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 flow_ctrl rx 904~~~~~~~~~~~~~~~~ 905 906Set the link flow control parameter on a port:: 907 908 testpmd> set flow_ctrl rx (on|off) tx (on|off) (high_water) (low_water) \ 909 (pause_time) (send_xon) mac_ctrl_frame_fwd (on|off) \ 910 autoneg (on|off) (port_id) 911 912Where: 913 914* ``high_water`` (integer): High threshold value to trigger XOFF. 915 916* ``low_water`` (integer): Low threshold value to trigger XON. 917 918* ``pause_time`` (integer): Pause quota in the Pause frame. 919 920* ``send_xon`` (0/1): Send XON frame. 921 922* ``mac_ctrl_frame_fwd``: Enable receiving MAC control frames. 923 924* ``autoneg``: Change the auto-negotiation parameter. 925 926set pfc_ctrl rx 927~~~~~~~~~~~~~~~ 928 929Set the priority flow control parameter on a port:: 930 931 testpmd> set pfc_ctrl rx (on|off) tx (on|off) (high_water) (low_water) \ 932 (pause_time) (priority) (port_id) 933 934Where: 935 936* ``high_water`` (integer): High threshold value. 937 938* ``low_water`` (integer): Low threshold value. 939 940* ``pause_time`` (integer): Pause quota in the Pause frame. 941 942* ``priority`` (0-7): VLAN User Priority. 943 944set stat_qmap 945~~~~~~~~~~~~~ 946 947Set statistics mapping (qmapping 0..15) for RX/TX queue on port:: 948 949 testpmd> set stat_qmap (tx|rx) (port_id) (queue_id) (qmapping) 950 951For example, to set rx queue 2 on port 0 to mapping 5:: 952 953 testpmd>set stat_qmap rx 0 2 5 954 955set port - rx/tx (for VF) 956~~~~~~~~~~~~~~~~~~~~~~~~~ 957 958Set VF receive/transmit from a port:: 959 960 testpmd> set port (port_id) vf (vf_id) (rx|tx) (on|off) 961 962set port - mac address filter (for VF) 963~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ 964 965Add/Remove unicast or multicast MAC addr filter for a VF:: 966 967 testpmd> set port (port_id) vf (vf_id) (mac_addr) \ 968 (exact-mac|exact-mac-vlan|hashmac|hashmac-vlan) (on|off) 969 970set port - rx mode(for VF) 971~~~~~~~~~~~~~~~~~~~~~~~~~~ 972 973Set the VF receive mode of a port:: 974 975 testpmd> set port (port_id) vf (vf_id) \ 976 rxmode (AUPE|ROPE|BAM|MPE) (on|off) 977 978The available receive modes are: 979 980* ``AUPE``: Accepts untagged VLAN. 981 982* ``ROPE``: Accepts unicast hash. 983 984* ``BAM``: Accepts broadcast packets. 985 986* ``MPE``: Accepts all multicast packets. 987 988set port - tx_rate (for Queue) 989~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ 990 991Set TX rate limitation for a queue on a port:: 992 993 testpmd> set port (port_id) queue (queue_id) rate (rate_value) 994 995set port - tx_rate (for VF) 996~~~~~~~~~~~~~~~~~~~~~~~~~~~ 997 998Set TX rate limitation for queues in VF on a port:: 999 1000 testpmd> set port (port_id) vf (vf_id) rate (rate_value) queue_mask (queue_mask) 1001 1002set port - mirror rule 1003~~~~~~~~~~~~~~~~~~~~~~ 1004 1005Set pool or vlan type mirror rule for a port:: 1006 1007 testpmd> set port (port_id) mirror-rule (rule_id) \ 1008 (pool-mirror-up|pool-mirror-down|vlan-mirror) \ 1009 (poolmask|vlanid[,vlanid]*) dst-pool (pool_id) (on|off) 1010 1011Set link mirror rule for a port:: 1012 1013 testpmd> set port (port_id) mirror-rule (rule_id) \ 1014 (uplink-mirror|downlink-mirror) dst-pool (pool_id) (on|off) 1015 1016For example to enable mirror traffic with vlan 0,1 to pool 0:: 1017 1018 set port 0 mirror-rule 0 vlan-mirror 0,1 dst-pool 0 on 1019 1020reset port - mirror rule 1021~~~~~~~~~~~~~~~~~~~~~~~~ 1022 1023Reset a mirror rule for a port:: 1024 1025 testpmd> reset port (port_id) mirror-rule (rule_id) 1026 1027set flush_rx 1028~~~~~~~~~~~~ 1029 1030Set the flush on RX streams before forwarding. 1031The default is flush ``on``. 1032Mainly used with PCAP drivers to turn off the default behavior of flushing the first 512 packets on RX streams:: 1033 1034 testpmd> set flush_rx off 1035 1036set bypass mode 1037~~~~~~~~~~~~~~~ 1038 1039Set the bypass mode for the lowest port on bypass enabled NIC:: 1040 1041 testpmd> set bypass mode (normal|bypass|isolate) (port_id) 1042 1043set bypass event 1044~~~~~~~~~~~~~~~~ 1045 1046Set the event required to initiate specified bypass mode for the lowest port on a bypass enabled:: 1047 1048 testpmd> set bypass event (timeout|os_on|os_off|power_on|power_off) \ 1049 mode (normal|bypass|isolate) (port_id) 1050 1051Where: 1052 1053* ``timeout``: Enable bypass after watchdog timeout. 1054 1055* ``os_on``: Enable bypass when OS/board is powered on. 1056 1057* ``os_off``: Enable bypass when OS/board is powered off. 1058 1059* ``power_on``: Enable bypass when power supply is turned on. 1060 1061* ``power_off``: Enable bypass when power supply is turned off. 1062 1063 1064set bypass timeout 1065~~~~~~~~~~~~~~~~~~ 1066 1067Set the bypass watchdog timeout to ``n`` seconds where 0 = instant:: 1068 1069 testpmd> set bypass timeout (0|1.5|2|3|4|8|16|32) 1070 1071show bypass config 1072~~~~~~~~~~~~~~~~~~ 1073 1074Show the bypass configuration for a bypass enabled NIC using the lowest port on the NIC:: 1075 1076 testpmd> show bypass config (port_id) 1077 1078set link up 1079~~~~~~~~~~~ 1080 1081Set link up for a port:: 1082 1083 testpmd> set link-up port (port id) 1084 1085set link down 1086~~~~~~~~~~~~~ 1087 1088Set link down for a port:: 1089 1090 testpmd> set link-down port (port id) 1091 1092E-tag set 1093~~~~~~~~~ 1094 1095Enable E-tag insertion for a VF on a port:: 1096 1097 testpmd> E-tag set insertion on port-tag-id (value) port (port_id) vf (vf_id) 1098 1099Disable E-tag insertion for a VF on a port:: 1100 1101 testpmd> E-tag set insertion off port (port_id) vf (vf_id) 1102 1103Enable/disable E-tag stripping on a port:: 1104 1105 testpmd> E-tag set stripping (on|off) port (port_id) 1106 1107Enable/disable E-tag based forwarding on a port:: 1108 1109 testpmd> E-tag set forwarding (on|off) port (port_id) 1110 1111Add an E-tag forwarding filter on a port:: 1112 1113 testpmd> E-tag set filter add e-tag-id (value) dst-pool (pool_id) port (port_id) 1114 1115Delete an E-tag forwarding filter on a port:: 1116 testpmd> E-tag set filter del e-tag-id (value) port (port_id) 1117 1118 1119Port Functions 1120-------------- 1121 1122The following sections show functions for configuring ports. 1123 1124.. note:: 1125 1126 Port configuration changes only become active when forwarding is started/restarted. 1127 1128port attach 1129~~~~~~~~~~~ 1130 1131Attach a port specified by pci address or virtual device args:: 1132 1133 testpmd> port attach (identifier) 1134 1135To attach a new pci device, the device should be recognized by kernel first. 1136Then it should be moved under DPDK management. 1137Finally the port can be attached to testpmd. 1138 1139For example, to move a pci device using ixgbe under DPDK management: 1140 1141.. code-block:: console 1142 1143 # Check the status of the available devices. 1144 ./usertools/dpdk-devbind.py --status 1145 1146 Network devices using DPDK-compatible driver 1147 ============================================ 1148 <none> 1149 1150 Network devices using kernel driver 1151 =================================== 1152 0000:0a:00.0 '82599ES 10-Gigabit' if=eth2 drv=ixgbe unused= 1153 1154 1155 # Bind the device to igb_uio. 1156 sudo ./usertools/dpdk-devbind.py -b igb_uio 0000:0a:00.0 1157 1158 1159 # Recheck the status of the devices. 1160 ./usertools/dpdk-devbind.py --status 1161 Network devices using DPDK-compatible driver 1162 ============================================ 1163 0000:0a:00.0 '82599ES 10-Gigabit' drv=igb_uio unused= 1164 1165To attach a port created by virtual device, above steps are not needed. 1166 1167For example, to attach a port whose pci address is 0000:0a:00.0. 1168 1169.. code-block:: console 1170 1171 testpmd> port attach 0000:0a:00.0 1172 Attaching a new port... 1173 EAL: PCI device 0000:0a:00.0 on NUMA socket -1 1174 EAL: probe driver: 8086:10fb rte_ixgbe_pmd 1175 EAL: PCI memory mapped at 0x7f83bfa00000 1176 EAL: PCI memory mapped at 0x7f83bfa80000 1177 PMD: eth_ixgbe_dev_init(): MAC: 2, PHY: 18, SFP+: 5 1178 PMD: eth_ixgbe_dev_init(): port 0 vendorID=0x8086 deviceID=0x10fb 1179 Port 0 is attached. Now total ports is 1 1180 Done 1181 1182For example, to attach a port created by pcap PMD. 1183 1184.. code-block:: console 1185 1186 testpmd> port attach net_pcap0 1187 Attaching a new port... 1188 PMD: Initializing pmd_pcap for net_pcap0 1189 PMD: Creating pcap-backed ethdev on numa socket 0 1190 Port 0 is attached. Now total ports is 1 1191 Done 1192 1193In this case, identifier is ``net_pcap0``. 1194This identifier format is the same as ``--vdev`` format of DPDK applications. 1195 1196For example, to re-attach a bonded port which has been previously detached, 1197the mode and slave parameters must be given. 1198 1199.. code-block:: console 1200 1201 testpmd> port attach net_bond_0,mode=0,slave=1 1202 Attaching a new port... 1203 EAL: Initializing pmd_bond for net_bond_0 1204 EAL: Create bonded device net_bond_0 on port 0 in mode 0 on socket 0. 1205 Port 0 is attached. Now total ports is 1 1206 Done 1207 1208 1209port detach 1210~~~~~~~~~~~ 1211 1212Detach a specific port:: 1213 1214 testpmd> port detach (port_id) 1215 1216Before detaching a port, the port should be stopped and closed. 1217 1218For example, to detach a pci device port 0. 1219 1220.. code-block:: console 1221 1222 testpmd> port stop 0 1223 Stopping ports... 1224 Done 1225 testpmd> port close 0 1226 Closing ports... 1227 Done 1228 1229 testpmd> port detach 0 1230 Detaching a port... 1231 EAL: PCI device 0000:0a:00.0 on NUMA socket -1 1232 EAL: remove driver: 8086:10fb rte_ixgbe_pmd 1233 EAL: PCI memory unmapped at 0x7f83bfa00000 1234 EAL: PCI memory unmapped at 0x7f83bfa80000 1235 Done 1236 1237 1238For example, to detach a virtual device port 0. 1239 1240.. code-block:: console 1241 1242 testpmd> port stop 0 1243 Stopping ports... 1244 Done 1245 testpmd> port close 0 1246 Closing ports... 1247 Done 1248 1249 testpmd> port detach 0 1250 Detaching a port... 1251 PMD: Closing pcap ethdev on numa socket 0 1252 Port 'net_pcap0' is detached. Now total ports is 0 1253 Done 1254 1255To remove a pci device completely from the system, first detach the port from testpmd. 1256Then the device should be moved under kernel management. 1257Finally the device can be removed using kernel pci hotplug functionality. 1258 1259For example, to move a pci device under kernel management: 1260 1261.. code-block:: console 1262 1263 sudo ./usertools/dpdk-devbind.py -b ixgbe 0000:0a:00.0 1264 1265 ./usertools/dpdk-devbind.py --status 1266 1267 Network devices using DPDK-compatible driver 1268 ============================================ 1269 <none> 1270 1271 Network devices using kernel driver 1272 =================================== 1273 0000:0a:00.0 '82599ES 10-Gigabit' if=eth2 drv=ixgbe unused=igb_uio 1274 1275To remove a port created by a virtual device, above steps are not needed. 1276 1277port start 1278~~~~~~~~~~ 1279 1280Start all ports or a specific port:: 1281 1282 testpmd> port start (port_id|all) 1283 1284port stop 1285~~~~~~~~~ 1286 1287Stop all ports or a specific port:: 1288 1289 testpmd> port stop (port_id|all) 1290 1291port close 1292~~~~~~~~~~ 1293 1294Close all ports or a specific port:: 1295 1296 testpmd> port close (port_id|all) 1297 1298port start/stop queue 1299~~~~~~~~~~~~~~~~~~~~~ 1300 1301Start/stop a rx/tx queue on a specific port:: 1302 1303 testpmd> port (port_id) (rxq|txq) (queue_id) (start|stop) 1304 1305Only take effect when port is started. 1306 1307port config - speed 1308~~~~~~~~~~~~~~~~~~~ 1309 1310Set the speed and duplex mode for all ports or a specific port:: 1311 1312 testpmd> port config (port_id|all) speed (10|100|1000|10000|25000|40000|50000|100000|auto) \ 1313 duplex (half|full|auto) 1314 1315port config - queues/descriptors 1316~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ 1317 1318Set number of queues/descriptors for rxq, txq, rxd and txd:: 1319 1320 testpmd> port config all (rxq|txq|rxd|txd) (value) 1321 1322This is equivalent to the ``--rxq``, ``--txq``, ``--rxd`` and ``--txd`` command-line options. 1323 1324port config - max-pkt-len 1325~~~~~~~~~~~~~~~~~~~~~~~~~ 1326 1327Set the maximum packet length:: 1328 1329 testpmd> port config all max-pkt-len (value) 1330 1331This is equivalent to the ``--max-pkt-len`` command-line option. 1332 1333port config - CRC Strip 1334~~~~~~~~~~~~~~~~~~~~~~~ 1335 1336Set hardware CRC stripping on or off for all ports:: 1337 1338 testpmd> port config all crc-strip (on|off) 1339 1340CRC stripping is off by default. 1341 1342The ``on`` option is equivalent to the ``--crc-strip`` command-line option. 1343 1344port config - scatter 1345~~~~~~~~~~~~~~~~~~~~~~~ 1346 1347Set RX scatter mode on or off for all ports:: 1348 1349 testpmd> port config all scatter (on|off) 1350 1351RX scatter mode is off by default. 1352 1353The ``on`` option is equivalent to the ``--enable-scatter`` command-line option. 1354 1355port config - TX queue flags 1356~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ 1357 1358Set a hexadecimal bitmap of TX queue flags for all ports:: 1359 1360 testpmd> port config all txqflags value 1361 1362This command is equivalent to the ``--txqflags`` command-line option. 1363 1364port config - RX Checksum 1365~~~~~~~~~~~~~~~~~~~~~~~~~ 1366 1367Set hardware RX checksum offload to on or off for all ports:: 1368 1369 testpmd> port config all rx-cksum (on|off) 1370 1371Checksum offload is off by default. 1372 1373The ``on`` option is equivalent to the ``--enable-rx-cksum`` command-line option. 1374 1375port config - VLAN 1376~~~~~~~~~~~~~~~~~~ 1377 1378Set hardware VLAN on or off for all ports:: 1379 1380 testpmd> port config all hw-vlan (on|off) 1381 1382Hardware VLAN is on by default. 1383 1384The ``off`` option is equivalent to the ``--disable-hw-vlan`` command-line option. 1385 1386port config - VLAN filter 1387~~~~~~~~~~~~~~~~~~~~~~~~~ 1388 1389Set hardware VLAN filter on or off for all ports:: 1390 1391 testpmd> port config all hw-vlan-filter (on|off) 1392 1393Hardware VLAN filter is on by default. 1394 1395The ``off`` option is equivalent to the ``--disable-hw-vlan-filter`` command-line option. 1396 1397port config - VLAN strip 1398~~~~~~~~~~~~~~~~~~~~~~~~ 1399 1400Set hardware VLAN strip on or off for all ports:: 1401 1402 testpmd> port config all hw-vlan-strip (on|off) 1403 1404Hardware VLAN strip is on by default. 1405 1406The ``off`` option is equivalent to the ``--disable-hw-vlan-strip`` command-line option. 1407 1408port config - VLAN extend 1409~~~~~~~~~~~~~~~~~~~~~~~~~ 1410 1411Set hardware VLAN extend on or off for all ports:: 1412 1413 testpmd> port config all hw-vlan-extend (on|off) 1414 1415Hardware VLAN extend is off by default. 1416 1417The ``off`` option is equivalent to the ``--disable-hw-vlan-extend`` command-line option. 1418 1419port config - Drop Packets 1420~~~~~~~~~~~~~~~~~~~~~~~~~~ 1421 1422Set packet drop for packets with no descriptors on or off for all ports:: 1423 1424 testpmd> port config all drop-en (on|off) 1425 1426Packet dropping for packets with no descriptors is off by default. 1427 1428The ``on`` option is equivalent to the ``--enable-drop-en`` command-line option. 1429 1430port config - RSS 1431~~~~~~~~~~~~~~~~~ 1432 1433Set the RSS (Receive Side Scaling) mode on or off:: 1434 1435 testpmd> port config all rss (all|ip|tcp|udp|sctp|ether|port|vxlan|geneve|nvgre|none) 1436 1437RSS is on by default. 1438 1439The ``none`` option is equivalent to the ``--disable-rss`` command-line option. 1440 1441port config - RSS Reta 1442~~~~~~~~~~~~~~~~~~~~~~ 1443 1444Set the RSS (Receive Side Scaling) redirection table:: 1445 1446 testpmd> port config all rss reta (hash,queue)[,(hash,queue)] 1447 1448port config - DCB 1449~~~~~~~~~~~~~~~~~ 1450 1451Set the DCB mode for an individual port:: 1452 1453 testpmd> port config (port_id) dcb vt (on|off) (traffic_class) pfc (on|off) 1454 1455The traffic class should be 4 or 8. 1456 1457port config - Burst 1458~~~~~~~~~~~~~~~~~~~ 1459 1460Set the number of packets per burst:: 1461 1462 testpmd> port config all burst (value) 1463 1464This is equivalent to the ``--burst`` command-line option. 1465 1466port config - Threshold 1467~~~~~~~~~~~~~~~~~~~~~~~ 1468 1469Set thresholds for TX/RX queues:: 1470 1471 testpmd> port config all (threshold) (value) 1472 1473Where the threshold type can be: 1474 1475* ``txpt:`` Set the prefetch threshold register of the TX rings, 0 <= value <= 255. 1476 1477* ``txht:`` Set the host threshold register of the TX rings, 0 <= value <= 255. 1478 1479* ``txwt:`` Set the write-back threshold register of the TX rings, 0 <= value <= 255. 1480 1481* ``rxpt:`` Set the prefetch threshold register of the RX rings, 0 <= value <= 255. 1482 1483* ``rxht:`` Set the host threshold register of the RX rings, 0 <= value <= 255. 1484 1485* ``rxwt:`` Set the write-back threshold register of the RX rings, 0 <= value <= 255. 1486 1487* ``txfreet:`` Set the transmit free threshold of the TX rings, 0 <= value <= txd. 1488 1489* ``rxfreet:`` Set the transmit free threshold of the RX rings, 0 <= value <= rxd. 1490 1491* ``txrst:`` Set the transmit RS bit threshold of TX rings, 0 <= value <= txd. 1492 1493These threshold options are also available from the command-line. 1494 1495port config - E-tag 1496~~~~~~~~~~~~~~~~~~~ 1497 1498Set the value of ether-type for E-tag:: 1499 1500 testpmd> port config (port_id|all) l2-tunnel E-tag ether-type (value) 1501 1502Enable/disable the E-tag support:: 1503 1504 testpmd> port config (port_id|all) l2-tunnel E-tag (enable|disable) 1505 1506 1507Link Bonding Functions 1508---------------------- 1509 1510The Link Bonding functions make it possible to dynamically create and 1511manage link bonding devices from within testpmd interactive prompt. 1512 1513create bonded device 1514~~~~~~~~~~~~~~~~~~~~ 1515 1516Create a new bonding device:: 1517 1518 testpmd> create bonded device (mode) (socket) 1519 1520For example, to create a bonded device in mode 1 on socket 0:: 1521 1522 testpmd> create bonded 1 0 1523 created new bonded device (port X) 1524 1525add bonding slave 1526~~~~~~~~~~~~~~~~~ 1527 1528Adds Ethernet device to a Link Bonding device:: 1529 1530 testpmd> add bonding slave (slave id) (port id) 1531 1532For example, to add Ethernet device (port 6) to a Link Bonding device (port 10):: 1533 1534 testpmd> add bonding slave 6 10 1535 1536 1537remove bonding slave 1538~~~~~~~~~~~~~~~~~~~~ 1539 1540Removes an Ethernet slave device from a Link Bonding device:: 1541 1542 testpmd> remove bonding slave (slave id) (port id) 1543 1544For example, to remove Ethernet slave device (port 6) to a Link Bonding device (port 10):: 1545 1546 testpmd> remove bonding slave 6 10 1547 1548set bonding mode 1549~~~~~~~~~~~~~~~~ 1550 1551Set the Link Bonding mode of a Link Bonding device:: 1552 1553 testpmd> set bonding mode (value) (port id) 1554 1555For example, to set the bonding mode of a Link Bonding device (port 10) to broadcast (mode 3):: 1556 1557 testpmd> set bonding mode 3 10 1558 1559set bonding primary 1560~~~~~~~~~~~~~~~~~~~ 1561 1562Set an Ethernet slave device as the primary device on a Link Bonding device:: 1563 1564 testpmd> set bonding primary (slave id) (port id) 1565 1566For example, to set the Ethernet slave device (port 6) as the primary port of a Link Bonding device (port 10):: 1567 1568 testpmd> set bonding primary 6 10 1569 1570set bonding mac 1571~~~~~~~~~~~~~~~ 1572 1573Set the MAC address of a Link Bonding device:: 1574 1575 testpmd> set bonding mac (port id) (mac) 1576 1577For example, to set the MAC address of a Link Bonding device (port 10) to 00:00:00:00:00:01:: 1578 1579 testpmd> set bonding mac 10 00:00:00:00:00:01 1580 1581set bonding xmit_balance_policy 1582~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ 1583 1584Set the transmission policy for a Link Bonding device when it is in Balance XOR mode:: 1585 1586 testpmd> set bonding xmit_balance_policy (port_id) (l2|l23|l34) 1587 1588For example, set a Link Bonding device (port 10) to use a balance policy of layer 3+4 (IP addresses & UDP ports):: 1589 1590 testpmd> set bonding xmit_balance_policy 10 l34 1591 1592 1593set bonding mon_period 1594~~~~~~~~~~~~~~~~~~~~~~ 1595 1596Set the link status monitoring polling period in milliseconds for a bonding device. 1597 1598This adds support for PMD slave devices which do not support link status interrupts. 1599When the mon_period is set to a value greater than 0 then all PMD's which do not support 1600link status ISR will be queried every polling interval to check if their link status has changed:: 1601 1602 testpmd> set bonding mon_period (port_id) (value) 1603 1604For example, to set the link status monitoring polling period of bonded device (port 5) to 150ms:: 1605 1606 testpmd> set bonding mon_period 5 150 1607 1608 1609show bonding config 1610~~~~~~~~~~~~~~~~~~~ 1611 1612Show the current configuration of a Link Bonding device:: 1613 1614 testpmd> show bonding config (port id) 1615 1616For example, 1617to show the configuration a Link Bonding device (port 9) with 3 slave devices (1, 3, 4) 1618in balance mode with a transmission policy of layer 2+3:: 1619 1620 testpmd> show bonding config 9 1621 Bonding mode: 2 1622 Balance Xmit Policy: BALANCE_XMIT_POLICY_LAYER23 1623 Slaves (3): [1 3 4] 1624 Active Slaves (3): [1 3 4] 1625 Primary: [3] 1626 1627 1628Register Functions 1629------------------ 1630 1631The Register Functions can be used to read from and write to registers on the network card referenced by a port number. 1632This is mainly useful for debugging purposes. 1633Reference should be made to the appropriate datasheet for the network card for details on the register addresses 1634and fields that can be accessed. 1635 1636read reg 1637~~~~~~~~ 1638 1639Display the value of a port register:: 1640 1641 testpmd> read reg (port_id) (address) 1642 1643For example, to examine the Flow Director control register (FDIRCTL, 0x0000EE000) on an Intel 82599 10 GbE Controller:: 1644 1645 testpmd> read reg 0 0xEE00 1646 port 0 PCI register at offset 0xEE00: 0x4A060029 (1241907241) 1647 1648read regfield 1649~~~~~~~~~~~~~ 1650 1651Display a port register bit field:: 1652 1653 testpmd> read regfield (port_id) (address) (bit_x) (bit_y) 1654 1655For example, reading the lowest two bits from the register in the example above:: 1656 1657 testpmd> read regfield 0 0xEE00 0 1 1658 port 0 PCI register at offset 0xEE00: bits[0, 1]=0x1 (1) 1659 1660read regbit 1661~~~~~~~~~~~ 1662 1663Display a single port register bit:: 1664 1665 testpmd> read regbit (port_id) (address) (bit_x) 1666 1667For example, reading the lowest bit from the register in the example above:: 1668 1669 testpmd> read regbit 0 0xEE00 0 1670 port 0 PCI register at offset 0xEE00: bit 0=1 1671 1672write reg 1673~~~~~~~~~ 1674 1675Set the value of a port register:: 1676 1677 testpmd> write reg (port_id) (address) (value) 1678 1679For example, to clear a register:: 1680 1681 testpmd> write reg 0 0xEE00 0x0 1682 port 0 PCI register at offset 0xEE00: 0x00000000 (0) 1683 1684write regfield 1685~~~~~~~~~~~~~~ 1686 1687Set bit field of a port register:: 1688 1689 testpmd> write regfield (port_id) (address) (bit_x) (bit_y) (value) 1690 1691For example, writing to the register cleared in the example above:: 1692 1693 testpmd> write regfield 0 0xEE00 0 1 2 1694 port 0 PCI register at offset 0xEE00: 0x00000002 (2) 1695 1696write regbit 1697~~~~~~~~~~~~ 1698 1699Set single bit value of a port register:: 1700 1701 testpmd> write regbit (port_id) (address) (bit_x) (value) 1702 1703For example, to set the high bit in the register from the example above:: 1704 1705 testpmd> write regbit 0 0xEE00 31 1 1706 port 0 PCI register at offset 0xEE00: 0x8000000A (2147483658) 1707 1708 1709Filter Functions 1710---------------- 1711 1712This section details the available filter functions that are available. 1713 1714Note these functions interface the deprecated legacy filtering framework, 1715superseded by *rte_flow*. See `Flow rules management`_. 1716 1717ethertype_filter 1718~~~~~~~~~~~~~~~~~~~~ 1719 1720Add or delete a L2 Ethertype filter, which identify packets by their L2 Ethertype mainly assign them to a receive queue:: 1721 1722 ethertype_filter (port_id) (add|del) (mac_addr|mac_ignr) (mac_address) \ 1723 ethertype (ether_type) (drop|fwd) queue (queue_id) 1724 1725The available information parameters are: 1726 1727* ``port_id``: The port which the Ethertype filter assigned on. 1728 1729* ``mac_addr``: Compare destination mac address. 1730 1731* ``mac_ignr``: Ignore destination mac address match. 1732 1733* ``mac_address``: Destination mac address to match. 1734 1735* ``ether_type``: The EtherType value want to match, 1736 for example 0x0806 for ARP packet. 0x0800 (IPv4) and 0x86DD (IPv6) are invalid. 1737 1738* ``queue_id``: The receive queue associated with this EtherType filter. 1739 It is meaningless when deleting or dropping. 1740 1741Example, to add/remove an ethertype filter rule:: 1742 1743 testpmd> ethertype_filter 0 add mac_ignr 00:11:22:33:44:55 \ 1744 ethertype 0x0806 fwd queue 3 1745 1746 testpmd> ethertype_filter 0 del mac_ignr 00:11:22:33:44:55 \ 1747 ethertype 0x0806 fwd queue 3 1748 17492tuple_filter 1750~~~~~~~~~~~~~~~~~ 1751 1752Add or delete a 2-tuple filter, 1753which identifies packets by specific protocol and destination TCP/UDP port 1754and forwards packets into one of the receive queues:: 1755 1756 2tuple_filter (port_id) (add|del) dst_port (dst_port_value) \ 1757 protocol (protocol_value) mask (mask_value) \ 1758 tcp_flags (tcp_flags_value) priority (prio_value) \ 1759 queue (queue_id) 1760 1761The available information parameters are: 1762 1763* ``port_id``: The port which the 2-tuple filter assigned on. 1764 1765* ``dst_port_value``: Destination port in L4. 1766 1767* ``protocol_value``: IP L4 protocol. 1768 1769* ``mask_value``: Participates in the match or not by bit for field above, 1b means participate. 1770 1771* ``tcp_flags_value``: TCP control bits. The non-zero value is invalid, when the pro_value is not set to 0x06 (TCP). 1772 1773* ``prio_value``: Priority of this filter. 1774 1775* ``queue_id``: The receive queue associated with this 2-tuple filter. 1776 1777Example, to add/remove an 2tuple filter rule:: 1778 1779 testpmd> 2tuple_filter 0 add dst_port 32 protocol 0x06 mask 0x03 \ 1780 tcp_flags 0x02 priority 3 queue 3 1781 1782 testpmd> 2tuple_filter 0 del dst_port 32 protocol 0x06 mask 0x03 \ 1783 tcp_flags 0x02 priority 3 queue 3 1784 17855tuple_filter 1786~~~~~~~~~~~~~~~~~ 1787 1788Add or delete a 5-tuple filter, 1789which consists of a 5-tuple (protocol, source and destination IP addresses, source and destination TCP/UDP/SCTP port) 1790and routes packets into one of the receive queues:: 1791 1792 5tuple_filter (port_id) (add|del) dst_ip (dst_address) src_ip \ 1793 (src_address) dst_port (dst_port_value) \ 1794 src_port (src_port_value) protocol (protocol_value) \ 1795 mask (mask_value) tcp_flags (tcp_flags_value) \ 1796 priority (prio_value) queue (queue_id) 1797 1798The available information parameters are: 1799 1800* ``port_id``: The port which the 5-tuple filter assigned on. 1801 1802* ``dst_address``: Destination IP address. 1803 1804* ``src_address``: Source IP address. 1805 1806* ``dst_port_value``: TCP/UDP destination port. 1807 1808* ``src_port_value``: TCP/UDP source port. 1809 1810* ``protocol_value``: L4 protocol. 1811 1812* ``mask_value``: Participates in the match or not by bit for field above, 1b means participate 1813 1814* ``tcp_flags_value``: TCP control bits. The non-zero value is invalid, when the protocol_value is not set to 0x06 (TCP). 1815 1816* ``prio_value``: The priority of this filter. 1817 1818* ``queue_id``: The receive queue associated with this 5-tuple filter. 1819 1820Example, to add/remove an 5tuple filter rule:: 1821 1822 testpmd> 5tuple_filter 0 add dst_ip 2.2.2.5 src_ip 2.2.2.4 \ 1823 dst_port 64 src_port 32 protocol 0x06 mask 0x1F \ 1824 flags 0x0 priority 3 queue 3 1825 1826 testpmd> 5tuple_filter 0 del dst_ip 2.2.2.5 src_ip 2.2.2.4 \ 1827 dst_port 64 src_port 32 protocol 0x06 mask 0x1F \ 1828 flags 0x0 priority 3 queue 3 1829 1830syn_filter 1831~~~~~~~~~~ 1832 1833Using the SYN filter, TCP packets whose *SYN* flag is set can be forwarded to a separate queue:: 1834 1835 syn_filter (port_id) (add|del) priority (high|low) queue (queue_id) 1836 1837The available information parameters are: 1838 1839* ``port_id``: The port which the SYN filter assigned on. 1840 1841* ``high``: This SYN filter has higher priority than other filters. 1842 1843* ``low``: This SYN filter has lower priority than other filters. 1844 1845* ``queue_id``: The receive queue associated with this SYN filter 1846 1847Example:: 1848 1849 testpmd> syn_filter 0 add priority high queue 3 1850 1851flex_filter 1852~~~~~~~~~~~ 1853 1854With flex filter, packets can be recognized by any arbitrary pattern within the first 128 bytes of the packet 1855and routed into one of the receive queues:: 1856 1857 flex_filter (port_id) (add|del) len (len_value) bytes (bytes_value) \ 1858 mask (mask_value) priority (prio_value) queue (queue_id) 1859 1860The available information parameters are: 1861 1862* ``port_id``: The port which the Flex filter is assigned on. 1863 1864* ``len_value``: Filter length in bytes, no greater than 128. 1865 1866* ``bytes_value``: A string in hexadecimal, means the value the flex filter needs to match. 1867 1868* ``mask_value``: A string in hexadecimal, bit 1 means corresponding byte participates in the match. 1869 1870* ``prio_value``: The priority of this filter. 1871 1872* ``queue_id``: The receive queue associated with this Flex filter. 1873 1874Example:: 1875 1876 testpmd> flex_filter 0 add len 16 bytes 0x00000000000000000000000008060000 \ 1877 mask 000C priority 3 queue 3 1878 1879 testpmd> flex_filter 0 del len 16 bytes 0x00000000000000000000000008060000 \ 1880 mask 000C priority 3 queue 3 1881 1882 1883.. _testpmd_flow_director: 1884 1885flow_director_filter 1886~~~~~~~~~~~~~~~~~~~~ 1887 1888The Flow Director works in receive mode to identify specific flows or sets of flows and route them to specific queues. 1889 1890Four types of filtering are supported which are referred to as Perfect Match, Signature, Perfect-mac-vlan and 1891Perfect-tunnel filters, the match mode is set by the ``--pkt-filter-mode`` command-line parameter: 1892 1893* Perfect 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 IP flow. 1896 1897* Signature filters. 1898 The hardware checks a match between a hash-based signature of the masked fields of the received packet. 1899 1900* Perfect-mac-vlan match filters. 1901 The hardware checks a match between the masked fields of the received packets and the programmed filters. 1902 The masked fields are for MAC VLAN flow. 1903 1904* Perfect-tunnel match filters. 1905 The hardware checks a match between the masked fields of the received packets and the programmed filters. 1906 The masked fields are for tunnel flow. 1907 1908The Flow Director filters can match the different fields for different type of packet: flow type, specific input set 1909per flow type and the flexible payload. 1910 1911The Flow Director can also mask out parts of all of these fields so that filters 1912are only applied to certain fields or parts of the fields. 1913 1914Different NICs may have different capabilities, command show port fdir (port_id) can be used to acquire the information. 1915 1916# Commands to add flow director filters of different flow types:: 1917 1918 flow_director_filter (port_id) mode IP (add|del|update) \ 1919 flow (ipv4-other|ipv4-frag|ipv6-other|ipv6-frag) \ 1920 src (src_ip_address) dst (dst_ip_address) \ 1921 tos (tos_value) proto (proto_value) ttl (ttl_value) \ 1922 vlan (vlan_value) flexbytes (flexbytes_value) \ 1923 (drop|fwd) pf|vf(vf_id) queue (queue_id) \ 1924 fd_id (fd_id_value) 1925 1926 flow_director_filter (port_id) mode IP (add|del|update) \ 1927 flow (ipv4-tcp|ipv4-udp|ipv6-tcp|ipv6-udp) \ 1928 src (src_ip_address) (src_port) \ 1929 dst (dst_ip_address) (dst_port) \ 1930 tos (tos_value) ttl (ttl_value) \ 1931 vlan (vlan_value) flexbytes (flexbytes_value) \ 1932 (drop|fwd) queue pf|vf(vf_id) (queue_id) \ 1933 fd_id (fd_id_value) 1934 1935 flow_director_filter (port_id) mode IP (add|del|update) \ 1936 flow (ipv4-sctp|ipv6-sctp) \ 1937 src (src_ip_address) (src_port) \ 1938 dst (dst_ip_address) (dst_port) \ 1939 tos (tos_value) ttl (ttl_value) \ 1940 tag (verification_tag) vlan (vlan_value) \ 1941 flexbytes (flexbytes_value) (drop|fwd) \ 1942 pf|vf(vf_id) queue (queue_id) fd_id (fd_id_value) 1943 1944 flow_director_filter (port_id) mode IP (add|del|update) flow l2_payload \ 1945 ether (ethertype) flexbytes (flexbytes_value) \ 1946 (drop|fwd) pf|vf(vf_id) queue (queue_id) 1947 fd_id (fd_id_value) 1948 1949 flow_director_filter (port_id) mode MAC-VLAN (add|del|update) \ 1950 mac (mac_address) vlan (vlan_value) \ 1951 flexbytes (flexbytes_value) (drop|fwd) \ 1952 queue (queue_id) fd_id (fd_id_value) 1953 1954 flow_director_filter (port_id) mode Tunnel (add|del|update) \ 1955 mac (mac_address) vlan (vlan_value) \ 1956 tunnel (NVGRE|VxLAN) tunnel-id (tunnel_id_value) \ 1957 flexbytes (flexbytes_value) (drop|fwd) \ 1958 queue (queue_id) fd_id (fd_id_value) 1959 1960For example, to add an ipv4-udp flow type filter:: 1961 1962 testpmd> flow_director_filter 0 mode IP add flow ipv4-udp src 2.2.2.3 32 \ 1963 dst 2.2.2.5 33 tos 2 ttl 40 vlan 0x1 flexbytes (0x88,0x48) \ 1964 fwd pf queue 1 fd_id 1 1965 1966For example, add an ipv4-other flow type filter:: 1967 1968 testpmd> flow_director_filter 0 mode IP add flow ipv4-other src 2.2.2.3 \ 1969 dst 2.2.2.5 tos 2 proto 20 ttl 40 vlan 0x1 \ 1970 flexbytes (0x88,0x48) fwd pf queue 1 fd_id 1 1971 1972flush_flow_director 1973~~~~~~~~~~~~~~~~~~~ 1974 1975Flush all flow director filters on a device:: 1976 1977 testpmd> flush_flow_director (port_id) 1978 1979Example, to flush all flow director filter on port 0:: 1980 1981 testpmd> flush_flow_director 0 1982 1983flow_director_mask 1984~~~~~~~~~~~~~~~~~~ 1985 1986Set flow director's input masks:: 1987 1988 flow_director_mask (port_id) mode IP vlan (vlan_value) \ 1989 src_mask (ipv4_src) (ipv6_src) (src_port) \ 1990 dst_mask (ipv4_dst) (ipv6_dst) (dst_port) 1991 1992 flow_director_mask (port_id) mode MAC-VLAN vlan (vlan_value) 1993 1994 flow_director_mask (port_id) mode Tunnel vlan (vlan_value) \ 1995 mac (mac_value) tunnel-type (tunnel_type_value) \ 1996 tunnel-id (tunnel_id_value) 1997 1998Example, to set flow director mask on port 0:: 1999 2000 testpmd> flow_director_mask 0 mode IP vlan 0xefff \ 2001 src_mask 255.255.255.255 \ 2002 FFFF:FFFF:FFFF:FFFF:FFFF:FFFF:FFFF:FFFF 0xFFFF \ 2003 dst_mask 255.255.255.255 \ 2004 FFFF:FFFF:FFFF:FFFF:FFFF:FFFF:FFFF:FFFF 0xFFFF 2005 2006flow_director_flex_mask 2007~~~~~~~~~~~~~~~~~~~~~~~ 2008 2009set masks of flow director's flexible payload based on certain flow type:: 2010 2011 testpmd> flow_director_flex_mask (port_id) \ 2012 flow (none|ipv4-other|ipv4-frag|ipv4-tcp|ipv4-udp|ipv4-sctp| \ 2013 ipv6-other|ipv6-frag|ipv6-tcp|ipv6-udp|ipv6-sctp| \ 2014 l2_payload|all) (mask) 2015 2016Example, to set flow director's flex mask for all flow type on port 0:: 2017 2018 testpmd> flow_director_flex_mask 0 flow all \ 2019 (0xff,0xff,0,0,0,0,0,0,0,0,0,0,0,0,0,0) 2020 2021 2022flow_director_flex_payload 2023~~~~~~~~~~~~~~~~~~~~~~~~~~ 2024 2025Configure flexible payload selection:: 2026 2027 flow_director_flex_payload (port_id) (raw|l2|l3|l4) (config) 2028 2029For example, to select the first 16 bytes from the offset 4 (bytes) of packet's payload as flexible payload:: 2030 2031 testpmd> flow_director_flex_payload 0 l4 \ 2032 (4,5,6,7,8,9,10,11,12,13,14,15,16,17,18,19) 2033 2034get_sym_hash_ena_per_port 2035~~~~~~~~~~~~~~~~~~~~~~~~~ 2036 2037Get symmetric hash enable configuration per port:: 2038 2039 get_sym_hash_ena_per_port (port_id) 2040 2041For example, to get symmetric hash enable configuration of port 1:: 2042 2043 testpmd> get_sym_hash_ena_per_port 1 2044 2045set_sym_hash_ena_per_port 2046~~~~~~~~~~~~~~~~~~~~~~~~~ 2047 2048Set symmetric hash enable configuration per port to enable or disable:: 2049 2050 set_sym_hash_ena_per_port (port_id) (enable|disable) 2051 2052For example, to set symmetric hash enable configuration of port 1 to enable:: 2053 2054 testpmd> set_sym_hash_ena_per_port 1 enable 2055 2056get_hash_global_config 2057~~~~~~~~~~~~~~~~~~~~~~ 2058 2059Get the global configurations of hash filters:: 2060 2061 get_hash_global_config (port_id) 2062 2063For example, to get the global configurations of hash filters of port 1:: 2064 2065 testpmd> get_hash_global_config 1 2066 2067set_hash_global_config 2068~~~~~~~~~~~~~~~~~~~~~~ 2069 2070Set the global configurations of hash filters:: 2071 2072 set_hash_global_config (port_id) (toeplitz|simple_xor|default) \ 2073 (ipv4|ipv4-frag|ipv4-tcp|ipv4-udp|ipv4-sctp|ipv4-other|ipv6|ipv6-frag| \ 2074 ipv6-tcp|ipv6-udp|ipv6-sctp|ipv6-other|l2_payload) \ 2075 (enable|disable) 2076 2077For example, to enable simple_xor for flow type of ipv6 on port 2:: 2078 2079 testpmd> set_hash_global_config 2 simple_xor ipv6 enable 2080 2081set_hash_input_set 2082~~~~~~~~~~~~~~~~~~ 2083 2084Set the input set for hash:: 2085 2086 set_hash_input_set (port_id) (ipv4-frag|ipv4-tcp|ipv4-udp|ipv4-sctp| \ 2087 ipv4-other|ipv6-frag|ipv6-tcp|ipv6-udp|ipv6-sctp|ipv6-other| \ 2088 l2_payload) (ovlan|ivlan|src-ipv4|dst-ipv4|src-ipv6|dst-ipv6|ipv4-tos| \ 2089 ipv4-proto|ipv6-tc|ipv6-next-header|udp-src-port|udp-dst-port| \ 2090 tcp-src-port|tcp-dst-port|sctp-src-port|sctp-dst-port|sctp-veri-tag| \ 2091 udp-key|gre-key|fld-1st|fld-2nd|fld-3rd|fld-4th|fld-5th|fld-6th|fld-7th| \ 2092 fld-8th|none) (select|add) 2093 2094For example, to add source IP to hash input set for flow type of ipv4-udp on port 0:: 2095 2096 testpmd> set_hash_input_set 0 ipv4-udp src-ipv4 add 2097 2098set_fdir_input_set 2099~~~~~~~~~~~~~~~~~~ 2100 2101The Flow Director filters can match the different fields for different type of packet, i.e. specific input set 2102on per flow type and the flexible payload. This command can be used to change input set for each flow type. 2103 2104Set the input set for flow director:: 2105 2106 set_fdir_input_set (port_id) (ipv4-frag|ipv4-tcp|ipv4-udp|ipv4-sctp| \ 2107 ipv4-other|ipv6|ipv6-frag|ipv6-tcp|ipv6-udp|ipv6-sctp|ipv6-other| \ 2108 l2_payload) (ivlan|ethertype|src-ipv4|dst-ipv4|src-ipv6|dst-ipv6|ipv4-tos| \ 2109 ipv4-proto|ipv4-ttl|ipv6-tc|ipv6-next-header|ipv6-hop-limits| \ 2110 tudp-src-port|udp-dst-port|cp-src-port|tcp-dst-port|sctp-src-port| \ 2111 sctp-dst-port|sctp-veri-tag|none) (select|add) 2112 2113For example to add source IP to FD input set for flow type of ipv4-udp on port 0:: 2114 2115 testpmd> set_fdir_input_set 0 ipv4-udp src-ipv4 add 2116 2117global_config 2118~~~~~~~~~~~~~ 2119 2120Set different GRE key length for input set:: 2121 2122 global_config (port_id) gre-key-len (number in bytes) 2123 2124For example to set GRE key length for input set to 4 bytes on port 0:: 2125 2126 testpmd> global_config 0 gre-key-len 4 2127 2128 2129.. _testpmd_rte_flow: 2130 2131Flow rules management 2132--------------------- 2133 2134Control of the generic flow API (*rte_flow*) is fully exposed through the 2135``flow`` command (validation, creation, destruction and queries). 2136 2137Considering *rte_flow* overlaps with all `Filter Functions`_, using both 2138features simultaneously may cause undefined side-effects and is therefore 2139not recommended. 2140 2141``flow`` syntax 2142~~~~~~~~~~~~~~~ 2143 2144Because the ``flow`` command uses dynamic tokens to handle the large number 2145of possible flow rules combinations, its behavior differs slightly from 2146other commands, in particular: 2147 2148- Pressing *?* or the *<tab>* key displays contextual help for the current 2149 token, not that of the entire command. 2150 2151- Optional and repeated parameters are supported (provided they are listed 2152 in the contextual help). 2153 2154The first parameter stands for the operation mode. Possible operations and 2155their general syntax are described below. They are covered in detail in the 2156following sections. 2157 2158- Check whether a flow rule can be created:: 2159 2160 flow validate {port_id} 2161 [group {group_id}] [priority {level}] [ingress] [egress] 2162 pattern {item} [/ {item} [...]] / end 2163 actions {action} [/ {action} [...]] / end 2164 2165- Create a flow rule:: 2166 2167 flow create {port_id} 2168 [group {group_id}] [priority {level}] [ingress] [egress] 2169 pattern {item} [/ {item} [...]] / end 2170 actions {action} [/ {action} [...]] / end 2171 2172- Destroy specific flow rules:: 2173 2174 flow destroy {port_id} rule {rule_id} [...] 2175 2176- Destroy all flow rules:: 2177 2178 flow flush {port_id} 2179 2180- Query an existing flow rule:: 2181 2182 flow query {port_id} {rule_id} {action} 2183 2184- List existing flow rules sorted by priority, filtered by group 2185 identifiers:: 2186 2187 flow list {port_id} [group {group_id}] [...] 2188 2189Validating flow rules 2190~~~~~~~~~~~~~~~~~~~~~ 2191 2192``flow validate`` reports whether a flow rule would be accepted by the 2193underlying device in its current state but stops short of creating it. It is 2194bound to ``rte_flow_validate()``:: 2195 2196 flow validate {port_id} 2197 [group {group_id}] [priority {level}] [ingress] [egress] 2198 pattern {item} [/ {item} [...]] / end 2199 actions {action} [/ {action} [...]] / end 2200 2201If successful, it will show:: 2202 2203 Flow rule validated 2204 2205Otherwise it will show an error message of the form:: 2206 2207 Caught error type [...] ([...]): [...] 2208 2209This command uses the same parameters as ``flow create``, their format is 2210described in `Creating flow rules`_. 2211 2212Check whether redirecting any Ethernet packet received on port 0 to RX queue 2213index 6 is supported:: 2214 2215 testpmd> flow validate 0 ingress pattern eth / end 2216 actions queue index 6 / end 2217 Flow rule validated 2218 testpmd> 2219 2220Port 0 does not support TCPv6 rules:: 2221 2222 testpmd> flow validate 0 ingress pattern eth / ipv6 / tcp / end 2223 actions drop / end 2224 Caught error type 9 (specific pattern item): Invalid argument 2225 testpmd> 2226 2227Creating flow rules 2228~~~~~~~~~~~~~~~~~~~ 2229 2230``flow create`` validates and creates the specified flow rule. It is bound 2231to ``rte_flow_create()``:: 2232 2233 flow create {port_id} 2234 [group {group_id}] [priority {level}] [ingress] [egress] 2235 pattern {item} [/ {item} [...]] / end 2236 actions {action} [/ {action} [...]] / end 2237 2238If successful, it will return a flow rule ID usable with other commands:: 2239 2240 Flow rule #[...] created 2241 2242Otherwise it will show an error message of the form:: 2243 2244 Caught error type [...] ([...]): [...] 2245 2246Parameters describe in the following order: 2247 2248- Attributes (*group*, *priority*, *ingress*, *egress* tokens). 2249- A matching pattern, starting with the *pattern* token and terminated by an 2250 *end* pattern item. 2251- Actions, starting with the *actions* token and terminated by an *end* 2252 action. 2253 2254These translate directly to *rte_flow* objects provided as-is to the 2255underlying functions. 2256 2257The shortest valid definition only comprises mandatory tokens:: 2258 2259 testpmd> flow create 0 pattern end actions end 2260 2261Note that PMDs may refuse rules that essentially do nothing such as this 2262one. 2263 2264**All unspecified object values are automatically initialized to 0.** 2265 2266Attributes 2267^^^^^^^^^^ 2268 2269These tokens affect flow rule attributes (``struct rte_flow_attr``) and are 2270specified before the ``pattern`` token. 2271 2272- ``group {group id}``: priority group. 2273- ``priority {level}``: priority level within group. 2274- ``ingress``: rule applies to ingress traffic. 2275- ``egress``: rule applies to egress traffic. 2276 2277Each instance of an attribute specified several times overrides the previous 2278value as shown below (group 4 is used):: 2279 2280 testpmd> flow create 0 group 42 group 24 group 4 [...] 2281 2282Note that once enabled, ``ingress`` and ``egress`` cannot be disabled. 2283 2284While not specifying a direction is an error, some rules may allow both 2285simultaneously. 2286 2287Most rules affect RX therefore contain the ``ingress`` token:: 2288 2289 testpmd> flow create 0 ingress pattern [...] 2290 2291Matching pattern 2292^^^^^^^^^^^^^^^^ 2293 2294A matching pattern starts after the ``pattern`` token. It is made of pattern 2295items and is terminated by a mandatory ``end`` item. 2296 2297Items are named after their type (*RTE_FLOW_ITEM_TYPE_* from ``enum 2298rte_flow_item_type``). 2299 2300The ``/`` token is used as a separator between pattern items as shown 2301below:: 2302 2303 testpmd> flow create 0 ingress pattern eth / ipv4 / udp / end [...] 2304 2305Note that protocol items like these must be stacked from lowest to highest 2306layer to make sense. For instance, the following rule is either invalid or 2307unlikely to match any packet:: 2308 2309 testpmd> flow create 0 ingress pattern eth / udp / ipv4 / end [...] 2310 2311More information on these restrictions can be found in the *rte_flow* 2312documentation. 2313 2314Several items support additional specification structures, for example 2315``ipv4`` allows specifying source and destination addresses as follows:: 2316 2317 testpmd> flow create 0 ingress pattern eth / ipv4 src is 10.1.1.1 2318 dst is 10.2.0.0 / end [...] 2319 2320This rule matches all IPv4 traffic with the specified properties. 2321 2322In this example, ``src`` and ``dst`` are field names of the underlying 2323``struct rte_flow_item_ipv4`` object. All item properties can be specified 2324in a similar fashion. 2325 2326The ``is`` token means that the subsequent value must be matched exactly, 2327and assigns ``spec`` and ``mask`` fields in ``struct rte_flow_item`` 2328accordingly. Possible assignment tokens are: 2329 2330- ``is``: match value perfectly (with full bit-mask). 2331- ``spec``: match value according to configured bit-mask. 2332- ``last``: specify upper bound to establish a range. 2333- ``mask``: specify bit-mask with relevant bits set to one. 2334- ``prefix``: generate bit-mask from a prefix length. 2335 2336These yield identical results:: 2337 2338 ipv4 src is 10.1.1.1 2339 2340:: 2341 2342 ipv4 src spec 10.1.1.1 src mask 255.255.255.255 2343 2344:: 2345 2346 ipv4 src spec 10.1.1.1 src prefix 32 2347 2348:: 2349 2350 ipv4 src is 10.1.1.1 src last 10.1.1.1 # range with a single value 2351 2352:: 2353 2354 ipv4 src is 10.1.1.1 src last 0 # 0 disables range 2355 2356Inclusive ranges can be defined with ``last``:: 2357 2358 ipv4 src is 10.1.1.1 src last 10.2.3.4 # 10.1.1.1 to 10.2.3.4 2359 2360Note that ``mask`` affects both ``spec`` and ``last``:: 2361 2362 ipv4 src is 10.1.1.1 src last 10.2.3.4 src mask 255.255.0.0 2363 # matches 10.1.0.0 to 10.2.255.255 2364 2365Properties can be modified multiple times:: 2366 2367 ipv4 src is 10.1.1.1 src is 10.1.2.3 src is 10.2.3.4 # matches 10.2.3.4 2368 2369:: 2370 2371 ipv4 src is 10.1.1.1 src prefix 24 src prefix 16 # matches 10.1.0.0/16 2372 2373Pattern items 2374^^^^^^^^^^^^^ 2375 2376This section lists supported pattern items and their attributes, if any. 2377 2378- ``end``: end list of pattern items. 2379 2380- ``void``: no-op pattern item. 2381 2382- ``invert``: perform actions when pattern does not match. 2383 2384- ``any``: match any protocol for the current layer. 2385 2386 - ``num {unsigned}``: number of layers covered. 2387 2388- ``pf``: match packets addressed to the physical function. 2389 2390- ``vf``: match packets addressed to a virtual function ID. 2391 2392 - ``id {unsigned}``: destination VF ID. 2393 2394- ``port``: device-specific physical port index to use. 2395 2396 - ``index {unsigned}``: physical port index. 2397 2398- ``raw``: match an arbitrary byte string. 2399 2400 - ``relative {boolean}``: look for pattern after the previous item. 2401 - ``search {boolean}``: search pattern from offset (see also limit). 2402 - ``offset {integer}``: absolute or relative offset for pattern. 2403 - ``limit {unsigned}``: search area limit for start of pattern. 2404 - ``pattern {string}``: byte string to look for. 2405 2406- ``eth``: match Ethernet header. 2407 2408 - ``dst {MAC-48}``: destination MAC. 2409 - ``src {MAC-48}``: source MAC. 2410 - ``type {unsigned}``: EtherType. 2411 2412- ``vlan``: match 802.1Q/ad VLAN tag. 2413 2414 - ``tpid {unsigned}``: tag protocol identifier. 2415 - ``tci {unsigned}``: tag control information. 2416 - ``pcp {unsigned}``: priority code point. 2417 - ``dei {unsigned}``: drop eligible indicator. 2418 - ``vid {unsigned}``: VLAN identifier. 2419 2420- ``ipv4``: match IPv4 header. 2421 2422 - ``tos {unsigned}``: type of service. 2423 - ``ttl {unsigned}``: time to live. 2424 - ``proto {unsigned}``: next protocol ID. 2425 - ``src {ipv4 address}``: source address. 2426 - ``dst {ipv4 address}``: destination address. 2427 2428- ``ipv6``: match IPv6 header. 2429 2430 - ``tc {unsigned}``: traffic class. 2431 - ``flow {unsigned}``: flow label. 2432 - ``proto {unsigned}``: protocol (next header). 2433 - ``hop {unsigned}``: hop limit. 2434 - ``src {ipv6 address}``: source address. 2435 - ``dst {ipv6 address}``: destination address. 2436 2437- ``icmp``: match ICMP header. 2438 2439 - ``type {unsigned}``: ICMP packet type. 2440 - ``code {unsigned}``: ICMP packet code. 2441 2442- ``udp``: match UDP header. 2443 2444 - ``src {unsigned}``: UDP source port. 2445 - ``dst {unsigned}``: UDP destination port. 2446 2447- ``tcp``: match TCP header. 2448 2449 - ``src {unsigned}``: TCP source port. 2450 - ``dst {unsigned}``: TCP destination port. 2451 2452- ``sctp``: match SCTP header. 2453 2454 - ``src {unsigned}``: SCTP source port. 2455 - ``dst {unsigned}``: SCTP destination port. 2456 - ``tag {unsigned}``: validation tag. 2457 - ``cksum {unsigned}``: checksum. 2458 2459- ``vxlan``: match VXLAN header. 2460 2461 - ``vni {unsigned}``: VXLAN identifier. 2462 2463Actions list 2464^^^^^^^^^^^^ 2465 2466A list of actions starts after the ``actions`` token in the same fashion as 2467`Matching pattern`_; actions are separated by ``/`` tokens and the list is 2468terminated by a mandatory ``end`` action. 2469 2470Actions are named after their type (*RTE_FLOW_ACTION_TYPE_* from ``enum 2471rte_flow_action_type``). 2472 2473Dropping all incoming UDPv4 packets can be expressed as follows:: 2474 2475 testpmd> flow create 0 ingress pattern eth / ipv4 / udp / end 2476 actions drop / end 2477 2478Several actions have configurable properties which must be specified when 2479there is no valid default value. For example, ``queue`` requires a target 2480queue index. 2481 2482This rule redirects incoming UDPv4 traffic to queue index 6:: 2483 2484 testpmd> flow create 0 ingress pattern eth / ipv4 / udp / end 2485 actions queue index 6 / end 2486 2487While this one could be rejected by PMDs (unspecified queue index):: 2488 2489 testpmd> flow create 0 ingress pattern eth / ipv4 / udp / end 2490 actions queue / end 2491 2492As defined by *rte_flow*, the list is not ordered, all actions of a given 2493rule are performed simultaneously. These are equivalent:: 2494 2495 queue index 6 / void / mark id 42 / end 2496 2497:: 2498 2499 void / mark id 42 / queue index 6 / end 2500 2501All actions in a list should have different types, otherwise only the last 2502action of a given type is taken into account:: 2503 2504 queue index 4 / queue index 5 / queue index 6 / end # will use queue 6 2505 2506:: 2507 2508 drop / drop / drop / end # drop is performed only once 2509 2510:: 2511 2512 mark id 42 / queue index 3 / mark id 24 / end # mark will be 24 2513 2514Considering they are performed simultaneously, opposite and overlapping 2515actions can sometimes be combined when the end result is unambiguous:: 2516 2517 drop / queue index 6 / end # drop has no effect 2518 2519:: 2520 2521 drop / dup index 6 / end # same as above 2522 2523:: 2524 2525 queue index 6 / rss queues 6 7 8 / end # queue has no effect 2526 2527:: 2528 2529 drop / passthru / end # drop has no effect 2530 2531Note that PMDs may still refuse such combinations. 2532 2533Actions 2534^^^^^^^ 2535 2536This section lists supported actions and their attributes, if any. 2537 2538- ``end``: end list of actions. 2539 2540- ``void``: no-op action. 2541 2542- ``passthru``: let subsequent rule process matched packets. 2543 2544- ``mark``: attach 32 bit value to packets. 2545 2546 - ``id {unsigned}``: 32 bit value to return with packets. 2547 2548- ``flag``: flag packets. 2549 2550- ``queue``: assign packets to a given queue index. 2551 2552 - ``index {unsigned}``: queue index to use. 2553 2554- ``drop``: drop packets (note: passthru has priority). 2555 2556- ``count``: enable counters for this rule. 2557 2558- ``dup``: duplicate packets to a given queue index. 2559 2560 - ``index {unsigned}``: queue index to duplicate packets to. 2561 2562- ``rss``: spread packets among several queues. 2563 2564 - ``queues [{unsigned} [...]] end``: queue indices to use. 2565 2566- ``pf``: redirect packets to physical device function. 2567 2568- ``vf``: redirect packets to virtual device function. 2569 2570 - ``original {boolean}``: use original VF ID if possible. 2571 - ``id {unsigned}``: VF ID to redirect packets to. 2572 2573Destroying flow rules 2574~~~~~~~~~~~~~~~~~~~~~ 2575 2576``flow destroy`` destroys one or more rules from their rule ID (as returned 2577by ``flow create``), this command calls ``rte_flow_destroy()`` as many 2578times as necessary:: 2579 2580 flow destroy {port_id} rule {rule_id} [...] 2581 2582If successful, it will show:: 2583 2584 Flow rule #[...] destroyed 2585 2586It does not report anything for rule IDs that do not exist. The usual error 2587message is shown when a rule cannot be destroyed:: 2588 2589 Caught error type [...] ([...]): [...] 2590 2591``flow flush`` destroys all rules on a device and does not take extra 2592arguments. It is bound to ``rte_flow_flush()``:: 2593 2594 flow flush {port_id} 2595 2596Any errors are reported as above. 2597 2598Creating several rules and destroying them:: 2599 2600 testpmd> flow create 0 ingress pattern eth / ipv6 / end 2601 actions queue index 2 / end 2602 Flow rule #0 created 2603 testpmd> flow create 0 ingress pattern eth / ipv4 / end 2604 actions queue index 3 / end 2605 Flow rule #1 created 2606 testpmd> flow destroy 0 rule 0 rule 1 2607 Flow rule #1 destroyed 2608 Flow rule #0 destroyed 2609 testpmd> 2610 2611The same result can be achieved using ``flow flush``:: 2612 2613 testpmd> flow create 0 ingress pattern eth / ipv6 / end 2614 actions queue index 2 / end 2615 Flow rule #0 created 2616 testpmd> flow create 0 ingress pattern eth / ipv4 / end 2617 actions queue index 3 / end 2618 Flow rule #1 created 2619 testpmd> flow flush 0 2620 testpmd> 2621 2622Non-existent rule IDs are ignored:: 2623 2624 testpmd> flow create 0 ingress pattern eth / ipv6 / end 2625 actions queue index 2 / end 2626 Flow rule #0 created 2627 testpmd> flow create 0 ingress pattern eth / ipv4 / end 2628 actions queue index 3 / end 2629 Flow rule #1 created 2630 testpmd> flow destroy 0 rule 42 rule 10 rule 2 2631 testpmd> 2632 testpmd> flow destroy 0 rule 0 2633 Flow rule #0 destroyed 2634 testpmd> 2635 2636Querying flow rules 2637~~~~~~~~~~~~~~~~~~~ 2638 2639``flow query`` queries a specific action of a flow rule having that 2640ability. Such actions collect information that can be reported using this 2641command. It is bound to ``rte_flow_query()``:: 2642 2643 flow query {port_id} {rule_id} {action} 2644 2645If successful, it will display either the retrieved data for known actions 2646or the following message:: 2647 2648 Cannot display result for action type [...] ([...]) 2649 2650Otherwise, it will complain either that the rule does not exist or that some 2651error occurred:: 2652 2653 Flow rule #[...] not found 2654 2655:: 2656 2657 Caught error type [...] ([...]): [...] 2658 2659Currently only the ``count`` action is supported. This action reports the 2660number of packets that hit the flow rule and the total number of bytes. Its 2661output has the following format:: 2662 2663 count: 2664 hits_set: [...] # whether "hits" contains a valid value 2665 bytes_set: [...] # whether "bytes" contains a valid value 2666 hits: [...] # number of packets 2667 bytes: [...] # number of bytes 2668 2669Querying counters for TCPv6 packets redirected to queue 6:: 2670 2671 testpmd> flow create 0 ingress pattern eth / ipv6 / tcp / end 2672 actions queue index 6 / count / end 2673 Flow rule #4 created 2674 testpmd> flow query 0 4 count 2675 count: 2676 hits_set: 1 2677 bytes_set: 0 2678 hits: 386446 2679 bytes: 0 2680 testpmd> 2681 2682Listing flow rules 2683~~~~~~~~~~~~~~~~~~ 2684 2685``flow list`` lists existing flow rules sorted by priority and optionally 2686filtered by group identifiers:: 2687 2688 flow list {port_id} [group {group_id}] [...] 2689 2690This command only fails with the following message if the device does not 2691exist:: 2692 2693 Invalid port [...] 2694 2695Output consists of a header line followed by a short description of each 2696flow rule, one per line. There is no output at all when no flow rules are 2697configured on the device:: 2698 2699 ID Group Prio Attr Rule 2700 [...] [...] [...] [...] [...] 2701 2702``Attr`` column flags: 2703 2704- ``i`` for ``ingress``. 2705- ``e`` for ``egress``. 2706 2707Creating several flow rules and listing them:: 2708 2709 testpmd> flow create 0 ingress pattern eth / ipv4 / end 2710 actions queue index 6 / end 2711 Flow rule #0 created 2712 testpmd> flow create 0 ingress pattern eth / ipv6 / end 2713 actions queue index 2 / end 2714 Flow rule #1 created 2715 testpmd> flow create 0 priority 5 ingress pattern eth / ipv4 / udp / end 2716 actions rss queues 6 7 8 end / end 2717 Flow rule #2 created 2718 testpmd> flow list 0 2719 ID Group Prio Attr Rule 2720 0 0 0 i- ETH IPV4 => QUEUE 2721 1 0 0 i- ETH IPV6 => QUEUE 2722 2 0 5 i- ETH IPV4 UDP => RSS 2723 testpmd> 2724 2725Rules are sorted by priority (i.e. group ID first, then priority level):: 2726 2727 testpmd> flow list 1 2728 ID Group Prio Attr Rule 2729 0 0 0 i- ETH => COUNT 2730 6 0 500 i- ETH IPV6 TCP => DROP COUNT 2731 5 0 1000 i- ETH IPV6 ICMP => QUEUE 2732 1 24 0 i- ETH IPV4 UDP => QUEUE 2733 4 24 10 i- ETH IPV4 TCP => DROP 2734 3 24 20 i- ETH IPV4 => DROP 2735 2 24 42 i- ETH IPV4 UDP => QUEUE 2736 7 63 0 i- ETH IPV6 UDP VXLAN => MARK QUEUE 2737 testpmd> 2738 2739Output can be limited to specific groups:: 2740 2741 testpmd> flow list 1 group 0 group 63 2742 ID Group Prio Attr Rule 2743 0 0 0 i- ETH => COUNT 2744 6 0 500 i- ETH IPV6 TCP => DROP COUNT 2745 5 0 1000 i- ETH IPV6 ICMP => QUEUE 2746 7 63 0 i- ETH IPV6 UDP VXLAN => MARK QUEUE 2747 testpmd> 2748