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 tx max bandwidth (for VF) 897~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ 898 899Set TX max absolute bandwidth (Mbps) for a VF from PF:: 900 901 testpmd> set vf tx max-bandwidth (port_id) (vf_id) (max_bandwidth) 902 903set tc tx min bandwidth (for VF) 904~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ 905 906Set all TCs' TX min relative bandwidth (%) for a VF from PF:: 907 908 testpmd> set vf tc tx min-bandwidth (port_id) (vf_id) (bw1, bw2, ...) 909 910set tc tx max bandwidth (for VF) 911~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ 912 913Set a TC's TX max absolute bandwidth (Mbps) for a VF from PF:: 914 915 testpmd> set vf tc tx max-bandwidth (port_id) (vf_id) (tc_no) (max_bandwidth) 916 917set tc strict link priority mode 918~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ 919 920Set some TCs' strict link priority mode on a physical port:: 921 922 testpmd> set tx strict-link-priority (port_id) (tc_bitmap) 923 924set flow_ctrl rx 925~~~~~~~~~~~~~~~~ 926 927Set the link flow control parameter on a port:: 928 929 testpmd> set flow_ctrl rx (on|off) tx (on|off) (high_water) (low_water) \ 930 (pause_time) (send_xon) mac_ctrl_frame_fwd (on|off) \ 931 autoneg (on|off) (port_id) 932 933Where: 934 935* ``high_water`` (integer): High threshold value to trigger XOFF. 936 937* ``low_water`` (integer): Low threshold value to trigger XON. 938 939* ``pause_time`` (integer): Pause quota in the Pause frame. 940 941* ``send_xon`` (0/1): Send XON frame. 942 943* ``mac_ctrl_frame_fwd``: Enable receiving MAC control frames. 944 945* ``autoneg``: Change the auto-negotiation parameter. 946 947set pfc_ctrl rx 948~~~~~~~~~~~~~~~ 949 950Set the priority flow control parameter on a port:: 951 952 testpmd> set pfc_ctrl rx (on|off) tx (on|off) (high_water) (low_water) \ 953 (pause_time) (priority) (port_id) 954 955Where: 956 957* ``high_water`` (integer): High threshold value. 958 959* ``low_water`` (integer): Low threshold value. 960 961* ``pause_time`` (integer): Pause quota in the Pause frame. 962 963* ``priority`` (0-7): VLAN User Priority. 964 965set stat_qmap 966~~~~~~~~~~~~~ 967 968Set statistics mapping (qmapping 0..15) for RX/TX queue on port:: 969 970 testpmd> set stat_qmap (tx|rx) (port_id) (queue_id) (qmapping) 971 972For example, to set rx queue 2 on port 0 to mapping 5:: 973 974 testpmd>set stat_qmap rx 0 2 5 975 976set port - rx/tx (for VF) 977~~~~~~~~~~~~~~~~~~~~~~~~~ 978 979Set VF receive/transmit from a port:: 980 981 testpmd> set port (port_id) vf (vf_id) (rx|tx) (on|off) 982 983set port - mac address filter (for VF) 984~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ 985 986Add/Remove unicast or multicast MAC addr filter for a VF:: 987 988 testpmd> set port (port_id) vf (vf_id) (mac_addr) \ 989 (exact-mac|exact-mac-vlan|hashmac|hashmac-vlan) (on|off) 990 991set port - rx mode(for VF) 992~~~~~~~~~~~~~~~~~~~~~~~~~~ 993 994Set the VF receive mode of a port:: 995 996 testpmd> set port (port_id) vf (vf_id) \ 997 rxmode (AUPE|ROPE|BAM|MPE) (on|off) 998 999The available receive modes are: 1000 1001* ``AUPE``: Accepts untagged VLAN. 1002 1003* ``ROPE``: Accepts unicast hash. 1004 1005* ``BAM``: Accepts broadcast packets. 1006 1007* ``MPE``: Accepts all multicast packets. 1008 1009set port - tx_rate (for Queue) 1010~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ 1011 1012Set TX rate limitation for a queue on a port:: 1013 1014 testpmd> set port (port_id) queue (queue_id) rate (rate_value) 1015 1016set port - tx_rate (for VF) 1017~~~~~~~~~~~~~~~~~~~~~~~~~~~ 1018 1019Set TX rate limitation for queues in VF on a port:: 1020 1021 testpmd> set port (port_id) vf (vf_id) rate (rate_value) queue_mask (queue_mask) 1022 1023set port - mirror rule 1024~~~~~~~~~~~~~~~~~~~~~~ 1025 1026Set pool or vlan type mirror rule for a port:: 1027 1028 testpmd> set port (port_id) mirror-rule (rule_id) \ 1029 (pool-mirror-up|pool-mirror-down|vlan-mirror) \ 1030 (poolmask|vlanid[,vlanid]*) dst-pool (pool_id) (on|off) 1031 1032Set link mirror rule for a port:: 1033 1034 testpmd> set port (port_id) mirror-rule (rule_id) \ 1035 (uplink-mirror|downlink-mirror) dst-pool (pool_id) (on|off) 1036 1037For example to enable mirror traffic with vlan 0,1 to pool 0:: 1038 1039 set port 0 mirror-rule 0 vlan-mirror 0,1 dst-pool 0 on 1040 1041reset port - mirror rule 1042~~~~~~~~~~~~~~~~~~~~~~~~ 1043 1044Reset a mirror rule for a port:: 1045 1046 testpmd> reset port (port_id) mirror-rule (rule_id) 1047 1048set flush_rx 1049~~~~~~~~~~~~ 1050 1051Set the flush on RX streams before forwarding. 1052The default is flush ``on``. 1053Mainly used with PCAP drivers to turn off the default behavior of flushing the first 512 packets on RX streams:: 1054 1055 testpmd> set flush_rx off 1056 1057set bypass mode 1058~~~~~~~~~~~~~~~ 1059 1060Set the bypass mode for the lowest port on bypass enabled NIC:: 1061 1062 testpmd> set bypass mode (normal|bypass|isolate) (port_id) 1063 1064set bypass event 1065~~~~~~~~~~~~~~~~ 1066 1067Set the event required to initiate specified bypass mode for the lowest port on a bypass enabled:: 1068 1069 testpmd> set bypass event (timeout|os_on|os_off|power_on|power_off) \ 1070 mode (normal|bypass|isolate) (port_id) 1071 1072Where: 1073 1074* ``timeout``: Enable bypass after watchdog timeout. 1075 1076* ``os_on``: Enable bypass when OS/board is powered on. 1077 1078* ``os_off``: Enable bypass when OS/board is powered off. 1079 1080* ``power_on``: Enable bypass when power supply is turned on. 1081 1082* ``power_off``: Enable bypass when power supply is turned off. 1083 1084 1085set bypass timeout 1086~~~~~~~~~~~~~~~~~~ 1087 1088Set the bypass watchdog timeout to ``n`` seconds where 0 = instant:: 1089 1090 testpmd> set bypass timeout (0|1.5|2|3|4|8|16|32) 1091 1092show bypass config 1093~~~~~~~~~~~~~~~~~~ 1094 1095Show the bypass configuration for a bypass enabled NIC using the lowest port on the NIC:: 1096 1097 testpmd> show bypass config (port_id) 1098 1099set link up 1100~~~~~~~~~~~ 1101 1102Set link up for a port:: 1103 1104 testpmd> set link-up port (port id) 1105 1106set link down 1107~~~~~~~~~~~~~ 1108 1109Set link down for a port:: 1110 1111 testpmd> set link-down port (port id) 1112 1113E-tag set 1114~~~~~~~~~ 1115 1116Enable E-tag insertion for a VF on a port:: 1117 1118 testpmd> E-tag set insertion on port-tag-id (value) port (port_id) vf (vf_id) 1119 1120Disable E-tag insertion for a VF on a port:: 1121 1122 testpmd> E-tag set insertion off port (port_id) vf (vf_id) 1123 1124Enable/disable E-tag stripping on a port:: 1125 1126 testpmd> E-tag set stripping (on|off) port (port_id) 1127 1128Enable/disable E-tag based forwarding on a port:: 1129 1130 testpmd> E-tag set forwarding (on|off) port (port_id) 1131 1132Add an E-tag forwarding filter on a port:: 1133 1134 testpmd> E-tag set filter add e-tag-id (value) dst-pool (pool_id) port (port_id) 1135 1136Delete an E-tag forwarding filter on a port:: 1137 testpmd> E-tag set filter del e-tag-id (value) port (port_id) 1138 1139 1140Port Functions 1141-------------- 1142 1143The following sections show functions for configuring ports. 1144 1145.. note:: 1146 1147 Port configuration changes only become active when forwarding is started/restarted. 1148 1149port attach 1150~~~~~~~~~~~ 1151 1152Attach a port specified by pci address or virtual device args:: 1153 1154 testpmd> port attach (identifier) 1155 1156To attach a new pci device, the device should be recognized by kernel first. 1157Then it should be moved under DPDK management. 1158Finally the port can be attached to testpmd. 1159 1160For example, to move a pci device using ixgbe under DPDK management: 1161 1162.. code-block:: console 1163 1164 # Check the status of the available devices. 1165 ./usertools/dpdk-devbind.py --status 1166 1167 Network devices using DPDK-compatible driver 1168 ============================================ 1169 <none> 1170 1171 Network devices using kernel driver 1172 =================================== 1173 0000:0a:00.0 '82599ES 10-Gigabit' if=eth2 drv=ixgbe unused= 1174 1175 1176 # Bind the device to igb_uio. 1177 sudo ./usertools/dpdk-devbind.py -b igb_uio 0000:0a:00.0 1178 1179 1180 # Recheck the status of the devices. 1181 ./usertools/dpdk-devbind.py --status 1182 Network devices using DPDK-compatible driver 1183 ============================================ 1184 0000:0a:00.0 '82599ES 10-Gigabit' drv=igb_uio unused= 1185 1186To attach a port created by virtual device, above steps are not needed. 1187 1188For example, to attach a port whose pci address is 0000:0a:00.0. 1189 1190.. code-block:: console 1191 1192 testpmd> port attach 0000:0a:00.0 1193 Attaching a new port... 1194 EAL: PCI device 0000:0a:00.0 on NUMA socket -1 1195 EAL: probe driver: 8086:10fb rte_ixgbe_pmd 1196 EAL: PCI memory mapped at 0x7f83bfa00000 1197 EAL: PCI memory mapped at 0x7f83bfa80000 1198 PMD: eth_ixgbe_dev_init(): MAC: 2, PHY: 18, SFP+: 5 1199 PMD: eth_ixgbe_dev_init(): port 0 vendorID=0x8086 deviceID=0x10fb 1200 Port 0 is attached. Now total ports is 1 1201 Done 1202 1203For example, to attach a port created by pcap PMD. 1204 1205.. code-block:: console 1206 1207 testpmd> port attach net_pcap0 1208 Attaching a new port... 1209 PMD: Initializing pmd_pcap for net_pcap0 1210 PMD: Creating pcap-backed ethdev on numa socket 0 1211 Port 0 is attached. Now total ports is 1 1212 Done 1213 1214In this case, identifier is ``net_pcap0``. 1215This identifier format is the same as ``--vdev`` format of DPDK applications. 1216 1217For example, to re-attach a bonded port which has been previously detached, 1218the mode and slave parameters must be given. 1219 1220.. code-block:: console 1221 1222 testpmd> port attach net_bond_0,mode=0,slave=1 1223 Attaching a new port... 1224 EAL: Initializing pmd_bond for net_bond_0 1225 EAL: Create bonded device net_bond_0 on port 0 in mode 0 on socket 0. 1226 Port 0 is attached. Now total ports is 1 1227 Done 1228 1229 1230port detach 1231~~~~~~~~~~~ 1232 1233Detach a specific port:: 1234 1235 testpmd> port detach (port_id) 1236 1237Before detaching a port, the port should be stopped and closed. 1238 1239For example, to detach a pci device port 0. 1240 1241.. code-block:: console 1242 1243 testpmd> port stop 0 1244 Stopping ports... 1245 Done 1246 testpmd> port close 0 1247 Closing ports... 1248 Done 1249 1250 testpmd> port detach 0 1251 Detaching a port... 1252 EAL: PCI device 0000:0a:00.0 on NUMA socket -1 1253 EAL: remove driver: 8086:10fb rte_ixgbe_pmd 1254 EAL: PCI memory unmapped at 0x7f83bfa00000 1255 EAL: PCI memory unmapped at 0x7f83bfa80000 1256 Done 1257 1258 1259For example, to detach a virtual device port 0. 1260 1261.. code-block:: console 1262 1263 testpmd> port stop 0 1264 Stopping ports... 1265 Done 1266 testpmd> port close 0 1267 Closing ports... 1268 Done 1269 1270 testpmd> port detach 0 1271 Detaching a port... 1272 PMD: Closing pcap ethdev on numa socket 0 1273 Port 'net_pcap0' is detached. Now total ports is 0 1274 Done 1275 1276To remove a pci device completely from the system, first detach the port from testpmd. 1277Then the device should be moved under kernel management. 1278Finally the device can be removed using kernel pci hotplug functionality. 1279 1280For example, to move a pci device under kernel management: 1281 1282.. code-block:: console 1283 1284 sudo ./usertools/dpdk-devbind.py -b ixgbe 0000:0a:00.0 1285 1286 ./usertools/dpdk-devbind.py --status 1287 1288 Network devices using DPDK-compatible driver 1289 ============================================ 1290 <none> 1291 1292 Network devices using kernel driver 1293 =================================== 1294 0000:0a:00.0 '82599ES 10-Gigabit' if=eth2 drv=ixgbe unused=igb_uio 1295 1296To remove a port created by a virtual device, above steps are not needed. 1297 1298port start 1299~~~~~~~~~~ 1300 1301Start all ports or a specific port:: 1302 1303 testpmd> port start (port_id|all) 1304 1305port stop 1306~~~~~~~~~ 1307 1308Stop all ports or a specific port:: 1309 1310 testpmd> port stop (port_id|all) 1311 1312port close 1313~~~~~~~~~~ 1314 1315Close all ports or a specific port:: 1316 1317 testpmd> port close (port_id|all) 1318 1319port start/stop queue 1320~~~~~~~~~~~~~~~~~~~~~ 1321 1322Start/stop a rx/tx queue on a specific port:: 1323 1324 testpmd> port (port_id) (rxq|txq) (queue_id) (start|stop) 1325 1326Only take effect when port is started. 1327 1328port config - speed 1329~~~~~~~~~~~~~~~~~~~ 1330 1331Set the speed and duplex mode for all ports or a specific port:: 1332 1333 testpmd> port config (port_id|all) speed (10|100|1000|10000|25000|40000|50000|100000|auto) \ 1334 duplex (half|full|auto) 1335 1336port config - queues/descriptors 1337~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ 1338 1339Set number of queues/descriptors for rxq, txq, rxd and txd:: 1340 1341 testpmd> port config all (rxq|txq|rxd|txd) (value) 1342 1343This is equivalent to the ``--rxq``, ``--txq``, ``--rxd`` and ``--txd`` command-line options. 1344 1345port config - max-pkt-len 1346~~~~~~~~~~~~~~~~~~~~~~~~~ 1347 1348Set the maximum packet length:: 1349 1350 testpmd> port config all max-pkt-len (value) 1351 1352This is equivalent to the ``--max-pkt-len`` command-line option. 1353 1354port config - CRC Strip 1355~~~~~~~~~~~~~~~~~~~~~~~ 1356 1357Set hardware CRC stripping on or off for all ports:: 1358 1359 testpmd> port config all crc-strip (on|off) 1360 1361CRC stripping is off by default. 1362 1363The ``on`` option is equivalent to the ``--crc-strip`` command-line option. 1364 1365port config - scatter 1366~~~~~~~~~~~~~~~~~~~~~~~ 1367 1368Set RX scatter mode on or off for all ports:: 1369 1370 testpmd> port config all scatter (on|off) 1371 1372RX scatter mode is off by default. 1373 1374The ``on`` option is equivalent to the ``--enable-scatter`` command-line option. 1375 1376port config - TX queue flags 1377~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ 1378 1379Set a hexadecimal bitmap of TX queue flags for all ports:: 1380 1381 testpmd> port config all txqflags value 1382 1383This command is equivalent to the ``--txqflags`` command-line option. 1384 1385port config - RX Checksum 1386~~~~~~~~~~~~~~~~~~~~~~~~~ 1387 1388Set hardware RX checksum offload to on or off for all ports:: 1389 1390 testpmd> port config all rx-cksum (on|off) 1391 1392Checksum offload is off by default. 1393 1394The ``on`` option is equivalent to the ``--enable-rx-cksum`` command-line option. 1395 1396port config - VLAN 1397~~~~~~~~~~~~~~~~~~ 1398 1399Set hardware VLAN on or off for all ports:: 1400 1401 testpmd> port config all hw-vlan (on|off) 1402 1403Hardware VLAN is on by default. 1404 1405The ``off`` option is equivalent to the ``--disable-hw-vlan`` command-line option. 1406 1407port config - VLAN filter 1408~~~~~~~~~~~~~~~~~~~~~~~~~ 1409 1410Set hardware VLAN filter on or off for all ports:: 1411 1412 testpmd> port config all hw-vlan-filter (on|off) 1413 1414Hardware VLAN filter is on by default. 1415 1416The ``off`` option is equivalent to the ``--disable-hw-vlan-filter`` command-line option. 1417 1418port config - VLAN strip 1419~~~~~~~~~~~~~~~~~~~~~~~~ 1420 1421Set hardware VLAN strip on or off for all ports:: 1422 1423 testpmd> port config all hw-vlan-strip (on|off) 1424 1425Hardware VLAN strip is on by default. 1426 1427The ``off`` option is equivalent to the ``--disable-hw-vlan-strip`` command-line option. 1428 1429port config - VLAN extend 1430~~~~~~~~~~~~~~~~~~~~~~~~~ 1431 1432Set hardware VLAN extend on or off for all ports:: 1433 1434 testpmd> port config all hw-vlan-extend (on|off) 1435 1436Hardware VLAN extend is off by default. 1437 1438The ``off`` option is equivalent to the ``--disable-hw-vlan-extend`` command-line option. 1439 1440port config - Drop Packets 1441~~~~~~~~~~~~~~~~~~~~~~~~~~ 1442 1443Set packet drop for packets with no descriptors on or off for all ports:: 1444 1445 testpmd> port config all drop-en (on|off) 1446 1447Packet dropping for packets with no descriptors is off by default. 1448 1449The ``on`` option is equivalent to the ``--enable-drop-en`` command-line option. 1450 1451port config - RSS 1452~~~~~~~~~~~~~~~~~ 1453 1454Set the RSS (Receive Side Scaling) mode on or off:: 1455 1456 testpmd> port config all rss (all|ip|tcp|udp|sctp|ether|port|vxlan|geneve|nvgre|none) 1457 1458RSS is on by default. 1459 1460The ``none`` option is equivalent to the ``--disable-rss`` command-line option. 1461 1462port config - RSS Reta 1463~~~~~~~~~~~~~~~~~~~~~~ 1464 1465Set the RSS (Receive Side Scaling) redirection table:: 1466 1467 testpmd> port config all rss reta (hash,queue)[,(hash,queue)] 1468 1469port config - DCB 1470~~~~~~~~~~~~~~~~~ 1471 1472Set the DCB mode for an individual port:: 1473 1474 testpmd> port config (port_id) dcb vt (on|off) (traffic_class) pfc (on|off) 1475 1476The traffic class should be 4 or 8. 1477 1478port config - Burst 1479~~~~~~~~~~~~~~~~~~~ 1480 1481Set the number of packets per burst:: 1482 1483 testpmd> port config all burst (value) 1484 1485This is equivalent to the ``--burst`` command-line option. 1486 1487port config - Threshold 1488~~~~~~~~~~~~~~~~~~~~~~~ 1489 1490Set thresholds for TX/RX queues:: 1491 1492 testpmd> port config all (threshold) (value) 1493 1494Where the threshold type can be: 1495 1496* ``txpt:`` Set the prefetch threshold register of the TX rings, 0 <= value <= 255. 1497 1498* ``txht:`` Set the host threshold register of the TX rings, 0 <= value <= 255. 1499 1500* ``txwt:`` Set the write-back threshold register of the TX rings, 0 <= value <= 255. 1501 1502* ``rxpt:`` Set the prefetch threshold register of the RX rings, 0 <= value <= 255. 1503 1504* ``rxht:`` Set the host threshold register of the RX rings, 0 <= value <= 255. 1505 1506* ``rxwt:`` Set the write-back threshold register of the RX rings, 0 <= value <= 255. 1507 1508* ``txfreet:`` Set the transmit free threshold of the TX rings, 0 <= value <= txd. 1509 1510* ``rxfreet:`` Set the transmit free threshold of the RX rings, 0 <= value <= rxd. 1511 1512* ``txrst:`` Set the transmit RS bit threshold of TX rings, 0 <= value <= txd. 1513 1514These threshold options are also available from the command-line. 1515 1516port config - E-tag 1517~~~~~~~~~~~~~~~~~~~ 1518 1519Set the value of ether-type for E-tag:: 1520 1521 testpmd> port config (port_id|all) l2-tunnel E-tag ether-type (value) 1522 1523Enable/disable the E-tag support:: 1524 1525 testpmd> port config (port_id|all) l2-tunnel E-tag (enable|disable) 1526 1527 1528Link Bonding Functions 1529---------------------- 1530 1531The Link Bonding functions make it possible to dynamically create and 1532manage link bonding devices from within testpmd interactive prompt. 1533 1534create bonded device 1535~~~~~~~~~~~~~~~~~~~~ 1536 1537Create a new bonding device:: 1538 1539 testpmd> create bonded device (mode) (socket) 1540 1541For example, to create a bonded device in mode 1 on socket 0:: 1542 1543 testpmd> create bonded 1 0 1544 created new bonded device (port X) 1545 1546add bonding slave 1547~~~~~~~~~~~~~~~~~ 1548 1549Adds Ethernet device to a Link Bonding device:: 1550 1551 testpmd> add bonding slave (slave id) (port id) 1552 1553For example, to add Ethernet device (port 6) to a Link Bonding device (port 10):: 1554 1555 testpmd> add bonding slave 6 10 1556 1557 1558remove bonding slave 1559~~~~~~~~~~~~~~~~~~~~ 1560 1561Removes an Ethernet slave device from a Link Bonding device:: 1562 1563 testpmd> remove bonding slave (slave id) (port id) 1564 1565For example, to remove Ethernet slave device (port 6) to a Link Bonding device (port 10):: 1566 1567 testpmd> remove bonding slave 6 10 1568 1569set bonding mode 1570~~~~~~~~~~~~~~~~ 1571 1572Set the Link Bonding mode of a Link Bonding device:: 1573 1574 testpmd> set bonding mode (value) (port id) 1575 1576For example, to set the bonding mode of a Link Bonding device (port 10) to broadcast (mode 3):: 1577 1578 testpmd> set bonding mode 3 10 1579 1580set bonding primary 1581~~~~~~~~~~~~~~~~~~~ 1582 1583Set an Ethernet slave device as the primary device on a Link Bonding device:: 1584 1585 testpmd> set bonding primary (slave id) (port id) 1586 1587For example, to set the Ethernet slave device (port 6) as the primary port of a Link Bonding device (port 10):: 1588 1589 testpmd> set bonding primary 6 10 1590 1591set bonding mac 1592~~~~~~~~~~~~~~~ 1593 1594Set the MAC address of a Link Bonding device:: 1595 1596 testpmd> set bonding mac (port id) (mac) 1597 1598For example, to set the MAC address of a Link Bonding device (port 10) to 00:00:00:00:00:01:: 1599 1600 testpmd> set bonding mac 10 00:00:00:00:00:01 1601 1602set bonding xmit_balance_policy 1603~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ 1604 1605Set the transmission policy for a Link Bonding device when it is in Balance XOR mode:: 1606 1607 testpmd> set bonding xmit_balance_policy (port_id) (l2|l23|l34) 1608 1609For example, set a Link Bonding device (port 10) to use a balance policy of layer 3+4 (IP addresses & UDP ports):: 1610 1611 testpmd> set bonding xmit_balance_policy 10 l34 1612 1613 1614set bonding mon_period 1615~~~~~~~~~~~~~~~~~~~~~~ 1616 1617Set the link status monitoring polling period in milliseconds for a bonding device. 1618 1619This adds support for PMD slave devices which do not support link status interrupts. 1620When the mon_period is set to a value greater than 0 then all PMD's which do not support 1621link status ISR will be queried every polling interval to check if their link status has changed:: 1622 1623 testpmd> set bonding mon_period (port_id) (value) 1624 1625For example, to set the link status monitoring polling period of bonded device (port 5) to 150ms:: 1626 1627 testpmd> set bonding mon_period 5 150 1628 1629 1630show bonding config 1631~~~~~~~~~~~~~~~~~~~ 1632 1633Show the current configuration of a Link Bonding device:: 1634 1635 testpmd> show bonding config (port id) 1636 1637For example, 1638to show the configuration a Link Bonding device (port 9) with 3 slave devices (1, 3, 4) 1639in balance mode with a transmission policy of layer 2+3:: 1640 1641 testpmd> show bonding config 9 1642 Bonding mode: 2 1643 Balance Xmit Policy: BALANCE_XMIT_POLICY_LAYER23 1644 Slaves (3): [1 3 4] 1645 Active Slaves (3): [1 3 4] 1646 Primary: [3] 1647 1648 1649Register Functions 1650------------------ 1651 1652The Register Functions can be used to read from and write to registers on the network card referenced by a port number. 1653This is mainly useful for debugging purposes. 1654Reference should be made to the appropriate datasheet for the network card for details on the register addresses 1655and fields that can be accessed. 1656 1657read reg 1658~~~~~~~~ 1659 1660Display the value of a port register:: 1661 1662 testpmd> read reg (port_id) (address) 1663 1664For example, to examine the Flow Director control register (FDIRCTL, 0x0000EE000) on an Intel 82599 10 GbE Controller:: 1665 1666 testpmd> read reg 0 0xEE00 1667 port 0 PCI register at offset 0xEE00: 0x4A060029 (1241907241) 1668 1669read regfield 1670~~~~~~~~~~~~~ 1671 1672Display a port register bit field:: 1673 1674 testpmd> read regfield (port_id) (address) (bit_x) (bit_y) 1675 1676For example, reading the lowest two bits from the register in the example above:: 1677 1678 testpmd> read regfield 0 0xEE00 0 1 1679 port 0 PCI register at offset 0xEE00: bits[0, 1]=0x1 (1) 1680 1681read regbit 1682~~~~~~~~~~~ 1683 1684Display a single port register bit:: 1685 1686 testpmd> read regbit (port_id) (address) (bit_x) 1687 1688For example, reading the lowest bit from the register in the example above:: 1689 1690 testpmd> read regbit 0 0xEE00 0 1691 port 0 PCI register at offset 0xEE00: bit 0=1 1692 1693write reg 1694~~~~~~~~~ 1695 1696Set the value of a port register:: 1697 1698 testpmd> write reg (port_id) (address) (value) 1699 1700For example, to clear a register:: 1701 1702 testpmd> write reg 0 0xEE00 0x0 1703 port 0 PCI register at offset 0xEE00: 0x00000000 (0) 1704 1705write regfield 1706~~~~~~~~~~~~~~ 1707 1708Set bit field of a port register:: 1709 1710 testpmd> write regfield (port_id) (address) (bit_x) (bit_y) (value) 1711 1712For example, writing to the register cleared in the example above:: 1713 1714 testpmd> write regfield 0 0xEE00 0 1 2 1715 port 0 PCI register at offset 0xEE00: 0x00000002 (2) 1716 1717write regbit 1718~~~~~~~~~~~~ 1719 1720Set single bit value of a port register:: 1721 1722 testpmd> write regbit (port_id) (address) (bit_x) (value) 1723 1724For example, to set the high bit in the register from the example above:: 1725 1726 testpmd> write regbit 0 0xEE00 31 1 1727 port 0 PCI register at offset 0xEE00: 0x8000000A (2147483658) 1728 1729 1730Filter Functions 1731---------------- 1732 1733This section details the available filter functions that are available. 1734 1735Note these functions interface the deprecated legacy filtering framework, 1736superseded by *rte_flow*. See `Flow rules management`_. 1737 1738ethertype_filter 1739~~~~~~~~~~~~~~~~~~~~ 1740 1741Add or delete a L2 Ethertype filter, which identify packets by their L2 Ethertype mainly assign them to a receive queue:: 1742 1743 ethertype_filter (port_id) (add|del) (mac_addr|mac_ignr) (mac_address) \ 1744 ethertype (ether_type) (drop|fwd) queue (queue_id) 1745 1746The available information parameters are: 1747 1748* ``port_id``: The port which the Ethertype filter assigned on. 1749 1750* ``mac_addr``: Compare destination mac address. 1751 1752* ``mac_ignr``: Ignore destination mac address match. 1753 1754* ``mac_address``: Destination mac address to match. 1755 1756* ``ether_type``: The EtherType value want to match, 1757 for example 0x0806 for ARP packet. 0x0800 (IPv4) and 0x86DD (IPv6) are invalid. 1758 1759* ``queue_id``: The receive queue associated with this EtherType filter. 1760 It is meaningless when deleting or dropping. 1761 1762Example, to add/remove an ethertype filter rule:: 1763 1764 testpmd> ethertype_filter 0 add mac_ignr 00:11:22:33:44:55 \ 1765 ethertype 0x0806 fwd queue 3 1766 1767 testpmd> ethertype_filter 0 del mac_ignr 00:11:22:33:44:55 \ 1768 ethertype 0x0806 fwd queue 3 1769 17702tuple_filter 1771~~~~~~~~~~~~~~~~~ 1772 1773Add or delete a 2-tuple filter, 1774which identifies packets by specific protocol and destination TCP/UDP port 1775and forwards packets into one of the receive queues:: 1776 1777 2tuple_filter (port_id) (add|del) dst_port (dst_port_value) \ 1778 protocol (protocol_value) mask (mask_value) \ 1779 tcp_flags (tcp_flags_value) priority (prio_value) \ 1780 queue (queue_id) 1781 1782The available information parameters are: 1783 1784* ``port_id``: The port which the 2-tuple filter assigned on. 1785 1786* ``dst_port_value``: Destination port in L4. 1787 1788* ``protocol_value``: IP L4 protocol. 1789 1790* ``mask_value``: Participates in the match or not by bit for field above, 1b means participate. 1791 1792* ``tcp_flags_value``: TCP control bits. The non-zero value is invalid, when the pro_value is not set to 0x06 (TCP). 1793 1794* ``prio_value``: Priority of this filter. 1795 1796* ``queue_id``: The receive queue associated with this 2-tuple filter. 1797 1798Example, to add/remove an 2tuple filter rule:: 1799 1800 testpmd> 2tuple_filter 0 add dst_port 32 protocol 0x06 mask 0x03 \ 1801 tcp_flags 0x02 priority 3 queue 3 1802 1803 testpmd> 2tuple_filter 0 del dst_port 32 protocol 0x06 mask 0x03 \ 1804 tcp_flags 0x02 priority 3 queue 3 1805 18065tuple_filter 1807~~~~~~~~~~~~~~~~~ 1808 1809Add or delete a 5-tuple filter, 1810which consists of a 5-tuple (protocol, source and destination IP addresses, source and destination TCP/UDP/SCTP port) 1811and routes packets into one of the receive queues:: 1812 1813 5tuple_filter (port_id) (add|del) dst_ip (dst_address) src_ip \ 1814 (src_address) dst_port (dst_port_value) \ 1815 src_port (src_port_value) protocol (protocol_value) \ 1816 mask (mask_value) tcp_flags (tcp_flags_value) \ 1817 priority (prio_value) queue (queue_id) 1818 1819The available information parameters are: 1820 1821* ``port_id``: The port which the 5-tuple filter assigned on. 1822 1823* ``dst_address``: Destination IP address. 1824 1825* ``src_address``: Source IP address. 1826 1827* ``dst_port_value``: TCP/UDP destination port. 1828 1829* ``src_port_value``: TCP/UDP source port. 1830 1831* ``protocol_value``: L4 protocol. 1832 1833* ``mask_value``: Participates in the match or not by bit for field above, 1b means participate 1834 1835* ``tcp_flags_value``: TCP control bits. The non-zero value is invalid, when the protocol_value is not set to 0x06 (TCP). 1836 1837* ``prio_value``: The priority of this filter. 1838 1839* ``queue_id``: The receive queue associated with this 5-tuple filter. 1840 1841Example, to add/remove an 5tuple filter rule:: 1842 1843 testpmd> 5tuple_filter 0 add dst_ip 2.2.2.5 src_ip 2.2.2.4 \ 1844 dst_port 64 src_port 32 protocol 0x06 mask 0x1F \ 1845 flags 0x0 priority 3 queue 3 1846 1847 testpmd> 5tuple_filter 0 del dst_ip 2.2.2.5 src_ip 2.2.2.4 \ 1848 dst_port 64 src_port 32 protocol 0x06 mask 0x1F \ 1849 flags 0x0 priority 3 queue 3 1850 1851syn_filter 1852~~~~~~~~~~ 1853 1854Using the SYN filter, TCP packets whose *SYN* flag is set can be forwarded to a separate queue:: 1855 1856 syn_filter (port_id) (add|del) priority (high|low) queue (queue_id) 1857 1858The available information parameters are: 1859 1860* ``port_id``: The port which the SYN filter assigned on. 1861 1862* ``high``: This SYN filter has higher priority than other filters. 1863 1864* ``low``: This SYN filter has lower priority than other filters. 1865 1866* ``queue_id``: The receive queue associated with this SYN filter 1867 1868Example:: 1869 1870 testpmd> syn_filter 0 add priority high queue 3 1871 1872flex_filter 1873~~~~~~~~~~~ 1874 1875With flex filter, packets can be recognized by any arbitrary pattern within the first 128 bytes of the packet 1876and routed into one of the receive queues:: 1877 1878 flex_filter (port_id) (add|del) len (len_value) bytes (bytes_value) \ 1879 mask (mask_value) priority (prio_value) queue (queue_id) 1880 1881The available information parameters are: 1882 1883* ``port_id``: The port which the Flex filter is assigned on. 1884 1885* ``len_value``: Filter length in bytes, no greater than 128. 1886 1887* ``bytes_value``: A string in hexadecimal, means the value the flex filter needs to match. 1888 1889* ``mask_value``: A string in hexadecimal, bit 1 means corresponding byte participates in the match. 1890 1891* ``prio_value``: The priority of this filter. 1892 1893* ``queue_id``: The receive queue associated with this Flex filter. 1894 1895Example:: 1896 1897 testpmd> flex_filter 0 add len 16 bytes 0x00000000000000000000000008060000 \ 1898 mask 000C priority 3 queue 3 1899 1900 testpmd> flex_filter 0 del len 16 bytes 0x00000000000000000000000008060000 \ 1901 mask 000C priority 3 queue 3 1902 1903 1904.. _testpmd_flow_director: 1905 1906flow_director_filter 1907~~~~~~~~~~~~~~~~~~~~ 1908 1909The Flow Director works in receive mode to identify specific flows or sets of flows and route them to specific queues. 1910 1911Four types of filtering are supported which are referred to as Perfect Match, Signature, Perfect-mac-vlan and 1912Perfect-tunnel filters, the match mode is set by the ``--pkt-filter-mode`` command-line parameter: 1913 1914* Perfect match filters. 1915 The hardware checks a match between the masked fields of the received packets and the programmed filters. 1916 The masked fields are for IP flow. 1917 1918* Signature filters. 1919 The hardware checks a match between a hash-based signature of the masked fields of the received packet. 1920 1921* Perfect-mac-vlan match filters. 1922 The hardware checks a match between the masked fields of the received packets and the programmed filters. 1923 The masked fields are for MAC VLAN flow. 1924 1925* Perfect-tunnel match filters. 1926 The hardware checks a match between the masked fields of the received packets and the programmed filters. 1927 The masked fields are for tunnel flow. 1928 1929The Flow Director filters can match the different fields for different type of packet: flow type, specific input set 1930per flow type and the flexible payload. 1931 1932The Flow Director can also mask out parts of all of these fields so that filters 1933are only applied to certain fields or parts of the fields. 1934 1935Different NICs may have different capabilities, command show port fdir (port_id) can be used to acquire the information. 1936 1937# Commands to add flow director filters of different flow types:: 1938 1939 flow_director_filter (port_id) mode IP (add|del|update) \ 1940 flow (ipv4-other|ipv4-frag|ipv6-other|ipv6-frag) \ 1941 src (src_ip_address) dst (dst_ip_address) \ 1942 tos (tos_value) proto (proto_value) ttl (ttl_value) \ 1943 vlan (vlan_value) flexbytes (flexbytes_value) \ 1944 (drop|fwd) pf|vf(vf_id) queue (queue_id) \ 1945 fd_id (fd_id_value) 1946 1947 flow_director_filter (port_id) mode IP (add|del|update) \ 1948 flow (ipv4-tcp|ipv4-udp|ipv6-tcp|ipv6-udp) \ 1949 src (src_ip_address) (src_port) \ 1950 dst (dst_ip_address) (dst_port) \ 1951 tos (tos_value) ttl (ttl_value) \ 1952 vlan (vlan_value) flexbytes (flexbytes_value) \ 1953 (drop|fwd) queue pf|vf(vf_id) (queue_id) \ 1954 fd_id (fd_id_value) 1955 1956 flow_director_filter (port_id) mode IP (add|del|update) \ 1957 flow (ipv4-sctp|ipv6-sctp) \ 1958 src (src_ip_address) (src_port) \ 1959 dst (dst_ip_address) (dst_port) \ 1960 tos (tos_value) ttl (ttl_value) \ 1961 tag (verification_tag) vlan (vlan_value) \ 1962 flexbytes (flexbytes_value) (drop|fwd) \ 1963 pf|vf(vf_id) queue (queue_id) fd_id (fd_id_value) 1964 1965 flow_director_filter (port_id) mode IP (add|del|update) flow l2_payload \ 1966 ether (ethertype) flexbytes (flexbytes_value) \ 1967 (drop|fwd) pf|vf(vf_id) queue (queue_id) 1968 fd_id (fd_id_value) 1969 1970 flow_director_filter (port_id) mode MAC-VLAN (add|del|update) \ 1971 mac (mac_address) vlan (vlan_value) \ 1972 flexbytes (flexbytes_value) (drop|fwd) \ 1973 queue (queue_id) fd_id (fd_id_value) 1974 1975 flow_director_filter (port_id) mode Tunnel (add|del|update) \ 1976 mac (mac_address) vlan (vlan_value) \ 1977 tunnel (NVGRE|VxLAN) tunnel-id (tunnel_id_value) \ 1978 flexbytes (flexbytes_value) (drop|fwd) \ 1979 queue (queue_id) fd_id (fd_id_value) 1980 1981For example, to add an ipv4-udp flow type filter:: 1982 1983 testpmd> flow_director_filter 0 mode IP add flow ipv4-udp src 2.2.2.3 32 \ 1984 dst 2.2.2.5 33 tos 2 ttl 40 vlan 0x1 flexbytes (0x88,0x48) \ 1985 fwd pf queue 1 fd_id 1 1986 1987For example, add an ipv4-other flow type filter:: 1988 1989 testpmd> flow_director_filter 0 mode IP add flow ipv4-other src 2.2.2.3 \ 1990 dst 2.2.2.5 tos 2 proto 20 ttl 40 vlan 0x1 \ 1991 flexbytes (0x88,0x48) fwd pf queue 1 fd_id 1 1992 1993flush_flow_director 1994~~~~~~~~~~~~~~~~~~~ 1995 1996Flush all flow director filters on a device:: 1997 1998 testpmd> flush_flow_director (port_id) 1999 2000Example, to flush all flow director filter on port 0:: 2001 2002 testpmd> flush_flow_director 0 2003 2004flow_director_mask 2005~~~~~~~~~~~~~~~~~~ 2006 2007Set flow director's input masks:: 2008 2009 flow_director_mask (port_id) mode IP vlan (vlan_value) \ 2010 src_mask (ipv4_src) (ipv6_src) (src_port) \ 2011 dst_mask (ipv4_dst) (ipv6_dst) (dst_port) 2012 2013 flow_director_mask (port_id) mode MAC-VLAN vlan (vlan_value) 2014 2015 flow_director_mask (port_id) mode Tunnel vlan (vlan_value) \ 2016 mac (mac_value) tunnel-type (tunnel_type_value) \ 2017 tunnel-id (tunnel_id_value) 2018 2019Example, to set flow director mask on port 0:: 2020 2021 testpmd> flow_director_mask 0 mode IP vlan 0xefff \ 2022 src_mask 255.255.255.255 \ 2023 FFFF:FFFF:FFFF:FFFF:FFFF:FFFF:FFFF:FFFF 0xFFFF \ 2024 dst_mask 255.255.255.255 \ 2025 FFFF:FFFF:FFFF:FFFF:FFFF:FFFF:FFFF:FFFF 0xFFFF 2026 2027flow_director_flex_mask 2028~~~~~~~~~~~~~~~~~~~~~~~ 2029 2030set masks of flow director's flexible payload based on certain flow type:: 2031 2032 testpmd> flow_director_flex_mask (port_id) \ 2033 flow (none|ipv4-other|ipv4-frag|ipv4-tcp|ipv4-udp|ipv4-sctp| \ 2034 ipv6-other|ipv6-frag|ipv6-tcp|ipv6-udp|ipv6-sctp| \ 2035 l2_payload|all) (mask) 2036 2037Example, to set flow director's flex mask for all flow type on port 0:: 2038 2039 testpmd> flow_director_flex_mask 0 flow all \ 2040 (0xff,0xff,0,0,0,0,0,0,0,0,0,0,0,0,0,0) 2041 2042 2043flow_director_flex_payload 2044~~~~~~~~~~~~~~~~~~~~~~~~~~ 2045 2046Configure flexible payload selection:: 2047 2048 flow_director_flex_payload (port_id) (raw|l2|l3|l4) (config) 2049 2050For example, to select the first 16 bytes from the offset 4 (bytes) of packet's payload as flexible payload:: 2051 2052 testpmd> flow_director_flex_payload 0 l4 \ 2053 (4,5,6,7,8,9,10,11,12,13,14,15,16,17,18,19) 2054 2055get_sym_hash_ena_per_port 2056~~~~~~~~~~~~~~~~~~~~~~~~~ 2057 2058Get symmetric hash enable configuration per port:: 2059 2060 get_sym_hash_ena_per_port (port_id) 2061 2062For example, to get symmetric hash enable configuration of port 1:: 2063 2064 testpmd> get_sym_hash_ena_per_port 1 2065 2066set_sym_hash_ena_per_port 2067~~~~~~~~~~~~~~~~~~~~~~~~~ 2068 2069Set symmetric hash enable configuration per port to enable or disable:: 2070 2071 set_sym_hash_ena_per_port (port_id) (enable|disable) 2072 2073For example, to set symmetric hash enable configuration of port 1 to enable:: 2074 2075 testpmd> set_sym_hash_ena_per_port 1 enable 2076 2077get_hash_global_config 2078~~~~~~~~~~~~~~~~~~~~~~ 2079 2080Get the global configurations of hash filters:: 2081 2082 get_hash_global_config (port_id) 2083 2084For example, to get the global configurations of hash filters of port 1:: 2085 2086 testpmd> get_hash_global_config 1 2087 2088set_hash_global_config 2089~~~~~~~~~~~~~~~~~~~~~~ 2090 2091Set the global configurations of hash filters:: 2092 2093 set_hash_global_config (port_id) (toeplitz|simple_xor|default) \ 2094 (ipv4|ipv4-frag|ipv4-tcp|ipv4-udp|ipv4-sctp|ipv4-other|ipv6|ipv6-frag| \ 2095 ipv6-tcp|ipv6-udp|ipv6-sctp|ipv6-other|l2_payload) \ 2096 (enable|disable) 2097 2098For example, to enable simple_xor for flow type of ipv6 on port 2:: 2099 2100 testpmd> set_hash_global_config 2 simple_xor ipv6 enable 2101 2102set_hash_input_set 2103~~~~~~~~~~~~~~~~~~ 2104 2105Set the input set for hash:: 2106 2107 set_hash_input_set (port_id) (ipv4-frag|ipv4-tcp|ipv4-udp|ipv4-sctp| \ 2108 ipv4-other|ipv6-frag|ipv6-tcp|ipv6-udp|ipv6-sctp|ipv6-other| \ 2109 l2_payload) (ovlan|ivlan|src-ipv4|dst-ipv4|src-ipv6|dst-ipv6|ipv4-tos| \ 2110 ipv4-proto|ipv6-tc|ipv6-next-header|udp-src-port|udp-dst-port| \ 2111 tcp-src-port|tcp-dst-port|sctp-src-port|sctp-dst-port|sctp-veri-tag| \ 2112 udp-key|gre-key|fld-1st|fld-2nd|fld-3rd|fld-4th|fld-5th|fld-6th|fld-7th| \ 2113 fld-8th|none) (select|add) 2114 2115For example, to add source IP to hash input set for flow type of ipv4-udp on port 0:: 2116 2117 testpmd> set_hash_input_set 0 ipv4-udp src-ipv4 add 2118 2119set_fdir_input_set 2120~~~~~~~~~~~~~~~~~~ 2121 2122The Flow Director filters can match the different fields for different type of packet, i.e. specific input set 2123on per flow type and the flexible payload. This command can be used to change input set for each flow type. 2124 2125Set the input set for flow director:: 2126 2127 set_fdir_input_set (port_id) (ipv4-frag|ipv4-tcp|ipv4-udp|ipv4-sctp| \ 2128 ipv4-other|ipv6|ipv6-frag|ipv6-tcp|ipv6-udp|ipv6-sctp|ipv6-other| \ 2129 l2_payload) (ivlan|ethertype|src-ipv4|dst-ipv4|src-ipv6|dst-ipv6|ipv4-tos| \ 2130 ipv4-proto|ipv4-ttl|ipv6-tc|ipv6-next-header|ipv6-hop-limits| \ 2131 tudp-src-port|udp-dst-port|cp-src-port|tcp-dst-port|sctp-src-port| \ 2132 sctp-dst-port|sctp-veri-tag|none) (select|add) 2133 2134For example to add source IP to FD input set for flow type of ipv4-udp on port 0:: 2135 2136 testpmd> set_fdir_input_set 0 ipv4-udp src-ipv4 add 2137 2138global_config 2139~~~~~~~~~~~~~ 2140 2141Set different GRE key length for input set:: 2142 2143 global_config (port_id) gre-key-len (number in bytes) 2144 2145For example to set GRE key length for input set to 4 bytes on port 0:: 2146 2147 testpmd> global_config 0 gre-key-len 4 2148 2149 2150.. _testpmd_rte_flow: 2151 2152Flow rules management 2153--------------------- 2154 2155Control of the generic flow API (*rte_flow*) is fully exposed through the 2156``flow`` command (validation, creation, destruction and queries). 2157 2158Considering *rte_flow* overlaps with all `Filter Functions`_, using both 2159features simultaneously may cause undefined side-effects and is therefore 2160not recommended. 2161 2162``flow`` syntax 2163~~~~~~~~~~~~~~~ 2164 2165Because the ``flow`` command uses dynamic tokens to handle the large number 2166of possible flow rules combinations, its behavior differs slightly from 2167other commands, in particular: 2168 2169- Pressing *?* or the *<tab>* key displays contextual help for the current 2170 token, not that of the entire command. 2171 2172- Optional and repeated parameters are supported (provided they are listed 2173 in the contextual help). 2174 2175The first parameter stands for the operation mode. Possible operations and 2176their general syntax are described below. They are covered in detail in the 2177following sections. 2178 2179- Check whether a flow rule can be created:: 2180 2181 flow validate {port_id} 2182 [group {group_id}] [priority {level}] [ingress] [egress] 2183 pattern {item} [/ {item} [...]] / end 2184 actions {action} [/ {action} [...]] / end 2185 2186- Create a flow rule:: 2187 2188 flow create {port_id} 2189 [group {group_id}] [priority {level}] [ingress] [egress] 2190 pattern {item} [/ {item} [...]] / end 2191 actions {action} [/ {action} [...]] / end 2192 2193- Destroy specific flow rules:: 2194 2195 flow destroy {port_id} rule {rule_id} [...] 2196 2197- Destroy all flow rules:: 2198 2199 flow flush {port_id} 2200 2201- Query an existing flow rule:: 2202 2203 flow query {port_id} {rule_id} {action} 2204 2205- List existing flow rules sorted by priority, filtered by group 2206 identifiers:: 2207 2208 flow list {port_id} [group {group_id}] [...] 2209 2210Validating flow rules 2211~~~~~~~~~~~~~~~~~~~~~ 2212 2213``flow validate`` reports whether a flow rule would be accepted by the 2214underlying device in its current state but stops short of creating it. It is 2215bound to ``rte_flow_validate()``:: 2216 2217 flow validate {port_id} 2218 [group {group_id}] [priority {level}] [ingress] [egress] 2219 pattern {item} [/ {item} [...]] / end 2220 actions {action} [/ {action} [...]] / end 2221 2222If successful, it will show:: 2223 2224 Flow rule validated 2225 2226Otherwise it will show an error message of the form:: 2227 2228 Caught error type [...] ([...]): [...] 2229 2230This command uses the same parameters as ``flow create``, their format is 2231described in `Creating flow rules`_. 2232 2233Check whether redirecting any Ethernet packet received on port 0 to RX queue 2234index 6 is supported:: 2235 2236 testpmd> flow validate 0 ingress pattern eth / end 2237 actions queue index 6 / end 2238 Flow rule validated 2239 testpmd> 2240 2241Port 0 does not support TCPv6 rules:: 2242 2243 testpmd> flow validate 0 ingress pattern eth / ipv6 / tcp / end 2244 actions drop / end 2245 Caught error type 9 (specific pattern item): Invalid argument 2246 testpmd> 2247 2248Creating flow rules 2249~~~~~~~~~~~~~~~~~~~ 2250 2251``flow create`` validates and creates the specified flow rule. It is bound 2252to ``rte_flow_create()``:: 2253 2254 flow create {port_id} 2255 [group {group_id}] [priority {level}] [ingress] [egress] 2256 pattern {item} [/ {item} [...]] / end 2257 actions {action} [/ {action} [...]] / end 2258 2259If successful, it will return a flow rule ID usable with other commands:: 2260 2261 Flow rule #[...] created 2262 2263Otherwise it will show an error message of the form:: 2264 2265 Caught error type [...] ([...]): [...] 2266 2267Parameters describe in the following order: 2268 2269- Attributes (*group*, *priority*, *ingress*, *egress* tokens). 2270- A matching pattern, starting with the *pattern* token and terminated by an 2271 *end* pattern item. 2272- Actions, starting with the *actions* token and terminated by an *end* 2273 action. 2274 2275These translate directly to *rte_flow* objects provided as-is to the 2276underlying functions. 2277 2278The shortest valid definition only comprises mandatory tokens:: 2279 2280 testpmd> flow create 0 pattern end actions end 2281 2282Note that PMDs may refuse rules that essentially do nothing such as this 2283one. 2284 2285**All unspecified object values are automatically initialized to 0.** 2286 2287Attributes 2288^^^^^^^^^^ 2289 2290These tokens affect flow rule attributes (``struct rte_flow_attr``) and are 2291specified before the ``pattern`` token. 2292 2293- ``group {group id}``: priority group. 2294- ``priority {level}``: priority level within group. 2295- ``ingress``: rule applies to ingress traffic. 2296- ``egress``: rule applies to egress traffic. 2297 2298Each instance of an attribute specified several times overrides the previous 2299value as shown below (group 4 is used):: 2300 2301 testpmd> flow create 0 group 42 group 24 group 4 [...] 2302 2303Note that once enabled, ``ingress`` and ``egress`` cannot be disabled. 2304 2305While not specifying a direction is an error, some rules may allow both 2306simultaneously. 2307 2308Most rules affect RX therefore contain the ``ingress`` token:: 2309 2310 testpmd> flow create 0 ingress pattern [...] 2311 2312Matching pattern 2313^^^^^^^^^^^^^^^^ 2314 2315A matching pattern starts after the ``pattern`` token. It is made of pattern 2316items and is terminated by a mandatory ``end`` item. 2317 2318Items are named after their type (*RTE_FLOW_ITEM_TYPE_* from ``enum 2319rte_flow_item_type``). 2320 2321The ``/`` token is used as a separator between pattern items as shown 2322below:: 2323 2324 testpmd> flow create 0 ingress pattern eth / ipv4 / udp / end [...] 2325 2326Note that protocol items like these must be stacked from lowest to highest 2327layer to make sense. For instance, the following rule is either invalid or 2328unlikely to match any packet:: 2329 2330 testpmd> flow create 0 ingress pattern eth / udp / ipv4 / end [...] 2331 2332More information on these restrictions can be found in the *rte_flow* 2333documentation. 2334 2335Several items support additional specification structures, for example 2336``ipv4`` allows specifying source and destination addresses as follows:: 2337 2338 testpmd> flow create 0 ingress pattern eth / ipv4 src is 10.1.1.1 2339 dst is 10.2.0.0 / end [...] 2340 2341This rule matches all IPv4 traffic with the specified properties. 2342 2343In this example, ``src`` and ``dst`` are field names of the underlying 2344``struct rte_flow_item_ipv4`` object. All item properties can be specified 2345in a similar fashion. 2346 2347The ``is`` token means that the subsequent value must be matched exactly, 2348and assigns ``spec`` and ``mask`` fields in ``struct rte_flow_item`` 2349accordingly. Possible assignment tokens are: 2350 2351- ``is``: match value perfectly (with full bit-mask). 2352- ``spec``: match value according to configured bit-mask. 2353- ``last``: specify upper bound to establish a range. 2354- ``mask``: specify bit-mask with relevant bits set to one. 2355- ``prefix``: generate bit-mask from a prefix length. 2356 2357These yield identical results:: 2358 2359 ipv4 src is 10.1.1.1 2360 2361:: 2362 2363 ipv4 src spec 10.1.1.1 src mask 255.255.255.255 2364 2365:: 2366 2367 ipv4 src spec 10.1.1.1 src prefix 32 2368 2369:: 2370 2371 ipv4 src is 10.1.1.1 src last 10.1.1.1 # range with a single value 2372 2373:: 2374 2375 ipv4 src is 10.1.1.1 src last 0 # 0 disables range 2376 2377Inclusive ranges can be defined with ``last``:: 2378 2379 ipv4 src is 10.1.1.1 src last 10.2.3.4 # 10.1.1.1 to 10.2.3.4 2380 2381Note that ``mask`` affects both ``spec`` and ``last``:: 2382 2383 ipv4 src is 10.1.1.1 src last 10.2.3.4 src mask 255.255.0.0 2384 # matches 10.1.0.0 to 10.2.255.255 2385 2386Properties can be modified multiple times:: 2387 2388 ipv4 src is 10.1.1.1 src is 10.1.2.3 src is 10.2.3.4 # matches 10.2.3.4 2389 2390:: 2391 2392 ipv4 src is 10.1.1.1 src prefix 24 src prefix 16 # matches 10.1.0.0/16 2393 2394Pattern items 2395^^^^^^^^^^^^^ 2396 2397This section lists supported pattern items and their attributes, if any. 2398 2399- ``end``: end list of pattern items. 2400 2401- ``void``: no-op pattern item. 2402 2403- ``invert``: perform actions when pattern does not match. 2404 2405- ``any``: match any protocol for the current layer. 2406 2407 - ``num {unsigned}``: number of layers covered. 2408 2409- ``pf``: match packets addressed to the physical function. 2410 2411- ``vf``: match packets addressed to a virtual function ID. 2412 2413 - ``id {unsigned}``: destination VF ID. 2414 2415- ``port``: device-specific physical port index to use. 2416 2417 - ``index {unsigned}``: physical port index. 2418 2419- ``raw``: match an arbitrary byte string. 2420 2421 - ``relative {boolean}``: look for pattern after the previous item. 2422 - ``search {boolean}``: search pattern from offset (see also limit). 2423 - ``offset {integer}``: absolute or relative offset for pattern. 2424 - ``limit {unsigned}``: search area limit for start of pattern. 2425 - ``pattern {string}``: byte string to look for. 2426 2427- ``eth``: match Ethernet header. 2428 2429 - ``dst {MAC-48}``: destination MAC. 2430 - ``src {MAC-48}``: source MAC. 2431 - ``type {unsigned}``: EtherType. 2432 2433- ``vlan``: match 802.1Q/ad VLAN tag. 2434 2435 - ``tpid {unsigned}``: tag protocol identifier. 2436 - ``tci {unsigned}``: tag control information. 2437 - ``pcp {unsigned}``: priority code point. 2438 - ``dei {unsigned}``: drop eligible indicator. 2439 - ``vid {unsigned}``: VLAN identifier. 2440 2441- ``ipv4``: match IPv4 header. 2442 2443 - ``tos {unsigned}``: type of service. 2444 - ``ttl {unsigned}``: time to live. 2445 - ``proto {unsigned}``: next protocol ID. 2446 - ``src {ipv4 address}``: source address. 2447 - ``dst {ipv4 address}``: destination address. 2448 2449- ``ipv6``: match IPv6 header. 2450 2451 - ``tc {unsigned}``: traffic class. 2452 - ``flow {unsigned}``: flow label. 2453 - ``proto {unsigned}``: protocol (next header). 2454 - ``hop {unsigned}``: hop limit. 2455 - ``src {ipv6 address}``: source address. 2456 - ``dst {ipv6 address}``: destination address. 2457 2458- ``icmp``: match ICMP header. 2459 2460 - ``type {unsigned}``: ICMP packet type. 2461 - ``code {unsigned}``: ICMP packet code. 2462 2463- ``udp``: match UDP header. 2464 2465 - ``src {unsigned}``: UDP source port. 2466 - ``dst {unsigned}``: UDP destination port. 2467 2468- ``tcp``: match TCP header. 2469 2470 - ``src {unsigned}``: TCP source port. 2471 - ``dst {unsigned}``: TCP destination port. 2472 2473- ``sctp``: match SCTP header. 2474 2475 - ``src {unsigned}``: SCTP source port. 2476 - ``dst {unsigned}``: SCTP destination port. 2477 - ``tag {unsigned}``: validation tag. 2478 - ``cksum {unsigned}``: checksum. 2479 2480- ``vxlan``: match VXLAN header. 2481 2482 - ``vni {unsigned}``: VXLAN identifier. 2483 2484- ``mpls``: match MPLS header. 2485 2486 - ``label {unsigned}``: MPLS label. 2487 2488- ``gre``: match GRE header. 2489 2490 - ``protocol {unsigned}``: protocol type. 2491 2492Actions list 2493^^^^^^^^^^^^ 2494 2495A list of actions starts after the ``actions`` token in the same fashion as 2496`Matching pattern`_; actions are separated by ``/`` tokens and the list is 2497terminated by a mandatory ``end`` action. 2498 2499Actions are named after their type (*RTE_FLOW_ACTION_TYPE_* from ``enum 2500rte_flow_action_type``). 2501 2502Dropping all incoming UDPv4 packets can be expressed as follows:: 2503 2504 testpmd> flow create 0 ingress pattern eth / ipv4 / udp / end 2505 actions drop / end 2506 2507Several actions have configurable properties which must be specified when 2508there is no valid default value. For example, ``queue`` requires a target 2509queue index. 2510 2511This rule redirects incoming UDPv4 traffic to queue index 6:: 2512 2513 testpmd> flow create 0 ingress pattern eth / ipv4 / udp / end 2514 actions queue index 6 / end 2515 2516While this one could be rejected by PMDs (unspecified queue index):: 2517 2518 testpmd> flow create 0 ingress pattern eth / ipv4 / udp / end 2519 actions queue / end 2520 2521As defined by *rte_flow*, the list is not ordered, all actions of a given 2522rule are performed simultaneously. These are equivalent:: 2523 2524 queue index 6 / void / mark id 42 / end 2525 2526:: 2527 2528 void / mark id 42 / queue index 6 / end 2529 2530All actions in a list should have different types, otherwise only the last 2531action of a given type is taken into account:: 2532 2533 queue index 4 / queue index 5 / queue index 6 / end # will use queue 6 2534 2535:: 2536 2537 drop / drop / drop / end # drop is performed only once 2538 2539:: 2540 2541 mark id 42 / queue index 3 / mark id 24 / end # mark will be 24 2542 2543Considering they are performed simultaneously, opposite and overlapping 2544actions can sometimes be combined when the end result is unambiguous:: 2545 2546 drop / queue index 6 / end # drop has no effect 2547 2548:: 2549 2550 drop / dup index 6 / end # same as above 2551 2552:: 2553 2554 queue index 6 / rss queues 6 7 8 / end # queue has no effect 2555 2556:: 2557 2558 drop / passthru / end # drop has no effect 2559 2560Note that PMDs may still refuse such combinations. 2561 2562Actions 2563^^^^^^^ 2564 2565This section lists supported actions and their attributes, if any. 2566 2567- ``end``: end list of actions. 2568 2569- ``void``: no-op action. 2570 2571- ``passthru``: let subsequent rule process matched packets. 2572 2573- ``mark``: attach 32 bit value to packets. 2574 2575 - ``id {unsigned}``: 32 bit value to return with packets. 2576 2577- ``flag``: flag packets. 2578 2579- ``queue``: assign packets to a given queue index. 2580 2581 - ``index {unsigned}``: queue index to use. 2582 2583- ``drop``: drop packets (note: passthru has priority). 2584 2585- ``count``: enable counters for this rule. 2586 2587- ``dup``: duplicate packets to a given queue index. 2588 2589 - ``index {unsigned}``: queue index to duplicate packets to. 2590 2591- ``rss``: spread packets among several queues. 2592 2593 - ``queues [{unsigned} [...]] end``: queue indices to use. 2594 2595- ``pf``: redirect packets to physical device function. 2596 2597- ``vf``: redirect packets to virtual device function. 2598 2599 - ``original {boolean}``: use original VF ID if possible. 2600 - ``id {unsigned}``: VF ID to redirect packets to. 2601 2602Destroying flow rules 2603~~~~~~~~~~~~~~~~~~~~~ 2604 2605``flow destroy`` destroys one or more rules from their rule ID (as returned 2606by ``flow create``), this command calls ``rte_flow_destroy()`` as many 2607times as necessary:: 2608 2609 flow destroy {port_id} rule {rule_id} [...] 2610 2611If successful, it will show:: 2612 2613 Flow rule #[...] destroyed 2614 2615It does not report anything for rule IDs that do not exist. The usual error 2616message is shown when a rule cannot be destroyed:: 2617 2618 Caught error type [...] ([...]): [...] 2619 2620``flow flush`` destroys all rules on a device and does not take extra 2621arguments. It is bound to ``rte_flow_flush()``:: 2622 2623 flow flush {port_id} 2624 2625Any errors are reported as above. 2626 2627Creating several rules and destroying them:: 2628 2629 testpmd> flow create 0 ingress pattern eth / ipv6 / end 2630 actions queue index 2 / end 2631 Flow rule #0 created 2632 testpmd> flow create 0 ingress pattern eth / ipv4 / end 2633 actions queue index 3 / end 2634 Flow rule #1 created 2635 testpmd> flow destroy 0 rule 0 rule 1 2636 Flow rule #1 destroyed 2637 Flow rule #0 destroyed 2638 testpmd> 2639 2640The same result can be achieved using ``flow flush``:: 2641 2642 testpmd> flow create 0 ingress pattern eth / ipv6 / end 2643 actions queue index 2 / end 2644 Flow rule #0 created 2645 testpmd> flow create 0 ingress pattern eth / ipv4 / end 2646 actions queue index 3 / end 2647 Flow rule #1 created 2648 testpmd> flow flush 0 2649 testpmd> 2650 2651Non-existent rule IDs are ignored:: 2652 2653 testpmd> flow create 0 ingress pattern eth / ipv6 / end 2654 actions queue index 2 / end 2655 Flow rule #0 created 2656 testpmd> flow create 0 ingress pattern eth / ipv4 / end 2657 actions queue index 3 / end 2658 Flow rule #1 created 2659 testpmd> flow destroy 0 rule 42 rule 10 rule 2 2660 testpmd> 2661 testpmd> flow destroy 0 rule 0 2662 Flow rule #0 destroyed 2663 testpmd> 2664 2665Querying flow rules 2666~~~~~~~~~~~~~~~~~~~ 2667 2668``flow query`` queries a specific action of a flow rule having that 2669ability. Such actions collect information that can be reported using this 2670command. It is bound to ``rte_flow_query()``:: 2671 2672 flow query {port_id} {rule_id} {action} 2673 2674If successful, it will display either the retrieved data for known actions 2675or the following message:: 2676 2677 Cannot display result for action type [...] ([...]) 2678 2679Otherwise, it will complain either that the rule does not exist or that some 2680error occurred:: 2681 2682 Flow rule #[...] not found 2683 2684:: 2685 2686 Caught error type [...] ([...]): [...] 2687 2688Currently only the ``count`` action is supported. This action reports the 2689number of packets that hit the flow rule and the total number of bytes. Its 2690output has the following format:: 2691 2692 count: 2693 hits_set: [...] # whether "hits" contains a valid value 2694 bytes_set: [...] # whether "bytes" contains a valid value 2695 hits: [...] # number of packets 2696 bytes: [...] # number of bytes 2697 2698Querying counters for TCPv6 packets redirected to queue 6:: 2699 2700 testpmd> flow create 0 ingress pattern eth / ipv6 / tcp / end 2701 actions queue index 6 / count / end 2702 Flow rule #4 created 2703 testpmd> flow query 0 4 count 2704 count: 2705 hits_set: 1 2706 bytes_set: 0 2707 hits: 386446 2708 bytes: 0 2709 testpmd> 2710 2711Listing flow rules 2712~~~~~~~~~~~~~~~~~~ 2713 2714``flow list`` lists existing flow rules sorted by priority and optionally 2715filtered by group identifiers:: 2716 2717 flow list {port_id} [group {group_id}] [...] 2718 2719This command only fails with the following message if the device does not 2720exist:: 2721 2722 Invalid port [...] 2723 2724Output consists of a header line followed by a short description of each 2725flow rule, one per line. There is no output at all when no flow rules are 2726configured on the device:: 2727 2728 ID Group Prio Attr Rule 2729 [...] [...] [...] [...] [...] 2730 2731``Attr`` column flags: 2732 2733- ``i`` for ``ingress``. 2734- ``e`` for ``egress``. 2735 2736Creating several flow rules and listing them:: 2737 2738 testpmd> flow create 0 ingress pattern eth / ipv4 / end 2739 actions queue index 6 / end 2740 Flow rule #0 created 2741 testpmd> flow create 0 ingress pattern eth / ipv6 / end 2742 actions queue index 2 / end 2743 Flow rule #1 created 2744 testpmd> flow create 0 priority 5 ingress pattern eth / ipv4 / udp / end 2745 actions rss queues 6 7 8 end / end 2746 Flow rule #2 created 2747 testpmd> flow list 0 2748 ID Group Prio Attr Rule 2749 0 0 0 i- ETH IPV4 => QUEUE 2750 1 0 0 i- ETH IPV6 => QUEUE 2751 2 0 5 i- ETH IPV4 UDP => RSS 2752 testpmd> 2753 2754Rules are sorted by priority (i.e. group ID first, then priority level):: 2755 2756 testpmd> flow list 1 2757 ID Group Prio Attr Rule 2758 0 0 0 i- ETH => COUNT 2759 6 0 500 i- ETH IPV6 TCP => DROP COUNT 2760 5 0 1000 i- ETH IPV6 ICMP => QUEUE 2761 1 24 0 i- ETH IPV4 UDP => QUEUE 2762 4 24 10 i- ETH IPV4 TCP => DROP 2763 3 24 20 i- ETH IPV4 => DROP 2764 2 24 42 i- ETH IPV4 UDP => QUEUE 2765 7 63 0 i- ETH IPV6 UDP VXLAN => MARK QUEUE 2766 testpmd> 2767 2768Output can be limited to specific groups:: 2769 2770 testpmd> flow list 1 group 0 group 63 2771 ID Group Prio Attr Rule 2772 0 0 0 i- ETH => COUNT 2773 6 0 500 i- ETH IPV6 TCP => DROP COUNT 2774 5 0 1000 i- ETH IPV6 ICMP => QUEUE 2775 7 63 0 i- ETH IPV6 UDP VXLAN => MARK QUEUE 2776 testpmd> 2777