557 testpmd> set bonding primary (slave id) (port id) 1558 1559For example, to set the Ethernet slave device (port 6) as the primary port of a Link Bonding device (port 10):: 1560 1561 testpmd> set bonding primary 6 10 1562 1563set bonding mac 1564~~~~~~~~~~~~~~~ 1565 1566Set the MAC address of a Link Bonding device:: 1567 1568 testpmd> set bonding mac (port id) (mac) 1569 1570For example, to set the MAC address of a Link Bonding device (port 10) to 00:00:00:00:00:01:: 1571 1572 testpmd> set bonding mac 10 00:00:00:00:00:01 1573 1574set bonding xmit_balance_policy 1575~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ 1576 1577Set the transmission policy for a Link Bonding device when it is in Balance XOR mode:: 1578 1579 testpmd> set bonding xmit_balance_policy (port_id) (l2|l23|l34) 1580 1581For example, set a Link Bonding device (port 10) to use a balance policy of layer 3+4 (IP addresses & UDP ports):: 1582 1583 testpmd> set bonding xmit_balance_policy 10 l34 1584 1585 1586set bonding mon_period 1587~~~~~~~~~~~~~~~~~~~~~~ 1588 1589Set the link status monitoring polling period in milliseconds for a bonding device. 1590 1591This adds support for PMD slave devices which do not support link status interrupts. 1592When the mon_period is set to a value greater than 0 then all PMD's which do not support 1593link status ISR will be queried every polling interval to check if their link status has changed:: 1594 1595 testpmd> set bonding mon_period (port_id) (value) 1596 1597For example, to set the link status monitoring polling period of bonded device (port 5) to 150ms:: 1598 1599 testpmd> set bonding mon_period 5 150 1600 1601 1602show bonding config 1603~~~~~~~~~~~~~~~~~~~ 1604 1605Show the current configuration of a Link Bonding device:: 1606 1607 testpmd> show bonding config (port id) 1608 1609For example, 1610to show the configuration a Link Bonding device (port 9) with 3 slave devices (1, 3, 4) 1611in balance mode with a transmission policy of layer 2+3:: 1612 1613 testpmd> show bonding config 9 1614 Bonding mode: 2 1615 Balance Xmit Policy: BALANCE_XMIT_POLICY_LAYER23 1616 Slaves (3): [1 3 4] 1617 Active Slaves (3): [1 3 4] 1618 Primary: [3] 1619 1620 1621Register Functions 1622------------------ 1623 1624The Register Functions can be used to read from and write to registers on the network card referenced by a port number. 1625This is mainly useful for debugging purposes. 1626Reference should be made to the appropriate datasheet for the network card for details on the register addresses 1627and fields that can be accessed. 1628 1629read reg 1630~~~~~~~~ 1631 1632Display the value of a port register:: 1633 1634 testpmd> read reg (port_id) (address) 1635 1636For example, to examine the Flow Director control register (FDIRCTL, 0x0000EE000) on an Intel 82599 10 GbE Controller:: 1637 1638 testpmd> read reg 0 0xEE00 1639 port 0 PCI register at offset 0xEE00: 0x4A060029 (1241907241) 1640 1641read regfield 1642~~~~~~~~~~~~~ 1643 1644Display a port register bit field:: 1645 1646 testpmd> read regfield (port_id) (address) (bit_x) (bit_y) 1647 1648For example, reading the lowest two bits from the register in the example above:: 1649 1650 testpmd> read regfield 0 0xEE00 0 1 1651 port 0 PCI register at offset 0xEE00: bits[0, 1]=0x1 (1) 1652 1653read regbit 1654~~~~~~~~~~~ 1655 1656Display a single port register bit:: 1657 1658 testpmd> read regbit (port_id) (address) (bit_x) 1659 1660For example, reading the lowest bit from the register in the example above:: 1661 1662 testpmd> read regbit 0 0xEE00 0 1663 port 0 PCI register at offset 0xEE00: bit 0=1 1664 1665write reg 1666~~~~~~~~~ 1667 1668Set the value of a port register:: 1669 1670 testpmd> write reg (port_id) (address) (value) 1671 1672For example, to clear a register:: 1673 1674 testpmd> write reg 0 0xEE00 0x0 1675 port 0 PCI register at offset 0xEE00: 0x00000000 (0) 1676 1677write regfield 1678~~~~~~~~~~~~~~ 1679 1680Set bit field of a port register:: 1681 1682 testpmd> write regfield (port_id) (address) (bit_x) (bit_y) (value) 1683 1684For example, writing to the register cleared in the example above:: 1685 1686 testpmd> write regfield 0 0xEE00 0 1 2 1687 port 0 PCI register at offset 0xEE00: 0x00000002 (2) 1688 1689write regbit 1690~~~~~~~~~~~~ 1691 1692Set single bit value of a port register:: 1693 1694 testpmd> write regbit (port_id) (address) (bit_x) (value) 1695 1696For example, to set the high bit in the register from the example above:: 1697 1698 testpmd> write regbit 0 0xEE00 31 1 1699 port 0 PCI register at offset 0xEE00: 0x8000000A (2147483658) 1700 1701 1702Filter Functions 1703---------------- 1704 1705This section details the available filter functions that are available. 1706 1707Note these functions interface the deprecated legacy filtering framework, 1708superseded by *rte_flow*. See `Flow rules management`_. 1709 1710ethertype_filter 1711~~~~~~~~~~~~~~~~~~~~ 1712 1713Add or delete a L2 Ethertype filter, which identify packets by their L2 Ethertype mainly assign them to a receive queue:: 1714 1715 ethertype_filter (port_id) (add|del) (mac_addr|mac_ignr) (mac_address) \ 1716 ethertype (ether_type) (drop|fwd) queue (queue_id) 1717 1718The available information parameters are: 1719 1720* ``port_id``: The port which the Ethertype filter assigned on. 1721 1722* ``mac_addr``: Compare destination mac address. 1723 1724* ``mac_ignr``: Ignore destination mac address match. 1725 1726* ``mac_address``: Destination mac address to match. 1727 1728* ``ether_type``: The EtherType value want to match, 1729 for example 0x0806 for ARP packet. 0x0800 (IPv4) and 0x86DD (IPv6) are invalid. 1730 1731* ``queue_id``: The receive queue associated with this EtherType filter. 1732 It is meaningless when deleting or dropping. 1733 1734Example, to add/remove an ethertype filter rule:: 1735 1736 testpmd> ethertype_filter 0 add mac_ignr 00:11:22:33:44:55 \ 1737 ethertype 0x0806 fwd queue 3 1738 1739 testpmd> ethertype_filter 0 del mac_ignr 00:11:22:33:44:55 \ 1740 ethertype 0x0806 fwd queue 3 1741 17422tuple_filter 1743~~~~~~~~~~~~~~~~~ 1744 1745Add or delete a 2-tuple filter, 1746which identifies packets by specific protocol and destination TCP/UDP port 1747and forwards packets into one of the receive queues:: 1748 1749 2tuple_filter (port_id) (add|del) dst_port (dst_port_value) \ 1750 protocol (protocol_value) mask (mask_value) \ 1751 tcp_flags (tcp_flags_value) priority (prio_value) \ 1752 queue (queue_id) 1753 1754The available information parameters are: 1755 1756* ``port_id``: The port which the 2-tuple filter assigned on. 1757 1758* ``dst_port_value``: Destination port in L4. 1759 1760* ``protocol_value``: IP L4 protocol. 1761 1762* ``mask_value``: Participates in the match or not by bit for field above, 1b means participate. 1763 1764* ``tcp_flags_value``: TCP control bits. The non-zero value is invalid, when the pro_value is not set to 0x06 (TCP). 1765 1766* ``prio_value``: Priority of this filter. 1767 1768* ``queue_id``: The receive queue associated with this 2-tuple filter. 1769 1770Example, to add/remove an 2tuple filter rule:: 1771 1772 testpmd> 2tuple_filter 0 add dst_port 32 protocol 0x06 mask 0x03 \ 1773 tcp_flags 0x02 priority 3 queue 3 1774 1775 testpmd> 2tuple_filter 0 del dst_port 32 protocol 0x06 mask 0x03 \ 1776 tcp_flags 0x02 priority 3 queue 3 1777 17785tuple_filter 1779~~~~~~~~~~~~~~~~~ 1780 1781Add or delete a 5-tuple filter, 1782which consists of a 5-tuple (protocol, source and destination IP addresses, source and destination TCP/UDP/SCTP port) 1783and routes packets into one of the receive queues:: 1784 1785 5tuple_filter (port_id) (add|del) dst_ip (dst_address) src_ip \ 1786 (src_address) dst_port (dst_port_value) \ 1787 src_port (src_port_value) protocol (protocol_value) \ 1788 mask (mask_value) tcp_flags (tcp_flags_value) \ 1789 priority (prio_value) queue (queue_id) 1790 1791The available information parameters are: 1792 1793* ``port_id``: The port which the 5-tuple filter assigned on. 1794 1795* ``dst_address``: Destination IP address. 1796 1797* ``src_address``: Source IP address. 1798 1799* ``dst_port_value``: TCP/UDP destination port. 1800 1801* ``src_port_value``: TCP/UDP source port. 1802 1803* ``protocol_value``: L4 protocol. 1804 1805* ``mask_value``: Participates in the match or not by bit for field above, 1b means participate 1806 1807* ``tcp_flags_value``: TCP control bits. The non-zero value is invalid, when the protocol_value is not set to 0x06 (TCP). 1808 1809* ``prio_value``: The priority of this filter. 1810 1811* ``queue_id``: The receive queue associated with this 5-tuple filter. 1812 1813Example, to add/remove an 5tuple filter rule:: 1814 1815 testpmd> 5tuple_filter 0 add dst_ip 2.2.2.5 src_ip 2.2.2.4 \ 1816 dst_port 64 src_port 32 protocol 0x06 mask 0x1F \ 1817 flags 0x0 priority 3 queue 3 1818 1819 testpmd> 5tuple_filter 0 del dst_ip 2.2.2.5 src_ip 2.2.2.4 \ 1820 dst_port 64 src_port 32 protocol 0x06 mask 0x1F \ 1821 flags 0x0 priority 3 queue 3 1822 1823syn_filter 1824~~~~~~~~~~ 1825 1826Using the SYN filter, TCP packets whose *SYN* flag is set can be forwarded to a separate queue:: 1827 1828 syn_filter (port_id) (add|del) priority (high|low) queue (queue_id) 1829 1830The available information parameters are: 1831 1832* ``port_id``: The port which the SYN filter assigned on. 1833 1834* ``high``: This SYN filter has higher priority than other filters. 1835 1836* ``low``: This SYN filter has lower priority than other filters. 1837 1838* ``queue_id``: The receive queue associated with this SYN filter 1839 1840Example:: 1841 1842 testpmd> syn_filter 0 add priority high queue 3 1843 1844flex_filter 1845~~~~~~~~~~~ 1846 1847With flex filter, packets can be recognized by any arbitrary pattern within the first 128 bytes of the packet 1848and routed into one of the receive queues:: 1849 1850 flex_filter (port_id) (add|del) len (len_value) bytes (bytes_value) \ 1851 mask (mask_value) priority (prio_value) queue (queue_id) 1852 1853The available information parameters are: 1854 1855* ``port_id``: The port which the Flex filter is assigned on. 1856 1857* ``len_value``: Filter length in bytes, no greater than 128. 1858 1859* ``bytes_value``: A string in hexadecimal, means the value the flex filter needs to match. 1860 1861* ``mask_value``: A string in hexadecimal, bit 1 means corresponding byte participates in the match. 1862 1863* ``prio_value``: The priority of this filter. 1864 1865* ``queue_id``: The receive queue associated with this Flex filter. 1866 1867Example:: 1868 1869 testpmd> flex_filter 0 add len 16 bytes 0x00000000000000000000000008060000 \ 1870 mask 000C priority 3 queue 3 1871 1872 testpmd> flex_filter 0 del len 16 bytes 0x00000000000000000000000008060000 \ 1873 mask 000C priority 3 queue 3 1874 1875 1876.. _testpmd_flow_director: 1877 1878flow_director_filter 1879~~~~~~~~~~~~~~~~~~~~ 1880 1881The Flow Director works in receive mode to identify specific flows or sets of flows and route them to specific queues. 1882 1883Four types of filtering are supported which are referred to as Perfect Match, Signature, Perfect-mac-vlan and 1884Perfect-tunnel filters, the match mode is set by the ``--pkt-filter-mode`` command-line parameter: 1885 1886* Perfect 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 IP flow. 1889 1890* Signature filters. 1891 The hardware checks a match between a hash-based signature of the masked fields of the received packet. 1892 1893* Perfect-mac-vlan match filters. 1894 The hardware checks a match between the masked fields of the received packets and the programmed filters. 1895 The masked fields are for MAC VLAN flow. 1896 1897* Perfect-tunnel match filters. 1898 The hardware checks a match between the masked fields of the received packets and the programmed filters. 1899 The masked fields are for tunnel flow. 1900 1901The Flow Director filters can match the different fields for different type of packet: flow type, specific input set 1902per flow type and the flexible payload. 1903 1904The Flow Director can also mask out parts of all of these fields so that filters 1905are only applied to certain fields or parts of the fields. 1906 1907Different NICs may have different capabilities, command show port fdir (port_id) can be used to acquire the information. 1908 1909# Commands to add flow director filters of different flow types:: 1910 1911 flow_director_filter (port_id) mode IP (add|del|update) \ 1912 flow (ipv4-other|ipv4-frag|ipv6-other|ipv6-frag) \ 1913 src (src_ip_address) dst (dst_ip_address) \ 1914 tos (tos_value) proto (proto_value) ttl (ttl_value) \ 1915 vlan (vlan_value) flexbytes (flexbytes_value) \ 1916 (drop|fwd) pf|vf(vf_id) queue (queue_id) \ 1917 fd_id (fd_id_value) 1918 1919 flow_director_filter (port_id) mode IP (add|del|update) \ 1920 flow (ipv4-tcp|ipv4-udp|ipv6-tcp|ipv6-udp) \ 1921 src (src_ip_address) (src_port) \ 1922 dst (dst_ip_address) (dst_port) \ 1923 tos (tos_value) ttl (ttl_value) \ 1924 vlan (vlan_value) flexbytes (flexbytes_value) \ 1925 (drop|fwd) queue pf|vf(vf_id) (queue_id) \ 1926 fd_id (fd_id_value) 1927 1928 flow_director_filter (port_id) mode IP (add|del|update) \ 1929 flow (ipv4-sctp|ipv6-sctp) \ 1930 src (src_ip_address) (src_port) \ 1931 dst (dst_ip_address) (dst_port) \ 1932 tos (tos_value) ttl (ttl_value) \ 1933 tag (verification_tag) vlan (vlan_value) \ 1934 flexbytes (flexbytes_value) (drop|fwd) \ 1935 pf|vf(vf_id) queue (queue_id) fd_id (fd_id_value) 1936 1937 flow_director_filter (port_id) mode IP (add|del|update) flow l2_payload \ 1938 ether (ethertype) flexbytes (flexbytes_value) \ 1939 (drop|fwd) pf|vf(vf_id) queue (queue_id) 1940 fd_id (fd_id_value) 1941 1942 flow_director_filter (port_id) mode MAC-VLAN (add|del|update) \ 1943 mac (mac_address) vlan (vlan_value) \ 1944 flexbytes (flexbytes_value) (drop|fwd) \ 1945 queue (queue_id) fd_id (fd_id_value) 1946 1947 flow_director_filter (port_id) mode Tunnel (add|del|update) \ 1948 mac (mac_address) vlan (vlan_value) \ 1949 tunnel (NVGRE|VxLAN) tunnel-id (tunnel_id_value) \ 1950 flexbytes (flexbytes_value) (drop|fwd) \ 1951 queue (queue_id) fd_id (fd_id_value) 1952 1953For example, to add an ipv4-udp flow type filter:: 1954 1955 testpmd> flow_director_filter 0 mode IP add flow ipv4-udp src 2.2.2.3 32 \ 1956 dst 2.2.2.5 33 tos 2 ttl 40 vlan 0x1 flexbytes (0x88,0x48) \ 1957 fwd pf queue 1 fd_id 1 1958 1959For example, add an ipv4-other flow type filter:: 1960 1961 testpmd> flow_director_filter 0 mode IP add flow ipv4-other src 2.2.2.3 \ 1962 dst 2.2.2.5 tos 2 proto 20 ttl 40 vlan 0x1 \ 1963 flexbytes (0x88,0x48) fwd pf queue 1 fd_id 1 1964 1965flush_flow_director 1966~~~~~~~~~~~~~~~~~~~ 1967 1968Flush all flow director filters on a device:: 1969 1970 testpmd> flush_flow_director (port_id) 1971 1972Example, to flush all flow director filter on port 0:: 1973 1974 testpmd> flush_flow_director 0 1975 1976flow_director_mask 1977~~~~~~~~~~~~~~~~~~ 1978 1979Set flow director's input masks:: 1980 1981 flow_director_mask (port_id) mode IP vlan (vlan_value) \ 1982 src_mask (ipv4_src) (ipv6_src) (src_port) \ 1983 dst_mask (ipv4_dst) (ipv6_dst) (dst_port) 1984 1985 flow_director_mask (port_id) mode MAC-VLAN vlan (vlan_value) 1986 1987 flow_director_mask (port_id) mode Tunnel vlan (vlan_value) \ 1988 mac (mac_value) tunnel-type (tunnel_type_value) \ 1989 tunnel-id (tunnel_id_value) 1990 1991Example, to set flow director mask on port 0:: 1992 1993 testpmd> flow_director_mask 0 mode IP vlan 0xefff \ 1994 src_mask 255.255.255.255 \ 1995 FFFF:FFFF:FFFF:FFFF:FFFF:FFFF:FFFF:FFFF 0xFFFF \ 1996 dst_mask 255.255.255.255 \ 1997 FFFF:FFFF:FFFF:FFFF:FFFF:FFFF:FFFF:FFFF 0xFFFF 1998 1999flow_director_flex_mask 2000~~~~~~~~~~~~~~~~~~~~~~~ 2001 2002set masks of flow director's flexible payload based on certain flow type:: 2003 2004 testpmd> flow_director_flex_mask (port_id) \ 2005 flow (none|ipv4-other|ipv4-frag|ipv4-tcp|ipv4-udp|ipv4-sctp| \ 2006 ipv6-other|ipv6-frag|ipv6-tcp|ipv6-udp|ipv6-sctp| \ 2007 l2_payload|all) (mask) 2008 2009Example, to set flow director's flex mask for all flow type on port 0:: 2010 2011 testpmd> flow_director_flex_mask 0 flow all \ 2012 (0xff,0xff,0,0,0,0,0,0,0,0,0,0,0,0,0,0) 2013 2014 2015flow_director_flex_payload 2016~~~~~~~~~~~~~~~~~~~~~~~~~~ 2017 2018Configure flexible payload selection:: 2019 2020 flow_director_flex_payload (port_id) (raw|l2|l3|l4) (config) 2021 2022For example, to select the first 16 bytes from the offset 4 (bytes) of packet's payload as flexible payload:: 2023 2024 testpmd> flow_director_flex_payload 0 l4 \ 2025 (4,5,6,7,8,9,10,11,12,13,14,15,16,17,18,19) 2026 2027get_sym_hash_ena_per_port 2028~~~~~~~~~~~~~~~~~~~~~~~~~ 2029 2030Get symmetric hash enable configuration per port:: 2031 2032 get_sym_hash_ena_per_port (port_id) 2033 2034For example, to get symmetric hash enable configuration of port 1:: 2035 2036 testpmd> get_sym_hash_ena_per_port 1 2037 2038set_sym_hash_ena_per_port 2039~~~~~~~~~~~~~~~~~~~~~~~~~ 2040 2041Set symmetric hash enable configuration per port to enable or disable:: 2042 2043 set_sym_hash_ena_per_port (port_id) (enable|disable) 2044 2045For example, to set symmetric hash enable configuration of port 1 to enable:: 2046 2047 testpmd> set_sym_hash_ena_per_port 1 enable 2048 2049get_hash_global_config 2050~~~~~~~~~~~~~~~~~~~~~~ 2051 2052Get the global configurations of hash filters:: 2053 2054 get_hash_global_config (port_id) 2055 2056For example, to get the global configurations of hash filters of port 1:: 2057 2058 testpmd> get_hash_global_config 1 2059 2060set_hash_global_config 2061~~~~~~~~~~~~~~~~~~~~~~ 2062 2063Set the global configurations of hash filters:: 2064 2065 set_hash_global_config (port_id) (toeplitz|simple_xor|default) \ 2066 (ipv4|ipv4-frag|ipv4-tcp|ipv4-udp|ipv4-sctp|ipv4-other|ipv6|ipv6-frag| \ 2067 ipv6-tcp|ipv6-udp|ipv6-sctp|ipv6-other|l2_payload) \ 2068 (enable|disable) 2069 2070For example, to enable simple_xor for flow type of ipv6 on port 2:: 2071 2072 testpmd> set_hash_global_config 2 simple_xor ipv6 enable 2073 2074set_hash_input_set 2075~~~~~~~~~~~~~~~~~~ 2076 2077Set the input set for hash:: 2078 2079 set_hash_input_set (port_id) (ipv4-frag|ipv4-tcp|ipv4-udp|ipv4-sctp| \ 2080 ipv4-other|ipv6-frag|ipv6-tcp|ipv6-udp|ipv6-sctp|ipv6-other| \ 2081 l2_payload) (ovlan|ivlan|src-ipv4|dst-ipv4|src-ipv6|dst-ipv6|ipv4-tos| \ 2082 ipv4-proto|ipv6-tc|ipv6-next-header|udp-src-port|udp-dst-port| \ 2083 tcp-src-port|tcp-dst-port|sctp-src-port|sctp-dst-port|sctp-veri-tag| \ 2084 udp-key|gre-key|fld-1st|fld-2nd|fld-3rd|fld-4th|fld-5th|fld-6th|fld-7th| \ 2085 fld-8th|none) (select|add) 2086 2087For example, to add source IP to hash input set for flow type of ipv4-udp on port 0:: 2088 2089 testpmd> set_hash_input_set 0 ipv4-udp src-ipv4 add 2090 2091set_fdir_input_set 2092~~~~~~~~~~~~~~~~~~ 2093 2094The Flow Director filters can match the different fields for different type of packet, i.e. specific input set 2095on per flow type and the flexible payload. This command can be used to change input set for each flow type. 2096 2097Set the input set for flow director:: 2098 2099 set_fdir_input_set (port_id) (ipv4-frag|ipv4-tcp|ipv4-udp|ipv4-sctp| \ 2100 ipv4-other|ipv6|ipv6-frag|ipv6-tcp|ipv6-udp|ipv6-sctp|ipv6-other| \ 2101 l2_payload) (ivlan|ethertype|src-ipv4|dst-ipv4|src-ipv6|dst-ipv6|ipv4-tos| \ 2102 ipv4-proto|ipv4-ttl|ipv6-tc|ipv6-next-header|ipv6-hop-limits| \ 2103 tudp-src-port|udp-dst-port|cp-src-port|tcp-dst-port|sctp-src-port| \ 2104 sctp-dst-port|sctp-veri-tag|none) (select|add) 2105 2106For example to add source IP to FD input set for flow type of ipv4-udp on port 0:: 2107 2108 testpmd> set_fdir_input_set 0 ipv4-udp src-ipv4 add 2109 2110global_config 2111~~~~~~~~~~~~~ 2112 2113Set different GRE key length for input set:: 2114 2115 global_config (port_id) gre-key-len (number in bytes) 2116 2117For example to set GRE key length for input set to 4 bytes on port 0:: 2118 2119 testpmd> global_config 0 gre-key-len 4 2120 2121 2122.. _testpmd_rte_flow: 2123 2124Flow rules management 2125--------------------- 2126 2127Control of the generic flow API (*rte_flow*) is fully exposed through the 2128``flow`` command (validation, creation, destruction and queries). 2129 2130Considering *rte_flow* overlaps with all `Filter Functions`_, using both 2131features simultaneously may cause undefined side-effects and is therefore 2132not recommended. 2133 2134``flow`` syntax 2135~~~~~~~~~~~~~~~ 2136 2137Because the ``flow`` command uses dynamic tokens to handle the large number 2138of possible flow rules combinations, its behavior differs slightly from 2139other commands, in particular: 2140 2141- Pressing *?* or the *<tab>* key displays contextual help for the current 2142 token, not that of the entire command. 2143 2144- Optional and repeated parameters are supported (provided they are listed 2145 in the contextual help). 2146 2147The first parameter stands for the operation mode. Possible operations and 2148their general syntax are described below. They are covered in detail in the 2149following sections. 2150 2151- Check whether a flow rule can be created:: 2152 2153 flow validate {port_id} 2154 [group {group_id}] [priority {level}] [ingress] [egress] 2155 pattern {item} [/ {item} [...]] / end 2156 actions {action} [/ {action} [...]] / end 2157 2158- Create a flow rule:: 2159 2160 flow create {port_id} 2161 [group {group_id}] [priority {level}] [ingress] [egress] 2162 pattern {item} [/ {item} [...]] / end 2163 actions {action} [/ {action} [...]] / end 2164 2165- Destroy specific flow rules:: 2166 2167 flow destroy {port_id} rule {rule_id} [...] 2168 2169- Destroy all flow rules:: 2170 2171 flow flush {port_id} 2172 2173- Query an existing flow rule:: 2174 2175 flow query {port_id} {rule_id} {action} 2176 2177- List existing flow rules sorted by priority, filtered by group 2178 identifiers:: 2179 2180 flow list {port_id} [group {group_id}] [...] 2181 2182Validating flow rules 2183~~~~~~~~~~~~~~~~~~~~~ 2184 2185``flow validate`` reports whether a flow rule would be accepted by the 2186underlying device in its current state but stops short of creating it. It is 2187bound to ``rte_flow_validate()``:: 2188 2189 flow validate {port_id} 2190 [group {group_id}] [priority {level}] [ingress] [egress] 2191 pattern {item} [/ {item} [...]] / end 2192 actions {action} [/ {action} [...]] / end 2193 2194If successful, it will show:: 2195 2196 Flow rule validated 2197 2198Otherwise it will show an error message of the form:: 2199 2200 Caught error type [...] ([...]): [...] 2201 2202This command uses the same parameters as ``flow create``, their format is 2203described in `Creating flow rules`_. 2204 2205Check whether redirecting any Ethernet packet received on port 0 to RX queue 2206index 6 is supported:: 2207 2208 testpmd> flow validate 0 ingress pattern eth / end 2209 actions queue index 6 / end 2210 Flow rule validated 2211 testpmd> 2212 2213Port 0 does not support TCPv6 rules:: 2214 2215 testpmd> flow validate 0 ingress pattern eth / ipv6 / tcp / end 2216 actions drop / end 2217 Caught error type 9 (specific pattern item): Invalid argument 2218 testpmd> 2219 2220Creating flow rules 2221~~~~~~~~~~~~~~~~~~~ 2222 2223``flow create`` validates and creates the specified flow rule. It is bound 2224to ``rte_flow_create()``:: 2225 2226 flow create {port_id} 2227 [group {group_id}] [priority {level}] [ingress] [egress] 2228 pattern {item} [/ {item} [...]] / end 2229 actions {action} [/ {action} [...]] / end 2230 2231If successful, it will return a flow rule ID usable with other commands:: 2232 2233 Flow rule #[...] created 2234 2235Otherwise it will show an error message of the form:: 2236 2237 Caught error type [...] ([...]): [...] 2238 2239Parameters describe in the following order: 2240 2241- Attributes (*group*, *priority*, *ingress*, *egress* tokens). 2242- A matching pattern, starting with the *pattern* token and terminated by an 2243 *end* pattern item. 2244- Actions, starting with the *actions* token and terminated by an *end* 2245 action. 2246 2247These translate directly to *rte_flow* objects provided as-is to the 2248underlying functions. 2249 2250The shortest valid definition only comprises mandatory tokens:: 2251 2252 testpmd> flow create 0 pattern end actions end 2253 2254Note that PMDs may refuse rules that essentially do nothing such as this 2255one. 2256 2257**All unspecified object values are automatically initialized to 0.** 2258 2259Attributes 2260^^^^^^^^^^ 2261 2262These tokens affect flow rule attributes (``struct rte_flow_attr``) and are 2263specified before the ``pattern`` token. 2264 2265- ``group {group id}``: priority group. 2266- ``priority {level}``: priority level within group. 2267- ``ingress``: rule applies to ingress traffic. 2268- ``egress``: rule applies to egress traffic. 2269 2270Each instance of an attribute specified several times overrides the previous 2271value as shown below (group 4 is used):: 2272 2273 testpmd> flow create 0 group 42 group 24 group 4 [...] 2274 2275Note that once enabled, ``ingress`` and ``egress`` cannot be disabled. 2276 2277While not specifying a direction is an error, some rules may allow both 2278simultaneously. 2279 2280Most rules affect RX therefore contain the ``ingress`` token:: 2281 2282 testpmd> flow create 0 ingress pattern [...] 2283 2284Matching pattern 2285^^^^^^^^^^^^^^^^ 2286 2287A matching pattern starts after the ``pattern`` token. It is made of pattern 2288items and is terminated by a mandatory ``end`` item. 2289 2290Items are named after their type (*RTE_FLOW_ITEM_TYPE_* from ``enum 2291rte_flow_item_type``). 2292 2293The ``/`` token is used as a separator between pattern items as shown 2294below:: 2295 2296 testpmd> flow create 0 ingress pattern eth / ipv4 / udp / end [...] 2297 2298Note that protocol items like these must be stacked from lowest to highest 2299layer to make sense. For instance, the following rule is either invalid or 2300unlikely to match any packet:: 2301 2302 testpmd> flow create 0 ingress pattern eth / udp / ipv4 / end [...] 2303 2304More information on these restrictions can be found in the *rte_flow* 2305documentation. 2306 2307Several items support additional specification structures, for example 2308``ipv4`` allows specifying source and destination addresses as follows:: 2309 2310 testpmd> flow create 0 ingress pattern eth / ipv4 src is 10.1.1.1 2311 dst is 10.2.0.0 / end [...] 2312 2313This rule matches all IPv4 traffic with the specified properties. 2314 2315In this example, ``src`` and ``dst`` are field names of the underlying 2316``struct rte_flow_item_ipv4`` object. All item properties can be specified 2317in a similar fashion. 2318 2319The ``is`` token means that the subsequent value must be matched exactly, 2320and assigns ``spec`` and ``mask`` fields in ``struct rte_flow_item`` 2321accordingly. Possible assignment tokens are: 2322 2323- ``is``: match value perfectly (with full bit-mask). 2324- ``spec``: match value according to configured bit-mask. 2325- ``last``: specify upper bound to establish a range. 2326- ``mask``: specify bit-mask with relevant bits set to one. 2327- ``prefix``: generate bit-mask from a prefix length. 2328 2329These yield identical results:: 2330 2331 ipv4 src is 10.1.1.1 2332 2333:: 2334 2335 ipv4 src spec 10.1.1.1 src mask 255.255.255.255 2336 2337:: 2338 2339 ipv4 src spec 10.1.1.1 src prefix 32 2340 2341:: 2342 2343 ipv4 src is 10.1.1.1 src last 10.1.1.1 # range with a single value 2344 2345:: 2346 2347 ipv4 src is 10.1.1.1 src last 0 # 0 disables range 2348 2349Inclusive ranges can be defined with ``last``:: 2350 2351 ipv4 src is 10.1.1.1 src last 10.2.3.4 # 10.1.1.1 to 10.2.3.4 2352 2353Note that ``mask`` affects both ``spec`` and ``last``:: 2354 2355 ipv4 src is 10.1.1.1 src last 10.2.3.4 src mask 255.255.0.0 2356 # matches 10.1.0.0 to 10.2.255.255 2357 2358Properties can be modified multiple times:: 2359 2360 ipv4 src is 10.1.1.1 src is 10.1.2.3 src is 10.2.3.4 # matches 10.2.3.4 2361 2362:: 2363 2364 ipv4 src is 10.1.1.1 src prefix 24 src prefix 16 # matches 10.1.0.0/16 2365 2366Pattern items 2367^^^^^^^^^^^^^ 2368 2369This section lists supported pattern items and their attributes, if any. 2370 2371- ``end``: end list of pattern items. 2372 2373- ``void``: no-op pattern item. 2374 2375- ``invert``: perform actions when pattern does not match. 2376 2377- ``any``: match any protocol for the current layer. 2378 2379 - ``num {unsigned}``: number of layers covered. 2380 2381- ``pf``: match packets addressed to the physical function. 2382 2383- ``vf``: match packets addressed to a virtual function ID. 2384 2385 - ``id {unsigned}``: destination VF ID. 2386 2387- ``port``: device-specific physical port index to use. 2388 2389 - ``index {unsigned}``: physical port index. 2390 2391- ``raw``: match an arbitrary byte string. 2392 2393 - ``relative {boolean}``: look for pattern after the previous item. 2394 - ``search {boolean}``: search pattern from offset (see also limit). 2395 - ``offset {integer}``: absolute or relative offset for pattern. 2396 - ``limit {unsigned}``: search area limit for start of pattern. 2397 - ``pattern {string}``: byte string to look for. 2398 2399- ``eth``: match Ethernet header. 2400 2401 - ``dst {MAC-48}``: destination MAC. 2402 - ``src {MAC-48}``: source MAC. 2403 - ``type {unsigned}``: EtherType. 2404 2405- ``vlan``: match 802.1Q/ad VLAN tag. 2406 2407 - ``tpid {unsigned}``: tag protocol identifier. 2408 - ``tci {unsigned}``: tag control information. 2409 - ``pcp {unsigned}``: priority code point. 2410 - ``dei {unsigned}``: drop eligible indicator. 2411 - ``vid {unsigned}``: VLAN identifier. 2412 2413- ``ipv4``: match IPv4 header. 2414 2415 - ``tos {unsigned}``: type of service. 2416 - ``ttl {unsigned}``: time to live. 2417 - ``proto {unsigned}``: next protocol ID. 2418 - ``src {ipv4 address}``: source address. 2419 - ``dst {ipv4 address}``: destination address. 2420 2421- ``ipv6``: match IPv6 header. 2422 2423 - ``tc {unsigned}``: traffic class. 2424 - ``flow {unsigned}``: flow label. 2425 - ``proto {unsigned}``: protocol (next header). 2426 - ``hop {unsigned}``: hop limit. 2427 - ``src {ipv6 address}``: source address. 2428 - ``dst {ipv6 address}``: destination address. 2429 2430- ``icmp``: match ICMP header. 2431 2432 - ``type {unsigned}``: ICMP packet type. 2433 - ``code {unsigned}``: ICMP packet code. 2434 2435- ``udp``: match UDP header. 2436 2437 - ``src {unsigned}``: UDP source port. 2438 - ``dst {unsigned}``: UDP destination port. 2439 2440- ``tcp``: match TCP header. 2441 2442 - ``src {unsigned}``: TCP source port. 2443 - ``dst {unsigned}``: TCP destination port. 2444 2445- ``sctp``: match SCTP header. 2446 2447 - ``src {unsigned}``: SCTP source port. 2448 - ``dst {unsigned}``: SCTP destination port. 2449 - ``tag {unsigned}``: validation tag. 2450 - ``cksum {unsigned}``: checksum. 2451 2452- ``vxlan``: match VXLAN header. 2453 2454 - ``vni {unsigned}``: VXLAN identifier. 2455 2456Actions list 2457^^^^^^^^^^^^ 2458 2459A list of actions starts after the ``actions`` token in the same fashion as 2460`Matching pattern`_; actions are separated by ``/`` tokens and the list is 2461terminated by a mandatory ``end`` action. 2462 2463Actions are named after their type (*RTE_FLOW_ACTION_TYPE_* from ``enum 2464rte_flow_action_type``). 2465 2466Dropping all incoming UDPv4 packets can be expressed as follows:: 2467 2468 testpmd> flow create 0 ingress pattern eth / ipv4 / udp / end 2469 actions drop / end 2470 2471Several actions have configurable properties which must be specified when 2472there is no valid default value. For example, ``queue`` requires a target 2473queue index. 2474 2475This rule redirects incoming UDPv4 traffic to queue index 6:: 2476 2477 testpmd> flow create 0 ingress pattern eth / ipv4 / udp / end 2478 actions queue index 6 / end 2479 2480While this one could be rejected by PMDs (unspecified queue index):: 2481 2482 testpmd> flow create 0 ingress pattern eth / ipv4 / udp / end 2483 actions queue / end 2484 2485As defined by *rte_flow*, the list is not ordered, all actions of a given 2486rule are performed simultaneously. These are equivalent:: 2487 2488 queue index 6 / void / mark id 42 / end 2489 2490:: 2491 2492 void / mark id 42 / queue index 6 / end 2493 2494All actions in a list should have different types, otherwise only the last 2495action of a given type is taken into account:: 2496 2497 queue index 4 / queue index 5 / queue index 6 / end # will use queue 6 2498 2499:: 2500 2501 drop / drop / drop / end # drop is performed only once 2502 2503:: 2504 2505 mark id 42 / queue index 3 / mark id 24 / end # mark will be 24 2506 2507Considering they are performed simultaneously, opposite and overlapping 2508actions can sometimes be combined when the end result is unambiguous:: 2509 2510 drop / queue index 6 / end # drop has no effect 2511 2512:: 2513 2514 drop / dup index 6 / end # same as above 2515 2516:: 2517 2518 queue index 6 / rss queues 6 7 8 / end # queue has no effect 2519 2520:: 2521 2522 drop / passthru / end # drop has no effect 2523 2524Note that PMDs may still refuse such combinations. 2525 2526Actions 2527^^^^^^^ 2528 2529This section lists supported actions and their attributes, if any. 2530 2531- ``end``: end list of actions. 2532 2533- ``void``: no-op action. 2534 2535- ``passthru``: let subsequent rule process matched packets. 2536 2537- ``mark``: attach 32 bit value to packets. 2538 2539 - ``id {unsigned}``: 32 bit value to return with packets. 2540 2541- ``flag``: flag packets. 2542 2543- ``queue``: assign packets to a given queue index. 2544 2545 - ``index {unsigned}``: queue index to use. 2546 2547- ``drop``: drop packets (note: passthru has priority). 2548 2549- ``count``: enable counters for this rule. 2550 2551- ``dup``: duplicate packets to a given queue index. 2552 2553 - ``index {unsigned}``: queue index to duplicate packets to. 2554 2555- ``rss``: spread packets among several queues. 2556 2557 - ``queues [{unsigned} [...]] end``: queue indices to use. 2558 2559- ``pf``: redirect packets to physical device function. 2560 2561- ``vf``: redirect packets to virtual device function. 2562 2563 - ``original {boolean}``: use original VF ID if possible. 2564 - ``id {unsigned}``: VF ID to redirect packets to. 2565 2566Destroying flow rules 2567~~~~~~~~~~~~~~~~~~~~~ 2568 2569``flow destroy`` destroys one or more rules from their rule ID (as returned 2570by ``flow create``), this command calls ``rte_flow_destroy()`` as many 2571times as necessary:: 2572 2573 flow destroy {port_id} rule {rule_id} [...] 2574 2575If successful, it will show:: 2576 2577 Flow rule #[...] destroyed 2578 2579It does not report anything for rule IDs that do not exist. The usual error 2580message is shown when a rule cannot be destroyed:: 2581 2582 Caught error type [...] ([...]): [...] 2583 2584``flow flush`` destroys all rules on a device and does not take extra 2585arguments. It is bound to ``rte_flow_flush()``:: 2586 2587 flow flush {port_id} 2588 2589Any errors are reported as above. 2590 2591Creating several rules and destroying them:: 2592 2593 testpmd> flow create 0 ingress pattern eth / ipv6 / end 2594 actions queue index 2 / end 2595 Flow rule #0 created 2596 testpmd> flow create 0 ingress pattern eth / ipv4 / end 2597 actions queue index 3 / end 2598 Flow rule #1 created 2599 testpmd> flow destroy 0 rule 0 rule 1 2600 Flow rule #1 destroyed 2601 Flow rule #0 destroyed 2602 testpmd> 2603 2604The same result can be achieved using ``flow flush``:: 2605 2606 testpmd> flow create 0 ingress pattern eth / ipv6 / end 2607 actions queue index 2 / end 2608 Flow rule #0 created 2609 testpmd> flow create 0 ingress pattern eth / ipv4 / end 2610 actions queue index 3 / end 2611 Flow rule #1 created 2612 testpmd> flow flush 0 2613 testpmd> 2614 2615Non-existent rule IDs are ignored:: 2616 2617 testpmd> flow create 0 ingress pattern eth / ipv6 / end 2618 actions queue index 2 / end 2619 Flow rule #0 created 2620 testpmd> flow create 0 ingress pattern eth / ipv4 / end 2621 actions queue index 3 / end 2622 Flow rule #1 created 2623 testpmd> flow destroy 0 rule 42 rule 10 rule 2 2624 testpmd> 2625 testpmd> flow destroy 0 rule 0 2626 Flow rule #0 destroyed 2627 testpmd> 2628 2629Querying flow rules 2630~~~~~~~~~~~~~~~~~~~ 2631 2632``flow query`` queries a specific action of a flow rule having that 2633ability. Such actions collect information that can be reported using this 2634command. It is bound to ``rte_flow_query()``:: 2635 2636 flow query {port_id} {rule_id} {action} 2637 2638If successful, it will display either the retrieved data for known actions 2639or the following message:: 2640 2641 Cannot display result for action type [...] ([...]) 2642 2643Otherwise, it will complain either that the rule does not exist or that some 2644error occurred:: 2645 2646 Flow rule #[...] not found 2647 2648:: 2649 2650 Caught error type [...] ([...]): [...] 2651 2652Currently only the ``count`` action is supported. This action reports the 2653number of packets that hit the flow rule and the total number of bytes. Its 2654output has the following format:: 2655 2656 count: 2657 hits_set: [...] # whether "hits" contains a valid value 2658 bytes_set: [...] # whether "bytes" contains a valid value 2659 hits: [...] # number of packets 2660 bytes: [...] # number of bytes 2661 2662Querying counters for TCPv6 packets redirected to queue 6:: 2663 2664 testpmd> flow create 0 ingress pattern eth / ipv6 / tcp / end 2665 actions queue index 6 / count / end 2666 Flow rule #4 created 2667 testpmd> flow query 0 4 count 2668 count: 2669 hits_set: 1 2670 bytes_set: 0 2671 hits: 386446 2672 bytes: 0 2673 testpmd> 2674 2675Listing flow rules 2676~~~~~~~~~~~~~~~~~~ 2677 2678``flow list`` lists existing flow rules sorted by priority and optionally 2679filtered by group identifiers:: 2680 2681 flow list {port_id} [group {group_id}] [...] 2682 2683This command only fails with the following message if the device does not 2684exist:: 2685 2686 Invalid port [...] 2687 2688Output consists of a header line followed by a short description of each 2689flow rule, one per line. There is no output at all when no flow rules are 2690configured on the device:: 2691 2692 ID Group Prio Attr Rule 2693 [...] [...] [...] [...] [...] 2694 2695``Attr`` column flags: 2696 2697- ``i`` for ``ingress``. 2698- ``e`` for ``egress``. 2699 2700Creating several flow rules and listing them:: 2701 2702 testpmd> flow create 0 ingress pattern eth / ipv4 / end 2703 actions queue index 6 / end 2704 Flow rule #0 created 2705 testpmd> flow create 0 ingress pattern eth / ipv6 / end 2706 actions queue index 2 / end 2707 Flow rule #1 created 2708 testpmd> flow create 0 priority 5 ingress pattern eth / ipv4 / udp / end 2709 actions rss queues 6 7 8 end / end 2710 Flow rule #2 created 2711 testpmd> flow list 0 2712 ID Group Prio Attr Rule 2713 0 0 0 i- ETH IPV4 => QUEUE 2714 1 0 0 i- ETH IPV6 => QUEUE 2715 2 0 5 i- ETH IPV4 UDP => RSS 2716 testpmd> 2717 2718Rules are sorted by priority (i.e. group ID first, then priority level):: 2719 2720 testpmd> flow list 1 2721 ID Group Prio Attr Rule 2722 0 0 0 i- ETH => COUNT 2723 6 0 500 i- ETH IPV6 TCP => DROP COUNT 2724 5 0 1000 i- ETH IPV6 ICMP => QUEUE 2725 1 24 0 i- ETH IPV4 UDP => QUEUE 2726 4 24 10 i- ETH IPV4 TCP => DROP 2727 3 24 20 i- ETH IPV4 => DROP 2728 2 24 42 i- ETH IPV4 UDP => QUEUE 2729 7 63 0 i- ETH IPV6 UDP VXLAN => MARK QUEUE 2730 testpmd> 2731 2732Output can be limited to specific groups:: 2733 2734 testpmd> flow list 1 group 0 group 63 2735 ID Group Prio Attr Rule 2736 0 0 0 i- ETH => COUNT 2737 6 0 500 i- ETH IPV6 TCP => DROP COUNT 2738 5 0 1000 i- ETH IPV6 ICMP => QUEUE 2739 7 63 0 i- ETH IPV6 UDP VXLAN => MARK QUEUE 2740 testpmd> 2741