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 flow_ctrl rx 911~~~~~~~~~~~~~~~~ 912 913Set the link flow control parameter on a port:: 914 915 testpmd> set flow_ctrl rx (on|off) tx (on|off) (high_water) (low_water) \ 916 (pause_time) (send_xon) mac_ctrl_frame_fwd (on|off) \ 917 autoneg (on|off) (port_id) 918 919Where: 920 921* ``high_water`` (integer): High threshold value to trigger XOFF. 922 923* ``low_water`` (integer): Low threshold value to trigger XON. 924 925* ``pause_time`` (integer): Pause quota in the Pause frame. 926 927* ``send_xon`` (0/1): Send XON frame. 928 929* ``mac_ctrl_frame_fwd``: Enable receiving MAC control frames. 930 931* ``autoneg``: Change the auto-negotiation parameter. 932 933set pfc_ctrl rx 934~~~~~~~~~~~~~~~ 935 936Set the priority flow control parameter on a port:: 937 938 testpmd> set pfc_ctrl rx (on|off) tx (on|off) (high_water) (low_water) \ 939 (pause_time) (priority) (port_id) 940 941Where: 942 943* ``high_water`` (integer): High threshold value. 944 945* ``low_water`` (integer): Low threshold value. 946 947* ``pause_time`` (integer): Pause quota in the Pause frame. 948 949* ``priority`` (0-7): VLAN User Priority. 950 951set stat_qmap 952~~~~~~~~~~~~~ 953 954Set statistics mapping (qmapping 0..15) for RX/TX queue on port:: 955 956 testpmd> set stat_qmap (tx|rx) (port_id) (queue_id) (qmapping) 957 958For example, to set rx queue 2 on port 0 to mapping 5:: 959 960 testpmd>set stat_qmap rx 0 2 5 961 962set port - rx/tx (for VF) 963~~~~~~~~~~~~~~~~~~~~~~~~~ 964 965Set VF receive/transmit from a port:: 966 967 testpmd> set port (port_id) vf (vf_id) (rx|tx) (on|off) 968 969set port - mac address filter (for VF) 970~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ 971 972Add/Remove unicast or multicast MAC addr filter for a VF:: 973 974 testpmd> set port (port_id) vf (vf_id) (mac_addr) \ 975 (exact-mac|exact-mac-vlan|hashmac|hashmac-vlan) (on|off) 976 977set port - rx mode(for VF) 978~~~~~~~~~~~~~~~~~~~~~~~~~~ 979 980Set the VF receive mode of a port:: 981 982 testpmd> set port (port_id) vf (vf_id) \ 983 rxmode (AUPE|ROPE|BAM|MPE) (on|off) 984 985The available receive modes are: 986 987* ``AUPE``: Accepts untagged VLAN. 988 989* ``ROPE``: Accepts unicast hash. 990 991* ``BAM``: Accepts broadcast packets. 992 993* ``MPE``: Accepts all multicast packets. 994 995set port - tx_rate (for Queue) 996~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ 997 998Set TX rate limitation for a queue on a port:: 999 1000 testpmd> set port (port_id) queue (queue_id) rate (rate_value) 1001 1002set port - tx_rate (for VF) 1003~~~~~~~~~~~~~~~~~~~~~~~~~~~ 1004 1005Set TX rate limitation for queues in VF on a port:: 1006 1007 testpmd> set port (port_id) vf (vf_id) rate (rate_value) queue_mask (queue_mask) 1008 1009set port - mirror rule 1010~~~~~~~~~~~~~~~~~~~~~~ 1011 1012Set pool or vlan type mirror rule for a port:: 1013 1014 testpmd> set port (port_id) mirror-rule (rule_id) \ 1015 (pool-mirror-up|pool-mirror-down|vlan-mirror) \ 1016 (poolmask|vlanid[,vlanid]*) dst-pool (pool_id) (on|off) 1017 1018Set link mirror rule for a port:: 1019 1020 testpmd> set port (port_id) mirror-rule (rule_id) \ 1021 (uplink-mirror|downlink-mirror) dst-pool (pool_id) (on|off) 1022 1023For example to enable mirror traffic with vlan 0,1 to pool 0:: 1024 1025 set port 0 mirror-rule 0 vlan-mirror 0,1 dst-pool 0 on 1026 1027reset port - mirror rule 1028~~~~~~~~~~~~~~~~~~~~~~~~ 1029 1030Reset a mirror rule for a port:: 1031 1032 testpmd> reset port (port_id) mirror-rule (rule_id) 1033 1034set flush_rx 1035~~~~~~~~~~~~ 1036 1037Set the flush on RX streams before forwarding. 1038The default is flush ``on``. 1039Mainly used with PCAP drivers to turn off the default behavior of flushing the first 512 packets on RX streams:: 1040 1041 testpmd> set flush_rx off 1042 1043set bypass mode 1044~~~~~~~~~~~~~~~ 1045 1046Set the bypass mode for the lowest port on bypass enabled NIC:: 1047 1048 testpmd> set bypass mode (normal|bypass|isolate) (port_id) 1049 1050set bypass event 1051~~~~~~~~~~~~~~~~ 1052 1053Set the event required to initiate specified bypass mode for the lowest port on a bypass enabled:: 1054 1055 testpmd> set bypass event (timeout|os_on|os_off|power_on|power_off) \ 1056 mode (normal|bypass|isolate) (port_id) 1057 1058Where: 1059 1060* ``timeout``: Enable bypass after watchdog timeout. 1061 1062* ``os_on``: Enable bypass when OS/board is powered on. 1063 1064* ``os_off``: Enable bypass when OS/board is powered off. 1065 1066* ``power_on``: Enable bypass when power supply is turned on. 1067 1068* ``power_off``: Enable bypass when power supply is turned off. 1069 1070 1071set bypass timeout 1072~~~~~~~~~~~~~~~~~~ 1073 1074Set the bypass watchdog timeout to ``n`` seconds where 0 = instant:: 1075 1076 testpmd> set bypass timeout (0|1.5|2|3|4|8|16|32) 1077 1078show bypass config 1079~~~~~~~~~~~~~~~~~~ 1080 1081Show the bypass configuration for a bypass enabled NIC using the lowest port on the NIC:: 1082 1083 testpmd> show bypass config (port_id) 1084 1085set link up 1086~~~~~~~~~~~ 1087 1088Set link up for a port:: 1089 1090 testpmd> set link-up port (port id) 1091 1092set link down 1093~~~~~~~~~~~~~ 1094 1095Set link down for a port:: 1096 1097 testpmd> set link-down port (port id) 1098 1099E-tag set 1100~~~~~~~~~ 1101 1102Enable E-tag insertion for a VF on a port:: 1103 1104 testpmd> E-tag set insertion on port-tag-id (value) port (port_id) vf (vf_id) 1105 1106Disable E-tag insertion for a VF on a port:: 1107 1108 testpmd> E-tag set insertion off port (port_id) vf (vf_id) 1109 1110Enable/disable E-tag stripping on a port:: 1111 1112 testpmd> E-tag set stripping (on|off) port (port_id) 1113 1114Enable/disable E-tag based forwarding on a port:: 1115 1116 testpmd> E-tag set forwarding (on|off) port (port_id) 1117 1118Add an E-tag forwarding filter on a port:: 1119 1120 testpmd> E-tag set filter add e-tag-id (value) dst-pool (pool_id) port (port_id) 1121 1122Delete an E-tag forwarding filter on a port:: 1123 testpmd> E-tag set filter del e-tag-id (value) port (port_id) 1124 1125 1126Port Functions 1127-------------- 1128 1129The following sections show functions for configuring ports. 1130 1131.. note:: 1132 1133 Port configuration changes only become active when forwarding is started/restarted. 1134 1135port attach 1136~~~~~~~~~~~ 1137 1138Attach a port specified by pci address or virtual device args:: 1139 1140 testpmd> port attach (identifier) 1141 1142To attach a new pci device, the device should be recognized by kernel first. 1143Then it should be moved under DPDK management. 1144Finally the port can be attached to testpmd. 1145 1146For example, to move a pci device using ixgbe under DPDK management: 1147 1148.. code-block:: console 1149 1150 # Check the status of the available devices. 1151 ./usertools/dpdk-devbind.py --status 1152 1153 Network devices using DPDK-compatible driver 1154 ============================================ 1155 <none> 1156 1157 Network devices using kernel driver 1158 =================================== 1159 0000:0a:00.0 '82599ES 10-Gigabit' if=eth2 drv=ixgbe unused= 1160 1161 1162 # Bind the device to igb_uio. 1163 sudo ./usertools/dpdk-devbind.py -b igb_uio 0000:0a:00.0 1164 1165 1166 # Recheck the status of the devices. 1167 ./usertools/dpdk-devbind.py --status 1168 Network devices using DPDK-compatible driver 1169 ============================================ 1170 0000:0a:00.0 '82599ES 10-Gigabit' drv=igb_uio unused= 1171 1172To attach a port created by virtual device, above steps are not needed. 1173 1174For example, to attach a port whose pci address is 0000:0a:00.0. 1175 1176.. code-block:: console 1177 1178 testpmd> port attach 0000:0a:00.0 1179 Attaching a new port... 1180 EAL: PCI device 0000:0a:00.0 on NUMA socket -1 1181 EAL: probe driver: 8086:10fb rte_ixgbe_pmd 1182 EAL: PCI memory mapped at 0x7f83bfa00000 1183 EAL: PCI memory mapped at 0x7f83bfa80000 1184 PMD: eth_ixgbe_dev_init(): MAC: 2, PHY: 18, SFP+: 5 1185 PMD: eth_ixgbe_dev_init(): port 0 vendorID=0x8086 deviceID=0x10fb 1186 Port 0 is attached. Now total ports is 1 1187 Done 1188 1189For example, to attach a port created by pcap PMD. 1190 1191.. code-block:: console 1192 1193 testpmd> port attach net_pcap0 1194 Attaching a new port... 1195 PMD: Initializing pmd_pcap for net_pcap0 1196 PMD: Creating pcap-backed ethdev on numa socket 0 1197 Port 0 is attached. Now total ports is 1 1198 Done 1199 1200In this case, identifier is ``net_pcap0``. 1201This identifier format is the same as ``--vdev`` format of DPDK applications. 1202 1203For example, to re-attach a bonded port which has been previously detached, 1204the mode and slave parameters must be given. 1205 1206.. code-block:: console 1207 1208 testpmd> port attach net_bond_0,mode=0,slave=1 1209 Attaching a new port... 1210 EAL: Initializing pmd_bond for net_bond_0 1211 EAL: Create bonded device net_bond_0 on port 0 in mode 0 on socket 0. 1212 Port 0 is attached. Now total ports is 1 1213 Done 1214 1215 1216port detach 1217~~~~~~~~~~~ 1218 1219Detach a specific port:: 1220 1221 testpmd> port detach (port_id) 1222 1223Before detaching a port, the port should be stopped and closed. 1224 1225For example, to detach a pci device port 0. 1226 1227.. code-block:: console 1228 1229 testpmd> port stop 0 1230 Stopping ports... 1231 Done 1232 testpmd> port close 0 1233 Closing ports... 1234 Done 1235 1236 testpmd> port detach 0 1237 Detaching a port... 1238 EAL: PCI device 0000:0a:00.0 on NUMA socket -1 1239 EAL: remove driver: 8086:10fb rte_ixgbe_pmd 1240 EAL: PCI memory unmapped at 0x7f83bfa00000 1241 EAL: PCI memory unmapped at 0x7f83bfa80000 1242 Done 1243 1244 1245For example, to detach a virtual device port 0. 1246 1247.. code-block:: console 1248 1249 testpmd> port stop 0 1250 Stopping ports... 1251 Done 1252 testpmd> port close 0 1253 Closing ports... 1254 Done 1255 1256 testpmd> port detach 0 1257 Detaching a port... 1258 PMD: Closing pcap ethdev on numa socket 0 1259 Port 'net_pcap0' is detached. Now total ports is 0 1260 Done 1261 1262To remove a pci device completely from the system, first detach the port from testpmd. 1263Then the device should be moved under kernel management. 1264Finally the device can be removed using kernel pci hotplug functionality. 1265 1266For example, to move a pci device under kernel management: 1267 1268.. code-block:: console 1269 1270 sudo ./usertools/dpdk-devbind.py -b ixgbe 0000:0a:00.0 1271 1272 ./usertools/dpdk-devbind.py --status 1273 1274 Network devices using DPDK-compatible driver 1275 ============================================ 1276 <none> 1277 1278 Network devices using kernel driver 1279 =================================== 1280 0000:0a:00.0 '82599ES 10-Gigabit' if=eth2 drv=ixgbe unused=igb_uio 1281 1282To remove a port created by a virtual device, above steps are not needed. 1283 1284port start 1285~~~~~~~~~~ 1286 1287Start all ports or a specific port:: 1288 1289 testpmd> port start (port_id|all) 1290 1291port stop 1292~~~~~~~~~ 1293 1294Stop all ports or a specific port:: 1295 1296 testpmd> port stop (port_id|all) 1297 1298port close 1299~~~~~~~~~~ 1300 1301Close all ports or a specific port:: 1302 1303 testpmd> port close (port_id|all) 1304 1305port start/stop queue 1306~~~~~~~~~~~~~~~~~~~~~ 1307 1308Start/stop a rx/tx queue on a specific port:: 1309 1310 testpmd> port (port_id) (rxq|txq) (queue_id) (start|stop) 1311 1312Only take effect when port is started. 1313 1314port config - speed 1315~~~~~~~~~~~~~~~~~~~ 1316 1317Set the speed and duplex mode for all ports or a specific port:: 1318 1319 testpmd> port config (port_id|all) speed (10|100|1000|10000|25000|40000|50000|100000|auto) \ 1320 duplex (half|full|auto) 1321 1322port config - queues/descriptors 1323~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ 1324 1325Set number of queues/descriptors for rxq, txq, rxd and txd:: 1326 1327 testpmd> port config all (rxq|txq|rxd|txd) (value) 1328 1329This is equivalent to the ``--rxq``, ``--txq``, ``--rxd`` and ``--txd`` command-line options. 1330 1331port config - max-pkt-len 1332~~~~~~~~~~~~~~~~~~~~~~~~~ 1333 1334Set the maximum packet length:: 1335 1336 testpmd> port config all max-pkt-len (value) 1337 1338This is equivalent to the ``--max-pkt-len`` command-line option. 1339 1340port config - CRC Strip 1341~~~~~~~~~~~~~~~~~~~~~~~ 1342 1343Set hardware CRC stripping on or off for all ports:: 1344 1345 testpmd> port config all crc-strip (on|off) 1346 1347CRC stripping is off by default. 1348 1349The ``on`` option is equivalent to the ``--crc-strip`` command-line option. 1350 1351port config - scatter 1352~~~~~~~~~~~~~~~~~~~~~~~ 1353 1354Set RX scatter mode on or off for all ports:: 1355 1356 testpmd> port config all scatter (on|off) 1357 1358RX scatter mode is off by default. 1359 1360The ``on`` option is equivalent to the ``--enable-scatter`` command-line option. 1361 1362port config - TX queue flags 1363~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ 1364 1365Set a hexadecimal bitmap of TX queue flags for all ports:: 1366 1367 testpmd> port config all txqflags value 1368 1369This command is equivalent to the ``--txqflags`` command-line option. 1370 1371port config - RX Checksum 1372~~~~~~~~~~~~~~~~~~~~~~~~~ 1373 1374Set hardware RX checksum offload to on or off for all ports:: 1375 1376 testpmd> port config all rx-cksum (on|off) 1377 1378Checksum offload is off by default. 1379 1380The ``on`` option is equivalent to the ``--enable-rx-cksum`` command-line option. 1381 1382port config - VLAN 1383~~~~~~~~~~~~~~~~~~ 1384 1385Set hardware VLAN on or off for all ports:: 1386 1387 testpmd> port config all hw-vlan (on|off) 1388 1389Hardware VLAN is on by default. 1390 1391The ``off`` option is equivalent to the ``--disable-hw-vlan`` command-line option. 1392 1393port config - VLAN filter 1394~~~~~~~~~~~~~~~~~~~~~~~~~ 1395 1396Set hardware VLAN filter on or off for all ports:: 1397 1398 testpmd> port config all hw-vlan-filter (on|off) 1399 1400Hardware VLAN filter is on by default. 1401 1402The ``off`` option is equivalent to the ``--disable-hw-vlan-filter`` command-line option. 1403 1404port config - VLAN strip 1405~~~~~~~~~~~~~~~~~~~~~~~~ 1406 1407Set hardware VLAN strip on or off for all ports:: 1408 1409 testpmd> port config all hw-vlan-strip (on|off) 1410 1411Hardware VLAN strip is on by default. 1412 1413The ``off`` option is equivalent to the ``--disable-hw-vlan-strip`` command-line option. 1414 1415port config - VLAN extend 1416~~~~~~~~~~~~~~~~~~~~~~~~~ 1417 1418Set hardware VLAN extend on or off for all ports:: 1419 1420 testpmd> port config all hw-vlan-extend (on|off) 1421 1422Hardware VLAN extend is off by default. 1423 1424The ``off`` option is equivalent to the ``--disable-hw-vlan-extend`` command-line option. 1425 1426port config - Drop Packets 1427~~~~~~~~~~~~~~~~~~~~~~~~~~ 1428 1429Set packet drop for packets with no descriptors on or off for all ports:: 1430 1431 testpmd> port config all drop-en (on|off) 1432 1433Packet dropping for packets with no descriptors is off by default. 1434 1435The ``on`` option is equivalent to the ``--enable-drop-en`` command-line option. 1436 1437port config - RSS 1438~~~~~~~~~~~~~~~~~ 1439 1440Set the RSS (Receive Side Scaling) mode on or off:: 1441 1442 testpmd> port config all rss (all|ip|tcp|udp|sctp|ether|port|vxlan|geneve|nvgre|none) 1443 1444RSS is on by default. 1445 1446The ``none`` option is equivalent to the ``--disable-rss`` command-line option. 1447 1448port config - RSS Reta 1449~~~~~~~~~~~~~~~~~~~~~~ 1450 1451Set the RSS (Receive Side Scaling) redirection table:: 1452 1453 testpmd> port config all rss reta (hash,queue)[,(hash,queue)] 1454 1455port config - DCB 1456~~~~~~~~~~~~~~~~~ 1457 1458Set the DCB mode for an individual port:: 1459 1460 testpmd> port config (port_id) dcb vt (on|off) (traffic_class) pfc (on|off) 1461 1462The traffic class should be 4 or 8. 1463 1464port config - Burst 1465~~~~~~~~~~~~~~~~~~~ 1466 1467Set the number of packets per burst:: 1468 1469 testpmd> port config all burst (value) 1470 1471This is equivalent to the ``--burst`` command-line option. 1472 1473port config - Threshold 1474~~~~~~~~~~~~~~~~~~~~~~~ 1475 1476Set thresholds for TX/RX queues:: 1477 1478 testpmd> port config all (threshold) (value) 1479 1480Where the threshold type can be: 1481 1482* ``txpt:`` Set the prefetch threshold register of the TX rings, 0 <= value <= 255. 1483 1484* ``txht:`` Set the host threshold register of the TX rings, 0 <= value <= 255. 1485 1486* ``txwt:`` Set the write-back threshold register of the TX rings, 0 <= value <= 255. 1487 1488* ``rxpt:`` Set the prefetch threshold register of the RX rings, 0 <= value <= 255. 1489 1490* ``rxht:`` Set the host threshold register of the RX rings, 0 <= value <= 255. 1491 1492* ``rxwt:`` Set the write-back threshold register of the RX rings, 0 <= value <= 255. 1493 1494* ``txfreet:`` Set the transmit free threshold of the TX rings, 0 <= value <= txd. 1495 1496* ``rxfreet:`` Set the transmit free threshold of the RX rings, 0 <= value <= rxd. 1497 1498* ``txrst:`` Set the transmit RS bit threshold of TX rings, 0 <= value <= txd. 1499 1500These threshold options are also available from the command-line. 1501 1502port config - E-tag 1503~~~~~~~~~~~~~~~~~~~ 1504 1505Set the value of ether-type for E-tag:: 1506 1507 testpmd> port config (port_id|all) l2-tunnel E-tag ether-type (value) 1508 1509Enable/disable the E-tag support:: 1510 1511 testpmd> port config (port_id|all) l2-tunnel E-tag (enable|disable) 1512 1513 1514Link Bonding Functions 1515---------------------- 1516 1517The Link Bonding functions make it possible to dynamically create and 1518manage link bonding devices from within testpmd interactive prompt. 1519 1520create bonded device 1521~~~~~~~~~~~~~~~~~~~~ 1522 1523Create a new bonding device:: 1524 1525 testpmd> create bonded device (mode) (socket) 1526 1527For example, to create a bonded device in mode 1 on socket 0:: 1528 1529 testpmd> create bonded 1 0 1530 created new bonded device (port X) 1531 1532add bonding slave 1533~~~~~~~~~~~~~~~~~ 1534 1535Adds Ethernet device to a Link Bonding device:: 1536 1537 testpmd> add bonding slave (slave id) (port id) 1538 1539For example, to add Ethernet device (port 6) to a Link Bonding device (port 10):: 1540 1541 testpmd> add bonding slave 6 10 1542 1543 1544remove bonding slave 1545~~~~~~~~~~~~~~~~~~~~ 1546 1547Removes an Ethernet slave device from a Link Bonding device:: 1548 1549 testpmd> remove bonding slave (slave id) (port id) 1550 1551For example, to remove Ethernet slave device (port 6) to a Link Bonding device (port 10):: 1552 1553 testpmd> remove bonding slave 6 10 1554 1555set bonding mode 1556~~~~~~~~~~~~~~~~ 1557 1558Set the Link Bonding mode of a Link Bonding device:: 1559 1560 testpmd> set bonding mode (value) (port id) 1561 1562For example, to set the bonding mode of a Link Bonding device (port 10) to broadcast (mode 3):: 1563 1564 testpmd> set bonding mode 3 10 1565 1566set bonding primary 1567~~~~~~~~~~~~~~~~~~~ 1568 1569Set an Ethernet slave device as the primary device on a Link Bonding device:: 1570 1571 testpmd> set bonding primary (slave id) (port id) 1572 1573For example, to set the Ethernet slave device (port 6) as the primary port of a Link Bonding device (port 10):: 1574 1575 testpmd> set bonding primary 6 10 1576 1577set bonding mac 1578~~~~~~~~~~~~~~~ 1579 1580Set the MAC address of a Link Bonding device:: 1581 1582 testpmd> set bonding mac (port id) (mac) 1583 1584For example, to set the MAC address of a Link Bonding device (port 10) to 00:00:00:00:00:01:: 1585 1586 testpmd> set bonding mac 10 00:00:00:00:00:01 1587 1588set bonding xmit_balance_policy 1589~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ 1590 1591Set the transmission policy for a Link Bonding device when it is in Balance XOR mode:: 1592 1593 testpmd> set bonding xmit_balance_policy (port_id) (l2|l23|l34) 1594 1595For example, set a Link Bonding device (port 10) to use a balance policy of layer 3+4 (IP addresses & UDP ports):: 1596 1597 testpmd> set bonding xmit_balance_policy 10 l34 1598 1599 1600set bonding mon_period 1601~~~~~~~~~~~~~~~~~~~~~~ 1602 1603Set the link status monitoring polling period in milliseconds for a bonding device. 1604 1605This adds support for PMD slave devices which do not support link status interrupts. 1606When the mon_period is set to a value greater than 0 then all PMD's which do not support 1607link status ISR will be queried every polling interval to check if their link status has changed:: 1608 1609 testpmd> set bonding mon_period (port_id) (value) 1610 1611For example, to set the link status monitoring polling period of bonded device (port 5) to 150ms:: 1612 1613 testpmd> set bonding mon_period 5 150 1614 1615 1616show bonding config 1617~~~~~~~~~~~~~~~~~~~ 1618 1619Show the current configuration of a Link Bonding device:: 1620 1621 testpmd> show bonding config (port id) 1622 1623For example, 1624to show the configuration a Link Bonding device (port 9) with 3 slave devices (1, 3, 4) 1625in balance mode with a transmission policy of layer 2+3:: 1626 1627 testpmd> show bonding config 9 1628 Bonding mode: 2 1629 Balance Xmit Policy: BALANCE_XMIT_POLICY_LAYER23 1630 Slaves (3): [1 3 4] 1631 Active Slaves (3): [1 3 4] 1632 Primary: [3] 1633 1634 1635Register Functions 1636------------------ 1637 1638The Register Functions can be used to read from and write to registers on the network card referenced by a port number. 1639This is mainly useful for debugging purposes. 1640Reference should be made to the appropriate datasheet for the network card for details on the register addresses 1641and fields that can be accessed. 1642 1643read reg 1644~~~~~~~~ 1645 1646Display the value of a port register:: 1647 1648 testpmd> read reg (port_id) (address) 1649 1650For example, to examine the Flow Director control register (FDIRCTL, 0x0000EE000) on an Intel 82599 10 GbE Controller:: 1651 1652 testpmd> read reg 0 0xEE00 1653 port 0 PCI register at offset 0xEE00: 0x4A060029 (1241907241) 1654 1655read regfield 1656~~~~~~~~~~~~~ 1657 1658Display a port register bit field:: 1659 1660 testpmd> read regfield (port_id) (address) (bit_x) (bit_y) 1661 1662For example, reading the lowest two bits from the register in the example above:: 1663 1664 testpmd> read regfield 0 0xEE00 0 1 1665 port 0 PCI register at offset 0xEE00: bits[0, 1]=0x1 (1) 1666 1667read regbit 1668~~~~~~~~~~~ 1669 1670Display a single port register bit:: 1671 1672 testpmd> read regbit (port_id) (address) (bit_x) 1673 1674For example, reading the lowest bit from the register in the example above:: 1675 1676 testpmd> read regbit 0 0xEE00 0 1677 port 0 PCI register at offset 0xEE00: bit 0=1 1678 1679write reg 1680~~~~~~~~~ 1681 1682Set the value of a port register:: 1683 1684 testpmd> write reg (port_id) (address) (value) 1685 1686For example, to clear a register:: 1687 1688 testpmd> write reg 0 0xEE00 0x0 1689 port 0 PCI register at offset 0xEE00: 0x00000000 (0) 1690 1691write regfield 1692~~~~~~~~~~~~~~ 1693 1694Set bit field of a port register:: 1695 1696 testpmd> write regfield (port_id) (address) (bit_x) (bit_y) (value) 1697 1698For example, writing to the register cleared in the example above:: 1699 1700 testpmd> write regfield 0 0xEE00 0 1 2 1701 port 0 PCI register at offset 0xEE00: 0x00000002 (2) 1702 1703write regbit 1704~~~~~~~~~~~~ 1705 1706Set single bit value of a port register:: 1707 1708 testpmd> write regbit (port_id) (address) (bit_x) (value) 1709 1710For example, to set the high bit in the register from the example above:: 1711 1712 testpmd> write regbit 0 0xEE00 31 1 1713 port 0 PCI register at offset 0xEE00: 0x8000000A (2147483658) 1714 1715 1716Filter Functions 1717---------------- 1718 1719This section details the available filter functions that are available. 1720 1721Note these functions interface the deprecated legacy filtering framework, 1722superseded by *rte_flow*. See `Flow rules management`_. 1723 1724ethertype_filter 1725~~~~~~~~~~~~~~~~~~~~ 1726 1727Add or delete a L2 Ethertype filter, which identify packets by their L2 Ethertype mainly assign them to a receive queue:: 1728 1729 ethertype_filter (port_id) (add|del) (mac_addr|mac_ignr) (mac_address) \ 1730 ethertype (ether_type) (drop|fwd) queue (queue_id) 1731 1732The available information parameters are: 1733 1734* ``port_id``: The port which the Ethertype filter assigned on. 1735 1736* ``mac_addr``: Compare destination mac address. 1737 1738* ``mac_ignr``: Ignore destination mac address match. 1739 1740* ``mac_address``: Destination mac address to match. 1741 1742* ``ether_type``: The EtherType value want to match, 1743 for example 0x0806 for ARP packet. 0x0800 (IPv4) and 0x86DD (IPv6) are invalid. 1744 1745* ``queue_id``: The receive queue associated with this EtherType filter. 1746 It is meaningless when deleting or dropping. 1747 1748Example, to add/remove an ethertype filter rule:: 1749 1750 testpmd> ethertype_filter 0 add mac_ignr 00:11:22:33:44:55 \ 1751 ethertype 0x0806 fwd queue 3 1752 1753 testpmd> ethertype_filter 0 del mac_ignr 00:11:22:33:44:55 \ 1754 ethertype 0x0806 fwd queue 3 1755 17562tuple_filter 1757~~~~~~~~~~~~~~~~~ 1758 1759Add or delete a 2-tuple filter, 1760which identifies packets by specific protocol and destination TCP/UDP port 1761and forwards packets into one of the receive queues:: 1762 1763 2tuple_filter (port_id) (add|del) dst_port (dst_port_value) \ 1764 protocol (protocol_value) mask (mask_value) \ 1765 tcp_flags (tcp_flags_value) priority (prio_value) \ 1766 queue (queue_id) 1767 1768The available information parameters are: 1769 1770* ``port_id``: The port which the 2-tuple filter assigned on. 1771 1772* ``dst_port_value``: Destination port in L4. 1773 1774* ``protocol_value``: IP L4 protocol. 1775 1776* ``mask_value``: Participates in the match or not by bit for field above, 1b means participate. 1777 1778* ``tcp_flags_value``: TCP control bits. The non-zero value is invalid, when the pro_value is not set to 0x06 (TCP). 1779 1780* ``prio_value``: Priority of this filter. 1781 1782* ``queue_id``: The receive queue associated with this 2-tuple filter. 1783 1784Example, to add/remove an 2tuple filter rule:: 1785 1786 testpmd> 2tuple_filter 0 add dst_port 32 protocol 0x06 mask 0x03 \ 1787 tcp_flags 0x02 priority 3 queue 3 1788 1789 testpmd> 2tuple_filter 0 del dst_port 32 protocol 0x06 mask 0x03 \ 1790 tcp_flags 0x02 priority 3 queue 3 1791 17925tuple_filter 1793~~~~~~~~~~~~~~~~~ 1794 1795Add or delete a 5-tuple filter, 1796which consists of a 5-tuple (protocol, source and destination IP addresses, source and destination TCP/UDP/SCTP port) 1797and routes packets into one of the receive queues:: 1798 1799 5tuple_filter (port_id) (add|del) dst_ip (dst_address) src_ip \ 1800 (src_address) dst_port (dst_port_value) \ 1801 src_port (src_port_value) protocol (protocol_value) \ 1802 mask (mask_value) tcp_flags (tcp_flags_value) \ 1803 priority (prio_value) queue (queue_id) 1804 1805The available information parameters are: 1806 1807* ``port_id``: The port which the 5-tuple filter assigned on. 1808 1809* ``dst_address``: Destination IP address. 1810 1811* ``src_address``: Source IP address. 1812 1813* ``dst_port_value``: TCP/UDP destination port. 1814 1815* ``src_port_value``: TCP/UDP source port. 1816 1817* ``protocol_value``: L4 protocol. 1818 1819* ``mask_value``: Participates in the match or not by bit for field above, 1b means participate 1820 1821* ``tcp_flags_value``: TCP control bits. The non-zero value is invalid, when the protocol_value is not set to 0x06 (TCP). 1822 1823* ``prio_value``: The priority of this filter. 1824 1825* ``queue_id``: The receive queue associated with this 5-tuple filter. 1826 1827Example, to add/remove an 5tuple filter rule:: 1828 1829 testpmd> 5tuple_filter 0 add dst_ip 2.2.2.5 src_ip 2.2.2.4 \ 1830 dst_port 64 src_port 32 protocol 0x06 mask 0x1F \ 1831 flags 0x0 priority 3 queue 3 1832 1833 testpmd> 5tuple_filter 0 del dst_ip 2.2.2.5 src_ip 2.2.2.4 \ 1834 dst_port 64 src_port 32 protocol 0x06 mask 0x1F \ 1835 flags 0x0 priority 3 queue 3 1836 1837syn_filter 1838~~~~~~~~~~ 1839 1840Using the SYN filter, TCP packets whose *SYN* flag is set can be forwarded to a separate queue:: 1841 1842 syn_filter (port_id) (add|del) priority (high|low) queue (queue_id) 1843 1844The available information parameters are: 1845 1846* ``port_id``: The port which the SYN filter assigned on. 1847 1848* ``high``: This SYN filter has higher priority than other filters. 1849 1850* ``low``: This SYN filter has lower priority than other filters. 1851 1852* ``queue_id``: The receive queue associated with this SYN filter 1853 1854Example:: 1855 1856 testpmd> syn_filter 0 add priority high queue 3 1857 1858flex_filter 1859~~~~~~~~~~~ 1860 1861With flex filter, packets can be recognized by any arbitrary pattern within the first 128 bytes of the packet 1862and routed into one of the receive queues:: 1863 1864 flex_filter (port_id) (add|del) len (len_value) bytes (bytes_value) \ 1865 mask (mask_value) priority (prio_value) queue (queue_id) 1866 1867The available information parameters are: 1868 1869* ``port_id``: The port which the Flex filter is assigned on. 1870 1871* ``len_value``: Filter length in bytes, no greater than 128. 1872 1873* ``bytes_value``: A string in hexadecimal, means the value the flex filter needs to match. 1874 1875* ``mask_value``: A string in hexadecimal, bit 1 means corresponding byte participates in the match. 1876 1877* ``prio_value``: The priority of this filter. 1878 1879* ``queue_id``: The receive queue associated with this Flex filter. 1880 1881Example:: 1882 1883 testpmd> flex_filter 0 add len 16 bytes 0x00000000000000000000000008060000 \ 1884 mask 000C priority 3 queue 3 1885 1886 testpmd> flex_filter 0 del len 16 bytes 0x00000000000000000000000008060000 \ 1887 mask 000C priority 3 queue 3 1888 1889 1890.. _testpmd_flow_director: 1891 1892flow_director_filter 1893~~~~~~~~~~~~~~~~~~~~ 1894 1895The Flow Director works in receive mode to identify specific flows or sets of flows and route them to specific queues. 1896 1897Four types of filtering are supported which are referred to as Perfect Match, Signature, Perfect-mac-vlan and 1898Perfect-tunnel filters, the match mode is set by the ``--pkt-filter-mode`` command-line parameter: 1899 1900* Perfect 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 IP flow. 1903 1904* Signature filters. 1905 The hardware checks a match between a hash-based signature of the masked fields of the received packet. 1906 1907* Perfect-mac-vlan match filters. 1908 The hardware checks a match between the masked fields of the received packets and the programmed filters. 1909 The masked fields are for MAC VLAN flow. 1910 1911* Perfect-tunnel match filters. 1912 The hardware checks a match between the masked fields of the received packets and the programmed filters. 1913 The masked fields are for tunnel flow. 1914 1915The Flow Director filters can match the different fields for different type of packet: flow type, specific input set 1916per flow type and the flexible payload. 1917 1918The Flow Director can also mask out parts of all of these fields so that filters 1919are only applied to certain fields or parts of the fields. 1920 1921Different NICs may have different capabilities, command show port fdir (port_id) can be used to acquire the information. 1922 1923# Commands to add flow director filters of different flow types:: 1924 1925 flow_director_filter (port_id) mode IP (add|del|update) \ 1926 flow (ipv4-other|ipv4-frag|ipv6-other|ipv6-frag) \ 1927 src (src_ip_address) dst (dst_ip_address) \ 1928 tos (tos_value) proto (proto_value) ttl (ttl_value) \ 1929 vlan (vlan_value) flexbytes (flexbytes_value) \ 1930 (drop|fwd) pf|vf(vf_id) queue (queue_id) \ 1931 fd_id (fd_id_value) 1932 1933 flow_director_filter (port_id) mode IP (add|del|update) \ 1934 flow (ipv4-tcp|ipv4-udp|ipv6-tcp|ipv6-udp) \ 1935 src (src_ip_address) (src_port) \ 1936 dst (dst_ip_address) (dst_port) \ 1937 tos (tos_value) ttl (ttl_value) \ 1938 vlan (vlan_value) flexbytes (flexbytes_value) \ 1939 (drop|fwd) queue pf|vf(vf_id) (queue_id) \ 1940 fd_id (fd_id_value) 1941 1942 flow_director_filter (port_id) mode IP (add|del|update) \ 1943 flow (ipv4-sctp|ipv6-sctp) \ 1944 src (src_ip_address) (src_port) \ 1945 dst (dst_ip_address) (dst_port) \ 1946 tos (tos_value) ttl (ttl_value) \ 1947 tag (verification_tag) vlan (vlan_value) \ 1948 flexbytes (flexbytes_value) (drop|fwd) \ 1949 pf|vf(vf_id) queue (queue_id) fd_id (fd_id_value) 1950 1951 flow_director_filter (port_id) mode IP (add|del|update) flow l2_payload \ 1952 ether (ethertype) flexbytes (flexbytes_value) \ 1953 (drop|fwd) pf|vf(vf_id) queue (queue_id) 1954 fd_id (fd_id_value) 1955 1956 flow_director_filter (port_id) mode MAC-VLAN (add|del|update) \ 1957 mac (mac_address) vlan (vlan_value) \ 1958 flexbytes (flexbytes_value) (drop|fwd) \ 1959 queue (queue_id) fd_id (fd_id_value) 1960 1961 flow_director_filter (port_id) mode Tunnel (add|del|update) \ 1962 mac (mac_address) vlan (vlan_value) \ 1963 tunnel (NVGRE|VxLAN) tunnel-id (tunnel_id_value) \ 1964 flexbytes (flexbytes_value) (drop|fwd) \ 1965 queue (queue_id) fd_id (fd_id_value) 1966 1967For example, to add an ipv4-udp flow type filter:: 1968 1969 testpmd> flow_director_filter 0 mode IP add flow ipv4-udp src 2.2.2.3 32 \ 1970 dst 2.2.2.5 33 tos 2 ttl 40 vlan 0x1 flexbytes (0x88,0x48) \ 1971 fwd pf queue 1 fd_id 1 1972 1973For example, add an ipv4-other flow type filter:: 1974 1975 testpmd> flow_director_filter 0 mode IP add flow ipv4-other src 2.2.2.3 \ 1976 dst 2.2.2.5 tos 2 proto 20 ttl 40 vlan 0x1 \ 1977 flexbytes (0x88,0x48) fwd pf queue 1 fd_id 1 1978 1979flush_flow_director 1980~~~~~~~~~~~~~~~~~~~ 1981 1982Flush all flow director filters on a device:: 1983 1984 testpmd> flush_flow_director (port_id) 1985 1986Example, to flush all flow director filter on port 0:: 1987 1988 testpmd> flush_flow_director 0 1989 1990flow_director_mask 1991~~~~~~~~~~~~~~~~~~ 1992 1993Set flow director's input masks:: 1994 1995 flow_director_mask (port_id) mode IP vlan (vlan_value) \ 1996 src_mask (ipv4_src) (ipv6_src) (src_port) \ 1997 dst_mask (ipv4_dst) (ipv6_dst) (dst_port) 1998 1999 flow_director_mask (port_id) mode MAC-VLAN vlan (vlan_value) 2000 2001 flow_director_mask (port_id) mode Tunnel vlan (vlan_value) \ 2002 mac (mac_value) tunnel-type (tunnel_type_value) \ 2003 tunnel-id (tunnel_id_value) 2004 2005Example, to set flow director mask on port 0:: 2006 2007 testpmd> flow_director_mask 0 mode IP vlan 0xefff \ 2008 src_mask 255.255.255.255 \ 2009 FFFF:FFFF:FFFF:FFFF:FFFF:FFFF:FFFF:FFFF 0xFFFF \ 2010 dst_mask 255.255.255.255 \ 2011 FFFF:FFFF:FFFF:FFFF:FFFF:FFFF:FFFF:FFFF 0xFFFF 2012 2013flow_director_flex_mask 2014~~~~~~~~~~~~~~~~~~~~~~~ 2015 2016set masks of flow director's flexible payload based on certain flow type:: 2017 2018 testpmd> flow_director_flex_mask (port_id) \ 2019 flow (none|ipv4-other|ipv4-frag|ipv4-tcp|ipv4-udp|ipv4-sctp| \ 2020 ipv6-other|ipv6-frag|ipv6-tcp|ipv6-udp|ipv6-sctp| \ 2021 l2_payload|all) (mask) 2022 2023Example, to set flow director's flex mask for all flow type on port 0:: 2024 2025 testpmd> flow_director_flex_mask 0 flow all \ 2026 (0xff,0xff,0,0,0,0,0,0,0,0,0,0,0,0,0,0) 2027 2028 2029flow_director_flex_payload 2030~~~~~~~~~~~~~~~~~~~~~~~~~~ 2031 2032Configure flexible payload selection:: 2033 2034 flow_director_flex_payload (port_id) (raw|l2|l3|l4) (config) 2035 2036For example, to select the first 16 bytes from the offset 4 (bytes) of packet's payload as flexible payload:: 2037 2038 testpmd> flow_director_flex_payload 0 l4 \ 2039 (4,5,6,7,8,9,10,11,12,13,14,15,16,17,18,19) 2040 2041get_sym_hash_ena_per_port 2042~~~~~~~~~~~~~~~~~~~~~~~~~ 2043 2044Get symmetric hash enable configuration per port:: 2045 2046 get_sym_hash_ena_per_port (port_id) 2047 2048For example, to get symmetric hash enable configuration of port 1:: 2049 2050 testpmd> get_sym_hash_ena_per_port 1 2051 2052set_sym_hash_ena_per_port 2053~~~~~~~~~~~~~~~~~~~~~~~~~ 2054 2055Set symmetric hash enable configuration per port to enable or disable:: 2056 2057 set_sym_hash_ena_per_port (port_id) (enable|disable) 2058 2059For example, to set symmetric hash enable configuration of port 1 to enable:: 2060 2061 testpmd> set_sym_hash_ena_per_port 1 enable 2062 2063get_hash_global_config 2064~~~~~~~~~~~~~~~~~~~~~~ 2065 2066Get the global configurations of hash filters:: 2067 2068 get_hash_global_config (port_id) 2069 2070For example, to get the global configurations of hash filters of port 1:: 2071 2072 testpmd> get_hash_global_config 1 2073 2074set_hash_global_config 2075~~~~~~~~~~~~~~~~~~~~~~ 2076 2077Set the global configurations of hash filters:: 2078 2079 set_hash_global_config (port_id) (toeplitz|simple_xor|default) \ 2080 (ipv4|ipv4-frag|ipv4-tcp|ipv4-udp|ipv4-sctp|ipv4-other|ipv6|ipv6-frag| \ 2081 ipv6-tcp|ipv6-udp|ipv6-sctp|ipv6-other|l2_payload) \ 2082 (enable|disable) 2083 2084For example, to enable simple_xor for flow type of ipv6 on port 2:: 2085 2086 testpmd> set_hash_global_config 2 simple_xor ipv6 enable 2087 2088set_hash_input_set 2089~~~~~~~~~~~~~~~~~~ 2090 2091Set the input set for hash:: 2092 2093 set_hash_input_set (port_id) (ipv4-frag|ipv4-tcp|ipv4-udp|ipv4-sctp| \ 2094 ipv4-other|ipv6-frag|ipv6-tcp|ipv6-udp|ipv6-sctp|ipv6-other| \ 2095 l2_payload) (ovlan|ivlan|src-ipv4|dst-ipv4|src-ipv6|dst-ipv6|ipv4-tos| \ 2096 ipv4-proto|ipv6-tc|ipv6-next-header|udp-src-port|udp-dst-port| \ 2097 tcp-src-port|tcp-dst-port|sctp-src-port|sctp-dst-port|sctp-veri-tag| \ 2098 udp-key|gre-key|fld-1st|fld-2nd|fld-3rd|fld-4th|fld-5th|fld-6th|fld-7th| \ 2099 fld-8th|none) (select|add) 2100 2101For example, to add source IP to hash input set for flow type of ipv4-udp on port 0:: 2102 2103 testpmd> set_hash_input_set 0 ipv4-udp src-ipv4 add 2104 2105set_fdir_input_set 2106~~~~~~~~~~~~~~~~~~ 2107 2108The Flow Director filters can match the different fields for different type of packet, i.e. specific input set 2109on per flow type and the flexible payload. This command can be used to change input set for each flow type. 2110 2111Set the input set for flow director:: 2112 2113 set_fdir_input_set (port_id) (ipv4-frag|ipv4-tcp|ipv4-udp|ipv4-sctp| \ 2114 ipv4-other|ipv6|ipv6-frag|ipv6-tcp|ipv6-udp|ipv6-sctp|ipv6-other| \ 2115 l2_payload) (ivlan|ethertype|src-ipv4|dst-ipv4|src-ipv6|dst-ipv6|ipv4-tos| \ 2116 ipv4-proto|ipv4-ttl|ipv6-tc|ipv6-next-header|ipv6-hop-limits| \ 2117 tudp-src-port|udp-dst-port|cp-src-port|tcp-dst-port|sctp-src-port| \ 2118 sctp-dst-port|sctp-veri-tag|none) (select|add) 2119 2120For example to add source IP to FD input set for flow type of ipv4-udp on port 0:: 2121 2122 testpmd> set_fdir_input_set 0 ipv4-udp src-ipv4 add 2123 2124global_config 2125~~~~~~~~~~~~~ 2126 2127Set different GRE key length for input set:: 2128 2129 global_config (port_id) gre-key-len (number in bytes) 2130 2131For example to set GRE key length for input set to 4 bytes on port 0:: 2132 2133 testpmd> global_config 0 gre-key-len 4 2134 2135 2136.. _testpmd_rte_flow: 2137 2138Flow rules management 2139--------------------- 2140 2141Control of the generic flow API (*rte_flow*) is fully exposed through the 2142``flow`` command (validation, creation, destruction and queries). 2143 2144Considering *rte_flow* overlaps with all `Filter Functions`_, using both 2145features simultaneously may cause undefined side-effects and is therefore 2146not recommended. 2147 2148``flow`` syntax 2149~~~~~~~~~~~~~~~ 2150 2151Because the ``flow`` command uses dynamic tokens to handle the large number 2152of possible flow rules combinations, its behavior differs slightly from 2153other commands, in particular: 2154 2155- Pressing *?* or the *<tab>* key displays contextual help for the current 2156 token, not that of the entire command. 2157 2158- Optional and repeated parameters are supported (provided they are listed 2159 in the contextual help). 2160 2161The first parameter stands for the operation mode. Possible operations and 2162their general syntax are described below. They are covered in detail in the 2163following sections. 2164 2165- Check whether a flow rule can be created:: 2166 2167 flow validate {port_id} 2168 [group {group_id}] [priority {level}] [ingress] [egress] 2169 pattern {item} [/ {item} [...]] / end 2170 actions {action} [/ {action} [...]] / end 2171 2172- Create a flow rule:: 2173 2174 flow create {port_id} 2175 [group {group_id}] [priority {level}] [ingress] [egress] 2176 pattern {item} [/ {item} [...]] / end 2177 actions {action} [/ {action} [...]] / end 2178 2179- Destroy specific flow rules:: 2180 2181 flow destroy {port_id} rule {rule_id} [...] 2182 2183- Destroy all flow rules:: 2184 2185 flow flush {port_id} 2186 2187- Query an existing flow rule:: 2188 2189 flow query {port_id} {rule_id} {action} 2190 2191- List existing flow rules sorted by priority, filtered by group 2192 identifiers:: 2193 2194 flow list {port_id} [group {group_id}] [...] 2195 2196Validating flow rules 2197~~~~~~~~~~~~~~~~~~~~~ 2198 2199``flow validate`` reports whether a flow rule would be accepted by the 2200underlying device in its current state but stops short of creating it. It is 2201bound to ``rte_flow_validate()``:: 2202 2203 flow validate {port_id} 2204 [group {group_id}] [priority {level}] [ingress] [egress] 2205 pattern {item} [/ {item} [...]] / end 2206 actions {action} [/ {action} [...]] / end 2207 2208If successful, it will show:: 2209 2210 Flow rule validated 2211 2212Otherwise it will show an error message of the form:: 2213 2214 Caught error type [...] ([...]): [...] 2215 2216This command uses the same parameters as ``flow create``, their format is 2217described in `Creating flow rules`_. 2218 2219Check whether redirecting any Ethernet packet received on port 0 to RX queue 2220index 6 is supported:: 2221 2222 testpmd> flow validate 0 ingress pattern eth / end 2223 actions queue index 6 / end 2224 Flow rule validated 2225 testpmd> 2226 2227Port 0 does not support TCPv6 rules:: 2228 2229 testpmd> flow validate 0 ingress pattern eth / ipv6 / tcp / end 2230 actions drop / end 2231 Caught error type 9 (specific pattern item): Invalid argument 2232 testpmd> 2233 2234Creating flow rules 2235~~~~~~~~~~~~~~~~~~~ 2236 2237``flow create`` validates and creates the specified flow rule. It is bound 2238to ``rte_flow_create()``:: 2239 2240 flow create {port_id} 2241 [group {group_id}] [priority {level}] [ingress] [egress] 2242 pattern {item} [/ {item} [...]] / end 2243 actions {action} [/ {action} [...]] / end 2244 2245If successful, it will return a flow rule ID usable with other commands:: 2246 2247 Flow rule #[...] created 2248 2249Otherwise it will show an error message of the form:: 2250 2251 Caught error type [...] ([...]): [...] 2252 2253Parameters describe in the following order: 2254 2255- Attributes (*group*, *priority*, *ingress*, *egress* tokens). 2256- A matching pattern, starting with the *pattern* token and terminated by an 2257 *end* pattern item. 2258- Actions, starting with the *actions* token and terminated by an *end* 2259 action. 2260 2261These translate directly to *rte_flow* objects provided as-is to the 2262underlying functions. 2263 2264The shortest valid definition only comprises mandatory tokens:: 2265 2266 testpmd> flow create 0 pattern end actions end 2267 2268Note that PMDs may refuse rules that essentially do nothing such as this 2269one. 2270 2271**All unspecified object values are automatically initialized to 0.** 2272 2273Attributes 2274^^^^^^^^^^ 2275 2276These tokens affect flow rule attributes (``struct rte_flow_attr``) and are 2277specified before the ``pattern`` token. 2278 2279- ``group {group id}``: priority group. 2280- ``priority {level}``: priority level within group. 2281- ``ingress``: rule applies to ingress traffic. 2282- ``egress``: rule applies to egress traffic. 2283 2284Each instance of an attribute specified several times overrides the previous 2285value as shown below (group 4 is used):: 2286 2287 testpmd> flow create 0 group 42 group 24 group 4 [...] 2288 2289Note that once enabled, ``ingress`` and ``egress`` cannot be disabled. 2290 2291While not specifying a direction is an error, some rules may allow both 2292simultaneously. 2293 2294Most rules affect RX therefore contain the ``ingress`` token:: 2295 2296 testpmd> flow create 0 ingress pattern [...] 2297 2298Matching pattern 2299^^^^^^^^^^^^^^^^ 2300 2301A matching pattern starts after the ``pattern`` token. It is made of pattern 2302items and is terminated by a mandatory ``end`` item. 2303 2304Items are named after their type (*RTE_FLOW_ITEM_TYPE_* from ``enum 2305rte_flow_item_type``). 2306 2307The ``/`` token is used as a separator between pattern items as shown 2308below:: 2309 2310 testpmd> flow create 0 ingress pattern eth / ipv4 / udp / end [...] 2311 2312Note that protocol items like these must be stacked from lowest to highest 2313layer to make sense. For instance, the following rule is either invalid or 2314unlikely to match any packet:: 2315 2316 testpmd> flow create 0 ingress pattern eth / udp / ipv4 / end [...] 2317 2318More information on these restrictions can be found in the *rte_flow* 2319documentation. 2320 2321Several items support additional specification structures, for example 2322``ipv4`` allows specifying source and destination addresses as follows:: 2323 2324 testpmd> flow create 0 ingress pattern eth / ipv4 src is 10.1.1.1 2325 dst is 10.2.0.0 / end [...] 2326 2327This rule matches all IPv4 traffic with the specified properties. 2328 2329In this example, ``src`` and ``dst`` are field names of the underlying 2330``struct rte_flow_item_ipv4`` object. All item properties can be specified 2331in a similar fashion. 2332 2333The ``is`` token means that the subsequent value must be matched exactly, 2334and assigns ``spec`` and ``mask`` fields in ``struct rte_flow_item`` 2335accordingly. Possible assignment tokens are: 2336 2337- ``is``: match value perfectly (with full bit-mask). 2338- ``spec``: match value according to configured bit-mask. 2339- ``last``: specify upper bound to establish a range. 2340- ``mask``: specify bit-mask with relevant bits set to one. 2341- ``prefix``: generate bit-mask from a prefix length. 2342 2343These yield identical results:: 2344 2345 ipv4 src is 10.1.1.1 2346 2347:: 2348 2349 ipv4 src spec 10.1.1.1 src mask 255.255.255.255 2350 2351:: 2352 2353 ipv4 src spec 10.1.1.1 src prefix 32 2354 2355:: 2356 2357 ipv4 src is 10.1.1.1 src last 10.1.1.1 # range with a single value 2358 2359:: 2360 2361 ipv4 src is 10.1.1.1 src last 0 # 0 disables range 2362 2363Inclusive ranges can be defined with ``last``:: 2364 2365 ipv4 src is 10.1.1.1 src last 10.2.3.4 # 10.1.1.1 to 10.2.3.4 2366 2367Note that ``mask`` affects both ``spec`` and ``last``:: 2368 2369 ipv4 src is 10.1.1.1 src last 10.2.3.4 src mask 255.255.0.0 2370 # matches 10.1.0.0 to 10.2.255.255 2371 2372Properties can be modified multiple times:: 2373 2374 ipv4 src is 10.1.1.1 src is 10.1.2.3 src is 10.2.3.4 # matches 10.2.3.4 2375 2376:: 2377 2378 ipv4 src is 10.1.1.1 src prefix 24 src prefix 16 # matches 10.1.0.0/16 2379 2380Pattern items 2381^^^^^^^^^^^^^ 2382 2383This section lists supported pattern items and their attributes, if any. 2384 2385- ``end``: end list of pattern items. 2386 2387- ``void``: no-op pattern item. 2388 2389- ``invert``: perform actions when pattern does not match. 2390 2391- ``any``: match any protocol for the current layer. 2392 2393 - ``num {unsigned}``: number of layers covered. 2394 2395- ``pf``: match packets addressed to the physical function. 2396 2397- ``vf``: match packets addressed to a virtual function ID. 2398 2399 - ``id {unsigned}``: destination VF ID. 2400 2401- ``port``: device-specific physical port index to use. 2402 2403 - ``index {unsigned}``: physical port index. 2404 2405- ``raw``: match an arbitrary byte string. 2406 2407 - ``relative {boolean}``: look for pattern after the previous item. 2408 - ``search {boolean}``: search pattern from offset (see also limit). 2409 - ``offset {integer}``: absolute or relative offset for pattern. 2410 - ``limit {unsigned}``: search area limit for start of pattern. 2411 - ``pattern {string}``: byte string to look for. 2412 2413- ``eth``: match Ethernet header. 2414 2415 - ``dst {MAC-48}``: destination MAC. 2416 - ``src {MAC-48}``: source MAC. 2417 - ``type {unsigned}``: EtherType. 2418 2419- ``vlan``: match 802.1Q/ad VLAN tag. 2420 2421 - ``tpid {unsigned}``: tag protocol identifier. 2422 - ``tci {unsigned}``: tag control information. 2423 - ``pcp {unsigned}``: priority code point. 2424 - ``dei {unsigned}``: drop eligible indicator. 2425 - ``vid {unsigned}``: VLAN identifier. 2426 2427- ``ipv4``: match IPv4 header. 2428 2429 - ``tos {unsigned}``: type of service. 2430 - ``ttl {unsigned}``: time to live. 2431 - ``proto {unsigned}``: next protocol ID. 2432 - ``src {ipv4 address}``: source address. 2433 - ``dst {ipv4 address}``: destination address. 2434 2435- ``ipv6``: match IPv6 header. 2436 2437 - ``tc {unsigned}``: traffic class. 2438 - ``flow {unsigned}``: flow label. 2439 - ``proto {unsigned}``: protocol (next header). 2440 - ``hop {unsigned}``: hop limit. 2441 - ``src {ipv6 address}``: source address. 2442 - ``dst {ipv6 address}``: destination address. 2443 2444- ``icmp``: match ICMP header. 2445 2446 - ``type {unsigned}``: ICMP packet type. 2447 - ``code {unsigned}``: ICMP packet code. 2448 2449- ``udp``: match UDP header. 2450 2451 - ``src {unsigned}``: UDP source port. 2452 - ``dst {unsigned}``: UDP destination port. 2453 2454- ``tcp``: match TCP header. 2455 2456 - ``src {unsigned}``: TCP source port. 2457 - ``dst {unsigned}``: TCP destination port. 2458 2459- ``sctp``: match SCTP header. 2460 2461 - ``src {unsigned}``: SCTP source port. 2462 - ``dst {unsigned}``: SCTP destination port. 2463 - ``tag {unsigned}``: validation tag. 2464 - ``cksum {unsigned}``: checksum. 2465 2466- ``vxlan``: match VXLAN header. 2467 2468 - ``vni {unsigned}``: VXLAN identifier. 2469 2470Actions list 2471^^^^^^^^^^^^ 2472 2473A list of actions starts after the ``actions`` token in the same fashion as 2474`Matching pattern`_; actions are separated by ``/`` tokens and the list is 2475terminated by a mandatory ``end`` action. 2476 2477Actions are named after their type (*RTE_FLOW_ACTION_TYPE_* from ``enum 2478rte_flow_action_type``). 2479 2480Dropping all incoming UDPv4 packets can be expressed as follows:: 2481 2482 testpmd> flow create 0 ingress pattern eth / ipv4 / udp / end 2483 actions drop / end 2484 2485Several actions have configurable properties which must be specified when 2486there is no valid default value. For example, ``queue`` requires a target 2487queue index. 2488 2489This rule redirects incoming UDPv4 traffic to queue index 6:: 2490 2491 testpmd> flow create 0 ingress pattern eth / ipv4 / udp / end 2492 actions queue index 6 / end 2493 2494While this one could be rejected by PMDs (unspecified queue index):: 2495 2496 testpmd> flow create 0 ingress pattern eth / ipv4 / udp / end 2497 actions queue / end 2498 2499As defined by *rte_flow*, the list is not ordered, all actions of a given 2500rule are performed simultaneously. These are equivalent:: 2501 2502 queue index 6 / void / mark id 42 / end 2503 2504:: 2505 2506 void / mark id 42 / queue index 6 / end 2507 2508All actions in a list should have different types, otherwise only the last 2509action of a given type is taken into account:: 2510 2511 queue index 4 / queue index 5 / queue index 6 / end # will use queue 6 2512 2513:: 2514 2515 drop / drop / drop / end # drop is performed only once 2516 2517:: 2518 2519 mark id 42 / queue index 3 / mark id 24 / end # mark will be 24 2520 2521Considering they are performed simultaneously, opposite and overlapping 2522actions can sometimes be combined when the end result is unambiguous:: 2523 2524 drop / queue index 6 / end # drop has no effect 2525 2526:: 2527 2528 drop / dup index 6 / end # same as above 2529 2530:: 2531 2532 queue index 6 / rss queues 6 7 8 / end # queue has no effect 2533 2534:: 2535 2536 drop / passthru / end # drop has no effect 2537 2538Note that PMDs may still refuse such combinations. 2539 2540Actions 2541^^^^^^^ 2542 2543This section lists supported actions and their attributes, if any. 2544 2545- ``end``: end list of actions. 2546 2547- ``void``: no-op action. 2548 2549- ``passthru``: let subsequent rule process matched packets. 2550 2551- ``mark``: attach 32 bit value to packets. 2552 2553 - ``id {unsigned}``: 32 bit value to return with packets. 2554 2555- ``flag``: flag packets. 2556 2557- ``queue``: assign packets to a given queue index. 2558 2559 - ``index {unsigned}``: queue index to use. 2560 2561- ``drop``: drop packets (note: passthru has priority). 2562 2563- ``count``: enable counters for this rule. 2564 2565- ``dup``: duplicate packets to a given queue index. 2566 2567 - ``index {unsigned}``: queue index to duplicate packets to. 2568 2569- ``rss``: spread packets among several queues. 2570 2571 - ``queues [{unsigned} [...]] end``: queue indices to use. 2572 2573- ``pf``: redirect packets to physical device function. 2574 2575- ``vf``: redirect packets to virtual device function. 2576 2577 - ``original {boolean}``: use original VF ID if possible. 2578 - ``id {unsigned}``: VF ID to redirect packets to. 2579 2580Destroying flow rules 2581~~~~~~~~~~~~~~~~~~~~~ 2582 2583``flow destroy`` destroys one or more rules from their rule ID (as returned 2584by ``flow create``), this command calls ``rte_flow_destroy()`` as many 2585times as necessary:: 2586 2587 flow destroy {port_id} rule {rule_id} [...] 2588 2589If successful, it will show:: 2590 2591 Flow rule #[...] destroyed 2592 2593It does not report anything for rule IDs that do not exist. The usual error 2594message is shown when a rule cannot be destroyed:: 2595 2596 Caught error type [...] ([...]): [...] 2597 2598``flow flush`` destroys all rules on a device and does not take extra 2599arguments. It is bound to ``rte_flow_flush()``:: 2600 2601 flow flush {port_id} 2602 2603Any errors are reported as above. 2604 2605Creating several rules and destroying them:: 2606 2607 testpmd> flow create 0 ingress pattern eth / ipv6 / end 2608 actions queue index 2 / end 2609 Flow rule #0 created 2610 testpmd> flow create 0 ingress pattern eth / ipv4 / end 2611 actions queue index 3 / end 2612 Flow rule #1 created 2613 testpmd> flow destroy 0 rule 0 rule 1 2614 Flow rule #1 destroyed 2615 Flow rule #0 destroyed 2616 testpmd> 2617 2618The same result can be achieved using ``flow flush``:: 2619 2620 testpmd> flow create 0 ingress pattern eth / ipv6 / end 2621 actions queue index 2 / end 2622 Flow rule #0 created 2623 testpmd> flow create 0 ingress pattern eth / ipv4 / end 2624 actions queue index 3 / end 2625 Flow rule #1 created 2626 testpmd> flow flush 0 2627 testpmd> 2628 2629Non-existent rule IDs are ignored:: 2630 2631 testpmd> flow create 0 ingress pattern eth / ipv6 / end 2632 actions queue index 2 / end 2633 Flow rule #0 created 2634 testpmd> flow create 0 ingress pattern eth / ipv4 / end 2635 actions queue index 3 / end 2636 Flow rule #1 created 2637 testpmd> flow destroy 0 rule 42 rule 10 rule 2 2638 testpmd> 2639 testpmd> flow destroy 0 rule 0 2640 Flow rule #0 destroyed 2641 testpmd> 2642 2643Querying flow rules 2644~~~~~~~~~~~~~~~~~~~ 2645 2646``flow query`` queries a specific action of a flow rule having that 2647ability. Such actions collect information that can be reported using this 2648command. It is bound to ``rte_flow_query()``:: 2649 2650 flow query {port_id} {rule_id} {action} 2651 2652If successful, it will display either the retrieved data for known actions 2653or the following message:: 2654 2655 Cannot display result for action type [...] ([...]) 2656 2657Otherwise, it will complain either that the rule does not exist or that some 2658error occurred:: 2659 2660 Flow rule #[...] not found 2661 2662:: 2663 2664 Caught error type [...] ([...]): [...] 2665 2666Currently only the ``count`` action is supported. This action reports the 2667number of packets that hit the flow rule and the total number of bytes. Its 2668output has the following format:: 2669 2670 count: 2671 hits_set: [...] # whether "hits" contains a valid value 2672 bytes_set: [...] # whether "bytes" contains a valid value 2673 hits: [...] # number of packets 2674 bytes: [...] # number of bytes 2675 2676Querying counters for TCPv6 packets redirected to queue 6:: 2677 2678 testpmd> flow create 0 ingress pattern eth / ipv6 / tcp / end 2679 actions queue index 6 / count / end 2680 Flow rule #4 created 2681 testpmd> flow query 0 4 count 2682 count: 2683 hits_set: 1 2684 bytes_set: 0 2685 hits: 386446 2686 bytes: 0 2687 testpmd> 2688 2689Listing flow rules 2690~~~~~~~~~~~~~~~~~~ 2691 2692``flow list`` lists existing flow rules sorted by priority and optionally 2693filtered by group identifiers:: 2694 2695 flow list {port_id} [group {group_id}] [...] 2696 2697This command only fails with the following message if the device does not 2698exist:: 2699 2700 Invalid port [...] 2701 2702Output consists of a header line followed by a short description of each 2703flow rule, one per line. There is no output at all when no flow rules are 2704configured on the device:: 2705 2706 ID Group Prio Attr Rule 2707 [...] [...] [...] [...] [...] 2708 2709``Attr`` column flags: 2710 2711- ``i`` for ``ingress``. 2712- ``e`` for ``egress``. 2713 2714Creating several flow rules and listing them:: 2715 2716 testpmd> flow create 0 ingress pattern eth / ipv4 / end 2717 actions queue index 6 / end 2718 Flow rule #0 created 2719 testpmd> flow create 0 ingress pattern eth / ipv6 / end 2720 actions queue index 2 / end 2721 Flow rule #1 created 2722 testpmd> flow create 0 priority 5 ingress pattern eth / ipv4 / udp / end 2723 actions rss queues 6 7 8 end / end 2724 Flow rule #2 created 2725 testpmd> flow list 0 2726 ID Group Prio Attr Rule 2727 0 0 0 i- ETH IPV4 => QUEUE 2728 1 0 0 i- ETH IPV6 => QUEUE 2729 2 0 5 i- ETH IPV4 UDP => RSS 2730 testpmd> 2731 2732Rules are sorted by priority (i.e. group ID first, then priority level):: 2733 2734 testpmd> flow list 1 2735 ID Group Prio Attr Rule 2736 0 0 0 i- ETH => COUNT 2737 6 0 500 i- ETH IPV6 TCP => DROP COUNT 2738 5 0 1000 i- ETH IPV6 ICMP => QUEUE 2739 1 24 0 i- ETH IPV4 UDP => QUEUE 2740 4 24 10 i- ETH IPV4 TCP => DROP 2741 3 24 20 i- ETH IPV4 => DROP 2742 2 24 42 i- ETH IPV4 UDP => QUEUE 2743 7 63 0 i- ETH IPV6 UDP VXLAN => MARK QUEUE 2744 testpmd> 2745 2746Output can be limited to specific groups:: 2747 2748 testpmd> flow list 1 group 0 group 63 2749 ID Group Prio Attr Rule 2750 0 0 0 i- ETH => COUNT 2751 6 0 500 i- ETH IPV6 TCP => DROP COUNT 2752 5 0 1000 i- ETH IPV6 ICMP => QUEUE 2753 7 63 0 i- ETH IPV6 UDP VXLAN => MARK QUEUE 2754 testpmd> 2755