xref: /dpdk/doc/guides/testpmd_app_ug/testpmd_funcs.rst (revision 675783e064129414a9b67ef9fb7c8a6eb1eeebdf)
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 X
55       info [Mul-choice STRING]: show|clear port info|stats|xstats|fdir|stat_qmap|dcb_tc all
56       stats [Mul-choice STRING]: show|clear port info|stats|xstats|fdir|stat_qmap|dcb_tc X
57       stats [Mul-choice STRING]: show|clear port info|stats|xstats|fdir|stat_qmap|dcb_tc 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) (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
150For example:
151
152.. code-block:: console
153
154   testpmd> show port info 0
155
156   ********************* Infos for port 0 *********************
157
158   MAC address: XX:XX:XX:XX:XX:XX
159   Connect to socket: 0
160   memory allocation on the socket: 0
161   Link status: up
162   Link speed: 40000 Mbps
163   Link duplex: full-duplex
164   Promiscuous mode: enabled
165   Allmulticast mode: disabled
166   Maximum number of MAC addresses: 64
167   Maximum number of MAC addresses of hash filtering: 0
168   VLAN offload:
169       strip on
170       filter on
171       qinq(extend) off
172   Redirection table size: 512
173   Supported flow types:
174     ipv4-frag
175     ipv4-tcp
176     ipv4-udp
177     ipv4-sctp
178     ipv4-other
179     ipv6-frag
180     ipv6-tcp
181     ipv6-udp
182     ipv6-sctp
183     ipv6-other
184     l2_payload
185     port
186     vxlan
187     geneve
188     nvgre
189
190show port rss reta
191~~~~~~~~~~~~~~~~~~
192
193Display the rss redirection table entry indicated by masks on port X::
194
195   testpmd> show port (port_id) rss reta (size) (mask0, mask1...)
196
197size is used to indicate the hardware supported reta size
198
199show port rss-hash
200~~~~~~~~~~~~~~~~~~
201
202Display the RSS hash functions and RSS hash key of a port::
203
204   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]
205
206clear port
207~~~~~~~~~~
208
209Clear the port statistics for a given port or for all ports::
210
211   testpmd> clear port (info|stats|xstats|fdir|stat_qmap) (port_id|all)
212
213For example::
214
215   testpmd> clear port stats all
216
217show (rxq|txq)
218~~~~~~~~~~~~~~
219
220Display information for a given port's RX/TX queue::
221
222   testpmd> show (rxq|txq) info (port_id) (queue_id)
223
224show config
225~~~~~~~~~~~
226
227Displays the configuration of the application.
228The configuration comes from the command-line, the runtime or the application defaults::
229
230   testpmd> show config (rxtx|cores|fwd|txpkts)
231
232The available information categories are:
233
234* ``rxtx``: RX/TX configuration items.
235
236* ``cores``: List of forwarding cores.
237
238* ``fwd``: Packet forwarding configuration.
239
240* ``txpkts``: Packets to TX configuration.
241
242For example:
243
244.. code-block:: console
245
246   testpmd> show config rxtx
247
248   io packet forwarding - CRC stripping disabled - packets/burst=16
249   nb forwarding cores=2 - nb forwarding ports=1
250   RX queues=1 - RX desc=128 - RX free threshold=0
251   RX threshold registers: pthresh=8 hthresh=8 wthresh=4
252   TX queues=1 - TX desc=512 - TX free threshold=0
253   TX threshold registers: pthresh=36 hthresh=0 wthresh=0
254   TX RS bit threshold=0 - TXQ flags=0x0
255
256set fwd
257~~~~~~~
258
259Set the packet forwarding mode::
260
261   testpmd> set fwd (io|mac|macswap|flowgen| \
262                     rxonly|txonly|csum|icmpecho) (""|retry)
263
264``retry`` can be specified for forwarding engines except ``rx_only``.
265
266The available information categories are:
267
268* ``io``: Forwards packets "as-is" in I/O mode.
269  This is the fastest possible forwarding operation as it does not access packets data.
270  This is the default mode.
271
272* ``mac``: Changes the source and the destination Ethernet addresses of packets before forwarding them.
273  Default application behaviour is to set source Ethernet address to that of the transmitting interface, and destination
274  address to a dummy value (set during init). The user may specify a target destination Ethernet address via the 'eth-peer' or
275  'eth-peer-configfile' command-line options. It is not currently possible to specify a specific source Ethernet address.
276
277* ``macswap``: MAC swap forwarding mode.
278  Swaps the source and the destination Ethernet addresses of packets before forwarding them.
279
280* ``flowgen``: Multi-flow generation mode.
281  Originates a number of flows (with varying destination IP addresses), and terminate receive traffic.
282
283* ``rxonly``: Receives packets but doesn't transmit them.
284
285* ``txonly``: Generates and transmits packets without receiving any.
286
287* ``csum``: Changes the checksum field with hardware or software methods depending on the offload flags on the packet.
288
289* ``icmpecho``: Receives a burst of packets, lookup for IMCP echo requests and, if any, send back ICMP echo replies.
290
291* ``ieee1588``: Demonstrate L2 IEEE1588 V2 PTP timestamping for RX and TX. Requires ``CONFIG_RTE_LIBRTE_IEEE1588=y``.
292
293Note: TX timestamping is only available in the "Full Featured" TX path. To force ``testpmd`` into this mode set ``--txqflags=0``.
294
295Example::
296
297   testpmd> set fwd rxonly
298
299   Set rxonly packet forwarding mode
300
301
302read rxd
303~~~~~~~~
304
305Display an RX descriptor for a port RX queue::
306
307   testpmd> read rxd (port_id) (queue_id) (rxd_id)
308
309For example::
310
311   testpmd> read rxd 0 0 4
312        0x0000000B - 0x001D0180 / 0x0000000B - 0x001D0180
313
314read txd
315~~~~~~~~
316
317Display a TX descriptor for a port TX queue::
318
319   testpmd> read txd (port_id) (queue_id) (txd_id)
320
321For example::
322
323   testpmd> read txd 0 0 4
324        0x00000001 - 0x24C3C440 / 0x000F0000 - 0x2330003C
325
326
327Configuration Functions
328-----------------------
329
330The testpmd application can be configured from the runtime as well as from the command-line.
331
332This section details the available configuration functions that are available.
333
334.. note::
335
336   Configuration changes only become active when forwarding is started/restarted.
337
338set default
339~~~~~~~~~~~
340
341Reset forwarding to the default configuration::
342
343   testpmd> set default
344
345set verbose
346~~~~~~~~~~~
347
348Set the debug verbosity level::
349
350   testpmd> set verbose (level)
351
352Currently the only available levels are 0 (silent except for error) and 1 (fully verbose).
353
354set nbport
355~~~~~~~~~~
356
357Set the number of ports used by the application:
358
359set nbport (num)
360
361This is equivalent to the ``--nb-ports`` command-line option.
362
363set nbcore
364~~~~~~~~~~
365
366Set the number of cores used by the application::
367
368   testpmd> set nbcore (num)
369
370This is equivalent to the ``--nb-cores`` command-line option.
371
372.. note::
373
374   The number of cores used must not be greater than number of ports used multiplied by the number of queues per port.
375
376set coremask
377~~~~~~~~~~~~
378
379Set the forwarding cores hexadecimal mask::
380
381   testpmd> set coremask (mask)
382
383This is equivalent to the ``--coremask`` command-line option.
384
385.. note::
386
387   The master lcore is reserved for command line parsing only and cannot be masked on for packet forwarding.
388
389set portmask
390~~~~~~~~~~~~
391
392Set the forwarding ports hexadecimal mask::
393
394   testpmd> set portmask (mask)
395
396This is equivalent to the ``--portmask`` command-line option.
397
398set burst
399~~~~~~~~~
400
401Set number of packets per burst::
402
403   testpmd> set burst (num)
404
405This is equivalent to the ``--burst command-line`` option.
406
407When retry is enabled, the transmit delay time and number of retries can also be set::
408
409   testpmd> set burst tx delay (microseconds) retry (num)
410
411set txpkts
412~~~~~~~~~~
413
414Set the length of each segment of the TX-ONLY packets or length of packet for FLOWGEN mode::
415
416   testpmd> set txpkts (x[,y]*)
417
418Where x[,y]* represents a CSV list of values, without white space.
419
420set txsplit
421~~~~~~~~~~~
422
423Set the split policy for the TX packets, applicable for TX-ONLY and CSUM forwarding modes::
424
425   testpmd> set txsplit (off|on|rand)
426
427Where:
428
429* ``off`` disable packet copy & split for CSUM mode.
430
431* ``on`` split outgoing packet into multiple segments. Size of each segment
432  and number of segments per packet is determined by ``set txpkts`` command
433  (see above).
434
435* ``rand`` same as 'on', but number of segments per each packet is a random value between 1 and total number of segments.
436
437set corelist
438~~~~~~~~~~~~
439
440Set the list of forwarding cores::
441
442   testpmd> set corelist (x[,y]*)
443
444For example, to change the forwarding cores:
445
446.. code-block:: console
447
448   testpmd> set corelist 3,1
449   testpmd> show config fwd
450
451   io packet forwarding - ports=2 - cores=2 - streams=2 - NUMA support disabled
452   Logical Core 3 (socket 0) forwards packets on 1 streams:
453   RX P=0/Q=0 (socket 0) -> TX P=1/Q=0 (socket 0) peer=02:00:00:00:00:01
454   Logical Core 1 (socket 0) forwards packets on 1 streams:
455   RX P=1/Q=0 (socket 0) -> TX P=0/Q=0 (socket 0) peer=02:00:00:00:00:00
456
457.. note::
458
459   The cores are used in the same order as specified on the command line.
460
461set portlist
462~~~~~~~~~~~~
463
464Set the list of forwarding ports::
465
466   testpmd> set portlist (x[,y]*)
467
468For example, to change the port forwarding:
469
470.. code-block:: console
471
472   testpmd> set portlist 0,2,1,3
473   testpmd> show config fwd
474
475   io packet forwarding - ports=4 - cores=1 - streams=4
476   Logical Core 3 (socket 0) forwards packets on 4 streams:
477   RX P=0/Q=0 (socket 0) -> TX P=2/Q=0 (socket 0) peer=02:00:00:00:00:01
478   RX P=2/Q=0 (socket 0) -> TX P=0/Q=0 (socket 0) peer=02:00:00:00:00:00
479   RX P=1/Q=0 (socket 0) -> TX P=3/Q=0 (socket 0) peer=02:00:00:00:00:03
480   RX P=3/Q=0 (socket 0) -> TX P=1/Q=0 (socket 0) peer=02:00:00:00:00:02
481
482set tx loopback
483~~~~~~~~~~~~~~~
484
485Enable/disable tx loopback::
486
487   testpmd> set tx loopback (port_id) (on|off)
488
489set drop enable
490~~~~~~~~~~~~~~~
491
492set drop enable bit for all queues::
493
494   testpmd> set all queues drop (port_id) (on|off)
495
496set split drop enable (for VF)
497~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
498
499set split drop enable bit for VF from PF::
500
501   testpmd> set vf split drop (port_id) (vf_id) (on|off)
502
503set mac antispoof (for VF)
504~~~~~~~~~~~~~~~~~~~~~~~~~~
505
506Set mac antispoof for a VF from the PF::
507
508   testpmd> set vf mac antispoof  (port_id) (vf_id) (on|off)
509
510set macsec offload
511~~~~~~~~~~~~~~~~~~
512
513Enable/disable MACsec offload::
514
515   testpmd> set macsec offload (port_id) on encrypt (on|off) replay-protect (on|off)
516   testpmd> set macsec offload (port_id) off
517
518set macsec sc
519~~~~~~~~~~~~~
520
521Configure MACsec secure connection (SC)::
522
523   testpmd> set macsec sc (tx|rx) (port_id) (mac) (pi)
524
525.. note::
526
527   The pi argument is ignored for tx.
528   Check the NIC Datasheet for hardware limits.
529
530set macsec sa
531~~~~~~~~~~~~~
532
533Configure MACsec secure association (SA)::
534
535   testpmd> set macsec sa (tx|rx) (port_id) (idx) (an) (pn) (key)
536
537.. note::
538
539   The IDX value must be 0 or 1.
540   Check the NIC Datasheet for hardware limits.
541
542set broadcast mode (for VF)
543~~~~~~~~~~~~~~~~~~~~~~~~~~~
544
545Set broadcast mode for a VF from the PF::
546
547   testpmd> set vf broadcast (port_id) (vf_id) (on|off)
548
549vlan set strip
550~~~~~~~~~~~~~~
551
552Set the VLAN strip on a port::
553
554   testpmd> vlan set strip (on|off) (port_id)
555
556vlan set stripq
557~~~~~~~~~~~~~~~
558
559Set the VLAN strip for a queue on a port::
560
561   testpmd> vlan set stripq (on|off) (port_id,queue_id)
562
563vlan set stripq (for VF)
564~~~~~~~~~~~~~~~~~~~~~~~~
565
566Set VLAN strip for all queues in a pool for a VF from the PF::
567
568   testpmd> set vf vlan stripq (port_id) (vf_id) (on|off)
569
570vlan set insert (for VF)
571~~~~~~~~~~~~~~~~~~~~~~~~
572
573Set VLAN insert for a VF from the PF::
574
575   testpmd> set vf vlan insert (port_id) (vf_id) (vlan_id)
576
577vlan set antispoof (for VF)
578~~~~~~~~~~~~~~~~~~~~~~~~~~~
579
580Set VLAN antispoof for a VF from the PF::
581
582   testpmd> set vf vlan antispoof (port_id) (vf_id) (on|off)
583
584vlan set filter
585~~~~~~~~~~~~~~~
586
587Set the VLAN filter on a port::
588
589   testpmd> vlan set filter (on|off) (port_id)
590
591vlan set qinq
592~~~~~~~~~~~~~
593
594Set the VLAN QinQ (extended queue in queue) on for a port::
595
596   testpmd> vlan set qinq (on|off) (port_id)
597
598vlan set tpid
599~~~~~~~~~~~~~
600
601Set the inner or outer VLAN TPID for packet filtering on a port::
602
603   testpmd> vlan set (inner|outer) tpid (value) (port_id)
604
605.. note::
606
607   TPID value must be a 16-bit number (value <= 65536).
608
609rx_vlan add
610~~~~~~~~~~~
611
612Add a VLAN ID, or all identifiers, to the set of VLAN identifiers filtered by port ID::
613
614   testpmd> rx_vlan add (vlan_id|all) (port_id)
615
616.. note::
617
618   VLAN filter must be set on that port. VLAN ID < 4096.
619   Depending on the NIC used, number of vlan_ids may be limited to the maximum entries
620   in VFTA table. This is important if enabling all vlan_ids.
621
622rx_vlan rm
623~~~~~~~~~~
624
625Remove a VLAN ID, or all identifiers, from the set of VLAN identifiers filtered by port ID::
626
627   testpmd> rx_vlan rm (vlan_id|all) (port_id)
628
629rx_vlan add (for VF)
630~~~~~~~~~~~~~~~~~~~~
631
632Add a VLAN ID, to the set of VLAN identifiers filtered for VF(s) for port ID::
633
634   testpmd> rx_vlan add (vlan_id) port (port_id) vf (vf_mask)
635
636rx_vlan rm (for VF)
637~~~~~~~~~~~~~~~~~~~
638
639Remove a VLAN ID, from the set of VLAN identifiers filtered for VF(s) for port ID::
640
641   testpmd> rx_vlan rm (vlan_id) port (port_id) vf (vf_mask)
642
643tunnel_filter add
644~~~~~~~~~~~~~~~~~
645
646Add a tunnel filter on a port::
647
648   testpmd> tunnel_filter add (port_id) (outer_mac) (inner_mac) (ip_addr) \
649            (inner_vlan) (vxlan|nvgre|ipingre) (imac-ivlan|imac-ivlan-tenid|\
650            imac-tenid|imac|omac-imac-tenid|oip|iip) (tenant_id) (queue_id)
651
652The available information categories are:
653
654* ``vxlan``: Set tunnel type as VXLAN.
655
656* ``nvgre``: Set tunnel type as NVGRE.
657
658* ``ipingre``: Set tunnel type as IP-in-GRE.
659
660* ``imac-ivlan``: Set filter type as Inner MAC and VLAN.
661
662* ``imac-ivlan-tenid``: Set filter type as Inner MAC, VLAN and tenant ID.
663
664* ``imac-tenid``: Set filter type as Inner MAC and tenant ID.
665
666* ``imac``: Set filter type as Inner MAC.
667
668* ``omac-imac-tenid``: Set filter type as Outer MAC, Inner MAC and tenant ID.
669
670* ``oip``: Set filter type as Outer IP.
671
672* ``iip``: Set filter type as Inner IP.
673
674Example::
675
676   testpmd> tunnel_filter add 0 68:05:CA:28:09:82 00:00:00:00:00:00 \
677            192.168.2.2 0 ipingre oip 1 1
678
679   Set an IP-in-GRE tunnel on port 0, and the filter type is Outer IP.
680
681tunnel_filter remove
682~~~~~~~~~~~~~~~~~~~~
683
684Remove a tunnel filter on a port::
685
686   testpmd> tunnel_filter rm (port_id) (outer_mac) (inner_mac) (ip_addr) \
687            (inner_vlan) (vxlan|nvgre|ipingre) (imac-ivlan|imac-ivlan-tenid|\
688            imac-tenid|imac|omac-imac-tenid|oip|iip) (tenant_id) (queue_id)
689
690rx_vxlan_port add
691~~~~~~~~~~~~~~~~~
692
693Add an UDP port for VXLAN packet filter on a port::
694
695   testpmd> rx_vxlan_port add (udp_port) (port_id)
696
697rx_vxlan_port remove
698~~~~~~~~~~~~~~~~~~~~
699
700Remove an UDP port for VXLAN packet filter on a port::
701
702   testpmd> rx_vxlan_port rm (udp_port) (port_id)
703
704tx_vlan set
705~~~~~~~~~~~
706
707Set hardware insertion of VLAN IDs in packets sent on a port::
708
709   testpmd> tx_vlan set (port_id) vlan_id[, vlan_id_outer]
710
711For example, set a single VLAN ID (5) insertion on port 0::
712
713   tx_vlan set 0 5
714
715Or, set double VLAN ID (inner: 2, outer: 3) insertion on port 1::
716
717   tx_vlan set 1 2 3
718
719
720tx_vlan set pvid
721~~~~~~~~~~~~~~~~
722
723Set port based hardware insertion of VLAN ID in packets sent on a port::
724
725   testpmd> tx_vlan set pvid (port_id) (vlan_id) (on|off)
726
727tx_vlan reset
728~~~~~~~~~~~~~
729
730Disable hardware insertion of a VLAN header in packets sent on a port::
731
732   testpmd> tx_vlan reset (port_id)
733
734csum set
735~~~~~~~~
736
737Select hardware or software calculation of the checksum when
738transmitting a packet using the ``csum`` forwarding engine::
739
740   testpmd> csum set (ip|udp|tcp|sctp|outer-ip) (hw|sw) (port_id)
741
742Where:
743
744* ``ip|udp|tcp|sctp`` always relate to  the inner layer.
745
746* ``outer-ip`` relates to the outer IP layer (only for IPv4) in the case where the packet is recognized
747  as a tunnel packet by the forwarding engine (vxlan, gre and ipip are
748  supported). See also the ``csum parse-tunnel`` command.
749
750.. note::
751
752   Check the NIC Datasheet for hardware limits.
753
754csum parse-tunnel
755~~~~~~~~~~~~~~~~~
756
757Define how tunneled packets should be handled by the csum forward
758engine::
759
760   testpmd> csum parse-tunnel (on|off) (tx_port_id)
761
762If enabled, the csum forward engine will try to recognize supported
763tunnel headers (vxlan, gre, ipip).
764
765If disabled, treat tunnel packets as non-tunneled packets (a inner
766header is handled as a packet payload).
767
768.. note::
769
770   The port argument is the TX port like in the ``csum set`` command.
771
772Example:
773
774Consider a packet in packet like the following::
775
776   eth_out/ipv4_out/udp_out/vxlan/eth_in/ipv4_in/tcp_in
777
778* If parse-tunnel is enabled, the ``ip|udp|tcp|sctp`` parameters of ``csum set``
779  command relate to the inner headers (here ``ipv4_in`` and ``tcp_in``), and the
780  ``outer-ip parameter`` relates to the outer headers (here ``ipv4_out``).
781
782* If parse-tunnel is disabled, the ``ip|udp|tcp|sctp`` parameters of ``csum  set``
783   command relate to the outer headers, here ``ipv4_out`` and ``udp_out``.
784
785csum show
786~~~~~~~~~
787
788Display tx checksum offload configuration::
789
790   testpmd> csum show (port_id)
791
792tso set
793~~~~~~~
794
795Enable TCP Segmentation Offload (TSO) in the ``csum`` forwarding engine::
796
797   testpmd> tso set (segsize) (port_id)
798
799.. note::
800
801   Check the NIC datasheet for hardware limits.
802
803tso show
804~~~~~~~~
805
806Display the status of TCP Segmentation Offload::
807
808   testpmd> tso show (port_id)
809
810mac_addr add
811~~~~~~~~~~~~
812
813Add an alternative MAC address to a port::
814
815   testpmd> mac_addr add (port_id) (XX:XX:XX:XX:XX:XX)
816
817mac_addr remove
818~~~~~~~~~~~~~~~
819
820Remove a MAC address from a port::
821
822   testpmd> mac_addr remove (port_id) (XX:XX:XX:XX:XX:XX)
823
824mac_addr add (for VF)
825~~~~~~~~~~~~~~~~~~~~~
826
827Add an alternative MAC address for a VF to a port::
828
829   testpmd> mac_add add port (port_id) vf (vf_id) (XX:XX:XX:XX:XX:XX)
830
831mac_addr set (for VF)
832~~~~~~~~~~~~~~~~~~~~~
833
834Set the MAC address for a VF from the PF::
835
836   testpmd> set vf mac addr (port_id) (vf_id) (XX:XX:XX:XX:XX:XX)
837
838set port-uta
839~~~~~~~~~~~~
840
841Set the unicast hash filter(s) on/off for a port::
842
843   testpmd> set port (port_id) uta (XX:XX:XX:XX:XX:XX|all) (on|off)
844
845set promisc
846~~~~~~~~~~~
847
848Set the promiscuous mode on for a port or for all ports.
849In promiscuous mode packets are not dropped if they aren't for the specified MAC address::
850
851   testpmd> set promisc (port_id|all) (on|off)
852
853set allmulti
854~~~~~~~~~~~~
855
856Set the allmulti mode for a port or for all ports::
857
858   testpmd> set allmulti (port_id|all) (on|off)
859
860Same as the ifconfig (8) option. Controls how multicast packets are handled.
861
862set promisc (for VF)
863~~~~~~~~~~~~~~~~~~~~
864
865Set the unicast promiscuous mode for a VF from PF.
866It's supported by Intel i40e NICs now.
867In promiscuous mode packets are not dropped if they aren't for the specified MAC address::
868
869   testpmd> set vf promisc (port_id) (vf_id) (on|off)
870
871set allmulticast (for VF)
872~~~~~~~~~~~~~~~~~~~~~~~~~
873
874Set the multicast promiscuous mode for a VF from PF.
875It's supported by Intel i40e NICs now.
876In promiscuous mode packets are not dropped if they aren't for the specified MAC address::
877
878   testpmd> set vf allmulti (port_id) (vf_id) (on|off)
879
880set flow_ctrl rx
881~~~~~~~~~~~~~~~~
882
883Set the link flow control parameter on a port::
884
885   testpmd> set flow_ctrl rx (on|off) tx (on|off) (high_water) (low_water) \
886            (pause_time) (send_xon) mac_ctrl_frame_fwd (on|off) \
887	    autoneg (on|off) (port_id)
888
889Where:
890
891* ``high_water`` (integer): High threshold value to trigger XOFF.
892
893* ``low_water`` (integer): Low threshold value to trigger XON.
894
895* ``pause_time`` (integer): Pause quota in the Pause frame.
896
897* ``send_xon`` (0/1): Send XON frame.
898
899* ``mac_ctrl_frame_fwd``: Enable receiving MAC control frames.
900
901* ``autoneg``: Change the auto-negotiation parameter.
902
903set pfc_ctrl rx
904~~~~~~~~~~~~~~~
905
906Set the priority flow control parameter on a port::
907
908   testpmd> set pfc_ctrl rx (on|off) tx (on|off) (high_water) (low_water) \
909            (pause_time) (priority) (port_id)
910
911Where:
912
913* ``high_water`` (integer): High threshold value.
914
915* ``low_water`` (integer): Low threshold value.
916
917* ``pause_time`` (integer): Pause quota in the Pause frame.
918
919* ``priority`` (0-7): VLAN User Priority.
920
921set stat_qmap
922~~~~~~~~~~~~~
923
924Set statistics mapping (qmapping 0..15) for RX/TX queue on port::
925
926   testpmd> set stat_qmap (tx|rx) (port_id) (queue_id) (qmapping)
927
928For example, to set rx queue 2 on port 0 to mapping 5::
929
930   testpmd>set stat_qmap rx 0 2 5
931
932set port - rx/tx (for VF)
933~~~~~~~~~~~~~~~~~~~~~~~~~
934
935Set VF receive/transmit from a port::
936
937   testpmd> set port (port_id) vf (vf_id) (rx|tx) (on|off)
938
939set port - mac address filter (for VF)
940~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
941
942Add/Remove unicast or multicast MAC addr filter for a VF::
943
944   testpmd> set port (port_id) vf (vf_id) (mac_addr) \
945            (exact-mac|exact-mac-vlan|hashmac|hashmac-vlan) (on|off)
946
947set port - rx mode(for VF)
948~~~~~~~~~~~~~~~~~~~~~~~~~~
949
950Set the VF receive mode of a port::
951
952   testpmd> set port (port_id) vf (vf_id) \
953            rxmode (AUPE|ROPE|BAM|MPE) (on|off)
954
955The available receive modes are:
956
957* ``AUPE``: Accepts untagged VLAN.
958
959* ``ROPE``: Accepts unicast hash.
960
961* ``BAM``: Accepts broadcast packets.
962
963* ``MPE``: Accepts all multicast packets.
964
965set port - tx_rate (for Queue)
966~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
967
968Set TX rate limitation for a queue on a port::
969
970   testpmd> set port (port_id) queue (queue_id) rate (rate_value)
971
972set port - tx_rate (for VF)
973~~~~~~~~~~~~~~~~~~~~~~~~~~~
974
975Set TX rate limitation for queues in VF on a port::
976
977   testpmd> set port (port_id) vf (vf_id) rate (rate_value) queue_mask (queue_mask)
978
979set port - mirror rule
980~~~~~~~~~~~~~~~~~~~~~~
981
982Set pool or vlan type mirror rule for a port::
983
984   testpmd> set port (port_id) mirror-rule (rule_id) \
985            (pool-mirror-up|pool-mirror-down|vlan-mirror) \
986            (poolmask|vlanid[,vlanid]*) dst-pool (pool_id) (on|off)
987
988Set link mirror rule for a port::
989
990   testpmd> set port (port_id) mirror-rule (rule_id) \
991           (uplink-mirror|downlink-mirror) dst-pool (pool_id) (on|off)
992
993For example to enable mirror traffic with vlan 0,1 to pool 0::
994
995   set port 0 mirror-rule 0 vlan-mirror 0,1 dst-pool 0 on
996
997reset port - mirror rule
998~~~~~~~~~~~~~~~~~~~~~~~~
999
1000Reset a mirror rule for a port::
1001
1002   testpmd> reset port (port_id) mirror-rule (rule_id)
1003
1004set flush_rx
1005~~~~~~~~~~~~
1006
1007Set the flush on RX streams before forwarding.
1008The default is flush ``on``.
1009Mainly used with PCAP drivers to turn off the default behavior of flushing the first 512 packets on RX streams::
1010
1011   testpmd> set flush_rx off
1012
1013set bypass mode
1014~~~~~~~~~~~~~~~
1015
1016Set the bypass mode for the lowest port on bypass enabled NIC::
1017
1018   testpmd> set bypass mode (normal|bypass|isolate) (port_id)
1019
1020set bypass event
1021~~~~~~~~~~~~~~~~
1022
1023Set the event required to initiate specified bypass mode for the lowest port on a bypass enabled::
1024
1025   testpmd> set bypass event (timeout|os_on|os_off|power_on|power_off) \
1026            mode (normal|bypass|isolate) (port_id)
1027
1028Where:
1029
1030* ``timeout``: Enable bypass after watchdog timeout.
1031
1032* ``os_on``: Enable bypass when OS/board is powered on.
1033
1034* ``os_off``: Enable bypass when OS/board is powered off.
1035
1036* ``power_on``: Enable bypass when power supply is turned on.
1037
1038* ``power_off``: Enable bypass when power supply is turned off.
1039
1040
1041set bypass timeout
1042~~~~~~~~~~~~~~~~~~
1043
1044Set the bypass watchdog timeout to ``n`` seconds where 0 = instant::
1045
1046   testpmd> set bypass timeout (0|1.5|2|3|4|8|16|32)
1047
1048show bypass config
1049~~~~~~~~~~~~~~~~~~
1050
1051Show the bypass configuration for a bypass enabled NIC using the lowest port on the NIC::
1052
1053   testpmd> show bypass config (port_id)
1054
1055set link up
1056~~~~~~~~~~~
1057
1058Set link up for a port::
1059
1060   testpmd> set link-up port (port id)
1061
1062set link down
1063~~~~~~~~~~~~~
1064
1065Set link down for a port::
1066
1067   testpmd> set link-down port (port id)
1068
1069E-tag set
1070~~~~~~~~~
1071
1072Enable E-tag insertion for a VF on a port::
1073
1074   testpmd> E-tag set insertion on port-tag-id (value) port (port_id) vf (vf_id)
1075
1076Disable E-tag insertion for a VF on a port::
1077
1078   testpmd> E-tag set insertion off port (port_id) vf (vf_id)
1079
1080Enable/disable E-tag stripping on a port::
1081
1082   testpmd> E-tag set stripping (on|off) port (port_id)
1083
1084Enable/disable E-tag based forwarding on a port::
1085
1086   testpmd> E-tag set forwarding (on|off) port (port_id)
1087
1088Add an E-tag forwarding filter on a port::
1089
1090   testpmd> E-tag set filter add e-tag-id (value) dst-pool (pool_id) port (port_id)
1091
1092Delete an E-tag forwarding filter on a port::
1093   testpmd> E-tag set filter del e-tag-id (value) port (port_id)
1094
1095
1096Port Functions
1097--------------
1098
1099The following sections show functions for configuring ports.
1100
1101.. note::
1102
1103   Port configuration changes only become active when forwarding is started/restarted.
1104
1105port attach
1106~~~~~~~~~~~
1107
1108Attach a port specified by pci address or virtual device args::
1109
1110   testpmd> port attach (identifier)
1111
1112To attach a new pci device, the device should be recognized by kernel first.
1113Then it should be moved under DPDK management.
1114Finally the port can be attached to testpmd.
1115
1116For example, to move a pci device using ixgbe under DPDK management:
1117
1118.. code-block:: console
1119
1120   # Check the status of the available devices.
1121   ./usertools/dpdk-devbind.py --status
1122
1123   Network devices using DPDK-compatible driver
1124   ============================================
1125   <none>
1126
1127   Network devices using kernel driver
1128   ===================================
1129   0000:0a:00.0 '82599ES 10-Gigabit' if=eth2 drv=ixgbe unused=
1130
1131
1132   # Bind the device to igb_uio.
1133   sudo ./usertools/dpdk-devbind.py -b igb_uio 0000:0a:00.0
1134
1135
1136   # Recheck the status of the devices.
1137   ./usertools/dpdk-devbind.py --status
1138   Network devices using DPDK-compatible driver
1139   ============================================
1140   0000:0a:00.0 '82599ES 10-Gigabit' drv=igb_uio unused=
1141
1142To attach a port created by virtual device, above steps are not needed.
1143
1144For example, to attach a port whose pci address is 0000:0a:00.0.
1145
1146.. code-block:: console
1147
1148   testpmd> port attach 0000:0a:00.0
1149   Attaching a new port...
1150   EAL: PCI device 0000:0a:00.0 on NUMA socket -1
1151   EAL:   probe driver: 8086:10fb rte_ixgbe_pmd
1152   EAL:   PCI memory mapped at 0x7f83bfa00000
1153   EAL:   PCI memory mapped at 0x7f83bfa80000
1154   PMD: eth_ixgbe_dev_init(): MAC: 2, PHY: 18, SFP+: 5
1155   PMD: eth_ixgbe_dev_init(): port 0 vendorID=0x8086 deviceID=0x10fb
1156   Port 0 is attached. Now total ports is 1
1157   Done
1158
1159For example, to attach a port created by pcap PMD.
1160
1161.. code-block:: console
1162
1163   testpmd> port attach net_pcap0
1164   Attaching a new port...
1165   PMD: Initializing pmd_pcap for net_pcap0
1166   PMD: Creating pcap-backed ethdev on numa socket 0
1167   Port 0 is attached. Now total ports is 1
1168   Done
1169
1170In this case, identifier is ``net_pcap0``.
1171This identifier format is the same as ``--vdev`` format of DPDK applications.
1172
1173For example, to re-attach a bonded port which has been previously detached,
1174the mode and slave parameters must be given.
1175
1176.. code-block:: console
1177
1178   testpmd> port attach net_bond_0,mode=0,slave=1
1179   Attaching a new port...
1180   EAL: Initializing pmd_bond for net_bond_0
1181   EAL: Create bonded device net_bond_0 on port 0 in mode 0 on socket 0.
1182   Port 0 is attached. Now total ports is 1
1183   Done
1184
1185
1186port detach
1187~~~~~~~~~~~
1188
1189Detach a specific port::
1190
1191   testpmd> port detach (port_id)
1192
1193Before detaching a port, the port should be stopped and closed.
1194
1195For example, to detach a pci device port 0.
1196
1197.. code-block:: console
1198
1199   testpmd> port stop 0
1200   Stopping ports...
1201   Done
1202   testpmd> port close 0
1203   Closing ports...
1204   Done
1205
1206   testpmd> port detach 0
1207   Detaching a port...
1208   EAL: PCI device 0000:0a:00.0 on NUMA socket -1
1209   EAL:   remove driver: 8086:10fb rte_ixgbe_pmd
1210   EAL:   PCI memory unmapped at 0x7f83bfa00000
1211   EAL:   PCI memory unmapped at 0x7f83bfa80000
1212   Done
1213
1214
1215For example, to detach a virtual device port 0.
1216
1217.. code-block:: console
1218
1219   testpmd> port stop 0
1220   Stopping ports...
1221   Done
1222   testpmd> port close 0
1223   Closing ports...
1224   Done
1225
1226   testpmd> port detach 0
1227   Detaching a port...
1228   PMD: Closing pcap ethdev on numa socket 0
1229   Port 'net_pcap0' is detached. Now total ports is 0
1230   Done
1231
1232To remove a pci device completely from the system, first detach the port from testpmd.
1233Then the device should be moved under kernel management.
1234Finally the device can be removed using kernel pci hotplug functionality.
1235
1236For example, to move a pci device under kernel management:
1237
1238.. code-block:: console
1239
1240   sudo ./usertools/dpdk-devbind.py -b ixgbe 0000:0a:00.0
1241
1242   ./usertools/dpdk-devbind.py --status
1243
1244   Network devices using DPDK-compatible driver
1245   ============================================
1246   <none>
1247
1248   Network devices using kernel driver
1249   ===================================
1250   0000:0a:00.0 '82599ES 10-Gigabit' if=eth2 drv=ixgbe unused=igb_uio
1251
1252To remove a port created by a virtual device, above steps are not needed.
1253
1254port start
1255~~~~~~~~~~
1256
1257Start all ports or a specific port::
1258
1259   testpmd> port start (port_id|all)
1260
1261port stop
1262~~~~~~~~~
1263
1264Stop all ports or a specific port::
1265
1266   testpmd> port stop (port_id|all)
1267
1268port close
1269~~~~~~~~~~
1270
1271Close all ports or a specific port::
1272
1273   testpmd> port close (port_id|all)
1274
1275port start/stop queue
1276~~~~~~~~~~~~~~~~~~~~~
1277
1278Start/stop a rx/tx queue on a specific port::
1279
1280   testpmd> port (port_id) (rxq|txq) (queue_id) (start|stop)
1281
1282Only take effect when port is started.
1283
1284port config - speed
1285~~~~~~~~~~~~~~~~~~~
1286
1287Set the speed and duplex mode for all ports or a specific port::
1288
1289   testpmd> port config (port_id|all) speed (10|100|1000|10000|25000|40000|50000|100000|auto) \
1290            duplex (half|full|auto)
1291
1292port config - queues/descriptors
1293~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
1294
1295Set number of queues/descriptors for rxq, txq, rxd and txd::
1296
1297   testpmd> port config all (rxq|txq|rxd|txd) (value)
1298
1299This is equivalent to the ``--rxq``, ``--txq``, ``--rxd`` and ``--txd`` command-line options.
1300
1301port config - max-pkt-len
1302~~~~~~~~~~~~~~~~~~~~~~~~~
1303
1304Set the maximum packet length::
1305
1306   testpmd> port config all max-pkt-len (value)
1307
1308This is equivalent to the ``--max-pkt-len`` command-line option.
1309
1310port config - CRC Strip
1311~~~~~~~~~~~~~~~~~~~~~~~
1312
1313Set hardware CRC stripping on or off for all ports::
1314
1315   testpmd> port config all crc-strip (on|off)
1316
1317CRC stripping is off by default.
1318
1319The ``on`` option is equivalent to the ``--crc-strip`` command-line option.
1320
1321port config - scatter
1322~~~~~~~~~~~~~~~~~~~~~~~
1323
1324Set RX scatter mode on or off for all ports::
1325
1326   testpmd> port config all scatter (on|off)
1327
1328RX scatter mode is off by default.
1329
1330The ``on`` option is equivalent to the ``--enable-scatter`` command-line option.
1331
1332port config - TX queue flags
1333~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
1334
1335Set a hexadecimal bitmap of TX queue flags for all ports::
1336
1337   testpmd> port config all txqflags value
1338
1339This command is equivalent to the ``--txqflags`` command-line option.
1340
1341port config - RX Checksum
1342~~~~~~~~~~~~~~~~~~~~~~~~~
1343
1344Set hardware RX checksum offload to on or off for all ports::
1345
1346   testpmd> port config all rx-cksum (on|off)
1347
1348Checksum offload is off by default.
1349
1350The ``on`` option is equivalent to the ``--enable-rx-cksum`` command-line option.
1351
1352port config - VLAN
1353~~~~~~~~~~~~~~~~~~
1354
1355Set hardware VLAN on or off for all ports::
1356
1357   testpmd> port config all hw-vlan (on|off)
1358
1359Hardware VLAN is on by default.
1360
1361The ``off`` option is equivalent to the ``--disable-hw-vlan`` command-line option.
1362
1363port config - VLAN filter
1364~~~~~~~~~~~~~~~~~~~~~~~~~
1365
1366Set hardware VLAN filter on or off for all ports::
1367
1368   testpmd> port config all hw-vlan-filter (on|off)
1369
1370Hardware VLAN filter is on by default.
1371
1372The ``off`` option is equivalent to the ``--disable-hw-vlan-filter`` command-line option.
1373
1374port config - VLAN strip
1375~~~~~~~~~~~~~~~~~~~~~~~~
1376
1377Set hardware VLAN strip on or off for all ports::
1378
1379   testpmd> port config all hw-vlan-strip (on|off)
1380
1381Hardware VLAN strip is on by default.
1382
1383The ``off`` option is equivalent to the ``--disable-hw-vlan-strip`` command-line option.
1384
1385port config - VLAN extend
1386~~~~~~~~~~~~~~~~~~~~~~~~~
1387
1388Set hardware VLAN extend on or off for all ports::
1389
1390   testpmd> port config all hw-vlan-extend (on|off)
1391
1392Hardware VLAN extend is off by default.
1393
1394The ``off`` option is equivalent to the ``--disable-hw-vlan-extend`` command-line option.
1395
1396port config - Drop Packets
1397~~~~~~~~~~~~~~~~~~~~~~~~~~
1398
1399Set packet drop for packets with no descriptors on or off for all ports::
1400
1401   testpmd> port config all drop-en (on|off)
1402
1403Packet dropping for packets with no descriptors is off by default.
1404
1405The ``on`` option is equivalent to the ``--enable-drop-en`` command-line option.
1406
1407port config - RSS
1408~~~~~~~~~~~~~~~~~
1409
1410Set the RSS (Receive Side Scaling) mode on or off::
1411
1412   testpmd> port config all rss (all|ip|tcp|udp|sctp|ether|port|vxlan|geneve|nvgre|none)
1413
1414RSS is on by default.
1415
1416The ``none`` option is equivalent to the ``--disable-rss`` command-line option.
1417
1418port config - RSS Reta
1419~~~~~~~~~~~~~~~~~~~~~~
1420
1421Set the RSS (Receive Side Scaling) redirection table::
1422
1423   testpmd> port config all rss reta (hash,queue)[,(hash,queue)]
1424
1425port config - DCB
1426~~~~~~~~~~~~~~~~~
1427
1428Set the DCB mode for an individual port::
1429
1430   testpmd> port config (port_id) dcb vt (on|off) (traffic_class) pfc (on|off)
1431
1432The traffic class should be 4 or 8.
1433
1434port config - Burst
1435~~~~~~~~~~~~~~~~~~~
1436
1437Set the number of packets per burst::
1438
1439   testpmd> port config all burst (value)
1440
1441This is equivalent to the ``--burst`` command-line option.
1442
1443port config - Threshold
1444~~~~~~~~~~~~~~~~~~~~~~~
1445
1446Set thresholds for TX/RX queues::
1447
1448   testpmd> port config all (threshold) (value)
1449
1450Where the threshold type can be:
1451
1452* ``txpt:`` Set the prefetch threshold register of the TX rings, 0 <= value <= 255.
1453
1454* ``txht:`` Set the host threshold register of the TX rings, 0 <= value <= 255.
1455
1456* ``txwt:`` Set the write-back threshold register of the TX rings, 0 <= value <= 255.
1457
1458* ``rxpt:`` Set the prefetch threshold register of the RX rings, 0 <= value <= 255.
1459
1460* ``rxht:`` Set the host threshold register of the RX rings, 0 <= value <= 255.
1461
1462* ``rxwt:`` Set the write-back threshold register of the RX rings, 0 <= value <= 255.
1463
1464* ``txfreet:`` Set the transmit free threshold of the TX rings, 0 <= value <= txd.
1465
1466* ``rxfreet:`` Set the transmit free threshold of the RX rings, 0 <= value <= rxd.
1467
1468* ``txrst:`` Set the transmit RS bit threshold of TX rings, 0 <= value <= txd.
1469
1470These threshold options are also available from the command-line.
1471
1472port config - E-tag
1473~~~~~~~~~~~~~~~~~~~
1474
1475Set the value of ether-type for E-tag::
1476
1477   testpmd> port config (port_id|all) l2-tunnel E-tag ether-type (value)
1478
1479Enable/disable the E-tag support::
1480
1481   testpmd> port config (port_id|all) l2-tunnel E-tag (enable|disable)
1482
1483
1484Link Bonding Functions
1485----------------------
1486
1487The Link Bonding functions make it possible to dynamically create and
1488manage link bonding devices from within testpmd interactive prompt.
1489
1490create bonded device
1491~~~~~~~~~~~~~~~~~~~~
1492
1493Create a new bonding device::
1494
1495   testpmd> create bonded device (mode) (socket)
1496
1497For example, to create a bonded device in mode 1 on socket 0::
1498
1499   testpmd> create bonded 1 0
1500   created new bonded device (port X)
1501
1502add bonding slave
1503~~~~~~~~~~~~~~~~~
1504
1505Adds Ethernet device to a Link Bonding device::
1506
1507   testpmd> add bonding slave (slave id) (port id)
1508
1509For example, to add Ethernet device (port 6) to a Link Bonding device (port 10)::
1510
1511   testpmd> add bonding slave 6 10
1512
1513
1514remove bonding slave
1515~~~~~~~~~~~~~~~~~~~~
1516
1517Removes an Ethernet slave device from a Link Bonding device::
1518
1519   testpmd> remove bonding slave (slave id) (port id)
1520
1521For example, to remove Ethernet slave device (port 6) to a Link Bonding device (port 10)::
1522
1523   testpmd> remove bonding slave 6 10
1524
1525set bonding mode
1526~~~~~~~~~~~~~~~~
1527
1528Set the Link Bonding mode of a Link Bonding device::
1529
1530   testpmd> set bonding mode (value) (port id)
1531
1532For example, to set the bonding mode of a Link Bonding device (port 10) to broadcast (mode 3)::
1533
1534   testpmd> set bonding mode 3 10
1535
1536set bonding primary
1537~~~~~~~~~~~~~~~~~~~
1538
1539Set an Ethernet slave device as the primary device on a Link Bonding device::
1540
1541   testpmd> set bonding primary (slave id) (port id)
1542
1543For example, to set the Ethernet slave device (port 6) as the primary port of a Link Bonding device (port 10)::
1544
1545   testpmd> set bonding primary 6 10
1546
1547set bonding mac
1548~~~~~~~~~~~~~~~
1549
1550Set the MAC address of a Link Bonding device::
1551
1552   testpmd> set bonding mac (port id) (mac)
1553
1554For example, to set the MAC address of a Link Bonding device (port 10) to 00:00:00:00:00:01::
1555
1556   testpmd> set bonding mac 10 00:00:00:00:00:01
1557
1558set bonding xmit_balance_policy
1559~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
1560
1561Set the transmission policy for a Link Bonding device when it is in Balance XOR mode::
1562
1563   testpmd> set bonding xmit_balance_policy (port_id) (l2|l23|l34)
1564
1565For example, set a Link Bonding device (port 10) to use a balance policy of layer 3+4 (IP addresses & UDP ports)::
1566
1567   testpmd> set bonding xmit_balance_policy 10 l34
1568
1569
1570set bonding mon_period
1571~~~~~~~~~~~~~~~~~~~~~~
1572
1573Set the link status monitoring polling period in milliseconds for a bonding device.
1574
1575This adds support for PMD slave devices which do not support link status interrupts.
1576When the mon_period is set to a value greater than 0 then all PMD's which do not support
1577link status ISR will be queried every polling interval to check if their link status has changed::
1578
1579   testpmd> set bonding mon_period (port_id) (value)
1580
1581For example, to set the link status monitoring polling period of bonded device (port 5) to 150ms::
1582
1583   testpmd> set bonding mon_period 5 150
1584
1585
1586show bonding config
1587~~~~~~~~~~~~~~~~~~~
1588
1589Show the current configuration of a Link Bonding device::
1590
1591   testpmd> show bonding config (port id)
1592
1593For example,
1594to show the configuration a Link Bonding device (port 9) with 3 slave devices (1, 3, 4)
1595in balance mode with a transmission policy of layer 2+3::
1596
1597   testpmd> show bonding config 9
1598        Bonding mode: 2
1599        Balance Xmit Policy: BALANCE_XMIT_POLICY_LAYER23
1600        Slaves (3): [1 3 4]
1601        Active Slaves (3): [1 3 4]
1602        Primary: [3]
1603
1604
1605Register Functions
1606------------------
1607
1608The Register Functions can be used to read from and write to registers on the network card referenced by a port number.
1609This is mainly useful for debugging purposes.
1610Reference should be made to the appropriate datasheet for the network card for details on the register addresses
1611and fields that can be accessed.
1612
1613read reg
1614~~~~~~~~
1615
1616Display the value of a port register::
1617
1618   testpmd> read reg (port_id) (address)
1619
1620For example, to examine the Flow Director control register (FDIRCTL, 0x0000EE000) on an Intel 82599 10 GbE Controller::
1621
1622   testpmd> read reg 0 0xEE00
1623   port 0 PCI register at offset 0xEE00: 0x4A060029 (1241907241)
1624
1625read regfield
1626~~~~~~~~~~~~~
1627
1628Display a port register bit field::
1629
1630   testpmd> read regfield (port_id) (address) (bit_x) (bit_y)
1631
1632For example, reading the lowest two bits from the register in the example above::
1633
1634   testpmd> read regfield 0 0xEE00 0 1
1635   port 0 PCI register at offset 0xEE00: bits[0, 1]=0x1 (1)
1636
1637read regbit
1638~~~~~~~~~~~
1639
1640Display a single port register bit::
1641
1642   testpmd> read regbit (port_id) (address) (bit_x)
1643
1644For example, reading the lowest bit from the register in the example above::
1645
1646   testpmd> read regbit 0 0xEE00 0
1647   port 0 PCI register at offset 0xEE00: bit 0=1
1648
1649write reg
1650~~~~~~~~~
1651
1652Set the value of a port register::
1653
1654   testpmd> write reg (port_id) (address) (value)
1655
1656For example, to clear a register::
1657
1658   testpmd> write reg 0 0xEE00 0x0
1659   port 0 PCI register at offset 0xEE00: 0x00000000 (0)
1660
1661write regfield
1662~~~~~~~~~~~~~~
1663
1664Set bit field of a port register::
1665
1666   testpmd> write regfield (port_id) (address) (bit_x) (bit_y) (value)
1667
1668For example, writing to the register cleared in the example above::
1669
1670   testpmd> write regfield 0 0xEE00 0 1 2
1671   port 0 PCI register at offset 0xEE00: 0x00000002 (2)
1672
1673write regbit
1674~~~~~~~~~~~~
1675
1676Set single bit value of a port register::
1677
1678   testpmd> write regbit (port_id) (address) (bit_x) (value)
1679
1680For example, to set the high bit in the register from the example above::
1681
1682   testpmd> write regbit 0 0xEE00 31 1
1683   port 0 PCI register at offset 0xEE00: 0x8000000A (2147483658)
1684
1685
1686Filter Functions
1687----------------
1688
1689This section details the available filter functions that are available.
1690
1691Note these functions interface the deprecated legacy filtering framework,
1692superseded by *rte_flow*. See `Flow rules management`_.
1693
1694ethertype_filter
1695~~~~~~~~~~~~~~~~~~~~
1696
1697Add or delete a L2 Ethertype filter, which identify packets by their L2 Ethertype mainly assign them to a receive queue::
1698
1699   ethertype_filter (port_id) (add|del) (mac_addr|mac_ignr) (mac_address) \
1700                    ethertype (ether_type) (drop|fwd) queue (queue_id)
1701
1702The available information parameters are:
1703
1704* ``port_id``: The port which the Ethertype filter assigned on.
1705
1706* ``mac_addr``: Compare destination mac address.
1707
1708* ``mac_ignr``: Ignore destination mac address match.
1709
1710* ``mac_address``: Destination mac address to match.
1711
1712* ``ether_type``: The EtherType value want to match,
1713  for example 0x0806 for ARP packet. 0x0800 (IPv4) and 0x86DD (IPv6) are invalid.
1714
1715* ``queue_id``: The receive queue associated with this EtherType filter.
1716  It is meaningless when deleting or dropping.
1717
1718Example, to add/remove an ethertype filter rule::
1719
1720   testpmd> ethertype_filter 0 add mac_ignr 00:11:22:33:44:55 \
1721                             ethertype 0x0806 fwd queue 3
1722
1723   testpmd> ethertype_filter 0 del mac_ignr 00:11:22:33:44:55 \
1724                             ethertype 0x0806 fwd queue 3
1725
17262tuple_filter
1727~~~~~~~~~~~~~~~~~
1728
1729Add or delete a 2-tuple filter,
1730which identifies packets by specific protocol and destination TCP/UDP port
1731and forwards packets into one of the receive queues::
1732
1733   2tuple_filter (port_id) (add|del) dst_port (dst_port_value) \
1734                 protocol (protocol_value) mask (mask_value) \
1735                 tcp_flags (tcp_flags_value) priority (prio_value) \
1736                 queue (queue_id)
1737
1738The available information parameters are:
1739
1740* ``port_id``: The port which the 2-tuple filter assigned on.
1741
1742* ``dst_port_value``: Destination port in L4.
1743
1744* ``protocol_value``: IP L4 protocol.
1745
1746* ``mask_value``: Participates in the match or not by bit for field above, 1b means participate.
1747
1748* ``tcp_flags_value``: TCP control bits. The non-zero value is invalid, when the pro_value is not set to 0x06 (TCP).
1749
1750* ``prio_value``: Priority of this filter.
1751
1752* ``queue_id``: The receive queue associated with this 2-tuple filter.
1753
1754Example, to add/remove an 2tuple filter rule::
1755
1756   testpmd> 2tuple_filter 0 add dst_port 32 protocol 0x06 mask 0x03 \
1757                          tcp_flags 0x02 priority 3 queue 3
1758
1759   testpmd> 2tuple_filter 0 del dst_port 32 protocol 0x06 mask 0x03 \
1760                          tcp_flags 0x02 priority 3 queue 3
1761
17625tuple_filter
1763~~~~~~~~~~~~~~~~~
1764
1765Add or delete a 5-tuple filter,
1766which consists of a 5-tuple (protocol, source and destination IP addresses, source and destination TCP/UDP/SCTP port)
1767and routes packets into one of the receive queues::
1768
1769   5tuple_filter (port_id) (add|del) dst_ip (dst_address) src_ip \
1770                 (src_address) dst_port (dst_port_value) \
1771                 src_port (src_port_value) protocol (protocol_value) \
1772                 mask (mask_value) tcp_flags (tcp_flags_value) \
1773                 priority (prio_value) queue (queue_id)
1774
1775The available information parameters are:
1776
1777* ``port_id``: The port which the 5-tuple filter assigned on.
1778
1779* ``dst_address``: Destination IP address.
1780
1781* ``src_address``: Source IP address.
1782
1783* ``dst_port_value``: TCP/UDP destination port.
1784
1785* ``src_port_value``: TCP/UDP source port.
1786
1787* ``protocol_value``: L4 protocol.
1788
1789* ``mask_value``: Participates in the match or not by bit for field above, 1b means participate
1790
1791* ``tcp_flags_value``: TCP control bits. The non-zero value is invalid, when the protocol_value is not set to 0x06 (TCP).
1792
1793* ``prio_value``: The priority of this filter.
1794
1795* ``queue_id``: The receive queue associated with this 5-tuple filter.
1796
1797Example, to add/remove an 5tuple filter rule::
1798
1799   testpmd> 5tuple_filter 0 add dst_ip 2.2.2.5 src_ip 2.2.2.4 \
1800            dst_port 64 src_port 32 protocol 0x06 mask 0x1F \
1801            flags 0x0 priority 3 queue 3
1802
1803   testpmd> 5tuple_filter 0 del dst_ip 2.2.2.5 src_ip 2.2.2.4 \
1804            dst_port 64 src_port 32 protocol 0x06 mask 0x1F \
1805            flags 0x0 priority 3 queue 3
1806
1807syn_filter
1808~~~~~~~~~~
1809
1810Using the  SYN filter, TCP packets whose *SYN* flag is set can be forwarded to a separate queue::
1811
1812   syn_filter (port_id) (add|del) priority (high|low) queue (queue_id)
1813
1814The available information parameters are:
1815
1816* ``port_id``: The port which the SYN filter assigned on.
1817
1818* ``high``: This SYN filter has higher priority than other filters.
1819
1820* ``low``: This SYN filter has lower priority than other filters.
1821
1822* ``queue_id``: The receive queue associated with this SYN filter
1823
1824Example::
1825
1826   testpmd> syn_filter 0 add priority high queue 3
1827
1828flex_filter
1829~~~~~~~~~~~
1830
1831With flex filter, packets can be recognized by any arbitrary pattern within the first 128 bytes of the packet
1832and routed into one of the receive queues::
1833
1834   flex_filter (port_id) (add|del) len (len_value) bytes (bytes_value) \
1835               mask (mask_value) priority (prio_value) queue (queue_id)
1836
1837The available information parameters are:
1838
1839* ``port_id``: The port which the Flex filter is assigned on.
1840
1841* ``len_value``: Filter length in bytes, no greater than 128.
1842
1843* ``bytes_value``: A string in hexadecimal, means the value the flex filter needs to match.
1844
1845* ``mask_value``: A string in hexadecimal, bit 1 means corresponding byte participates in the match.
1846
1847* ``prio_value``: The priority of this filter.
1848
1849* ``queue_id``: The receive queue associated with this Flex filter.
1850
1851Example::
1852
1853   testpmd> flex_filter 0 add len 16 bytes 0x00000000000000000000000008060000 \
1854                          mask 000C priority 3 queue 3
1855
1856   testpmd> flex_filter 0 del len 16 bytes 0x00000000000000000000000008060000 \
1857                          mask 000C priority 3 queue 3
1858
1859
1860.. _testpmd_flow_director:
1861
1862flow_director_filter
1863~~~~~~~~~~~~~~~~~~~~
1864
1865The Flow Director works in receive mode to identify specific flows or sets of flows and route them to specific queues.
1866
1867Four types of filtering are supported which are referred to as Perfect Match, Signature, Perfect-mac-vlan and
1868Perfect-tunnel filters, the match mode is set by the ``--pkt-filter-mode`` command-line parameter:
1869
1870* Perfect match filters.
1871  The hardware checks a match between the masked fields of the received packets and the programmed filters.
1872  The masked fields are for IP flow.
1873
1874* Signature filters.
1875  The hardware checks a match between a hash-based signature of the masked fields of the received packet.
1876
1877* Perfect-mac-vlan match filters.
1878  The hardware checks a match between the masked fields of the received packets and the programmed filters.
1879  The masked fields are for MAC VLAN flow.
1880
1881* Perfect-tunnel match filters.
1882  The hardware checks a match between the masked fields of the received packets and the programmed filters.
1883  The masked fields are for tunnel flow.
1884
1885The Flow Director filters can match the different fields for different type of packet: flow type, specific input set
1886per flow type and the flexible payload.
1887
1888The Flow Director can also mask out parts of all of these fields so that filters
1889are only applied to certain fields or parts of the fields.
1890
1891Different NICs may have different capabilities, command show port fdir (port_id) can be used to acquire the information.
1892
1893# Commands to add flow director filters of different flow types::
1894
1895   flow_director_filter (port_id) mode IP (add|del|update) \
1896                        flow (ipv4-other|ipv4-frag|ipv6-other|ipv6-frag) \
1897                        src (src_ip_address) dst (dst_ip_address) \
1898                        tos (tos_value) proto (proto_value) ttl (ttl_value) \
1899                        vlan (vlan_value) flexbytes (flexbytes_value) \
1900                        (drop|fwd) pf|vf(vf_id) queue (queue_id) \
1901                        fd_id (fd_id_value)
1902
1903   flow_director_filter (port_id) mode IP (add|del|update) \
1904                        flow (ipv4-tcp|ipv4-udp|ipv6-tcp|ipv6-udp) \
1905                        src (src_ip_address) (src_port) \
1906                        dst (dst_ip_address) (dst_port) \
1907                        tos (tos_value) ttl (ttl_value) \
1908                        vlan (vlan_value) flexbytes (flexbytes_value) \
1909                        (drop|fwd) queue pf|vf(vf_id) (queue_id) \
1910                        fd_id (fd_id_value)
1911
1912   flow_director_filter (port_id) mode IP (add|del|update) \
1913                        flow (ipv4-sctp|ipv6-sctp) \
1914                        src (src_ip_address) (src_port) \
1915                        dst (dst_ip_address) (dst_port) \
1916                        tos (tos_value) ttl (ttl_value) \
1917                        tag (verification_tag) vlan (vlan_value) \
1918                        flexbytes (flexbytes_value) (drop|fwd) \
1919                        pf|vf(vf_id) queue (queue_id) fd_id (fd_id_value)
1920
1921   flow_director_filter (port_id) mode IP (add|del|update) flow l2_payload \
1922                        ether (ethertype) 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 MAC-VLAN (add|del|update) \
1927                        mac (mac_address) vlan (vlan_value) \
1928                        flexbytes (flexbytes_value) (drop|fwd) \
1929                        queue (queue_id) fd_id (fd_id_value)
1930
1931   flow_director_filter (port_id) mode Tunnel (add|del|update) \
1932                        mac (mac_address) vlan (vlan_value) \
1933                        tunnel (NVGRE|VxLAN) tunnel-id (tunnel_id_value) \
1934                        flexbytes (flexbytes_value) (drop|fwd) \
1935                        queue (queue_id) fd_id (fd_id_value)
1936
1937For example, to add an ipv4-udp flow type filter::
1938
1939   testpmd> flow_director_filter 0 mode IP add flow ipv4-udp src 2.2.2.3 32 \
1940            dst 2.2.2.5 33 tos 2 ttl 40 vlan 0x1 flexbytes (0x88,0x48) \
1941            fwd pf queue 1 fd_id 1
1942
1943For example, add an ipv4-other flow type filter::
1944
1945   testpmd> flow_director_filter 0 mode IP add flow ipv4-other src 2.2.2.3 \
1946             dst 2.2.2.5 tos 2 proto 20 ttl 40 vlan 0x1 \
1947             flexbytes (0x88,0x48) fwd pf queue 1 fd_id 1
1948
1949flush_flow_director
1950~~~~~~~~~~~~~~~~~~~
1951
1952Flush all flow director filters on a device::
1953
1954   testpmd> flush_flow_director (port_id)
1955
1956Example, to flush all flow director filter on port 0::
1957
1958   testpmd> flush_flow_director 0
1959
1960flow_director_mask
1961~~~~~~~~~~~~~~~~~~
1962
1963Set flow director's input masks::
1964
1965   flow_director_mask (port_id) mode IP vlan (vlan_value) \
1966                      src_mask (ipv4_src) (ipv6_src) (src_port) \
1967                      dst_mask (ipv4_dst) (ipv6_dst) (dst_port)
1968
1969   flow_director_mask (port_id) mode MAC-VLAN vlan (vlan_value)
1970
1971   flow_director_mask (port_id) mode Tunnel vlan (vlan_value) \
1972                      mac (mac_value) tunnel-type (tunnel_type_value) \
1973                      tunnel-id (tunnel_id_value)
1974
1975Example, to set flow director mask on port 0::
1976
1977   testpmd> flow_director_mask 0 mode IP vlan 0xefff \
1978            src_mask 255.255.255.255 \
1979                FFFF:FFFF:FFFF:FFFF:FFFF:FFFF:FFFF:FFFF 0xFFFF \
1980            dst_mask 255.255.255.255 \
1981                FFFF:FFFF:FFFF:FFFF:FFFF:FFFF:FFFF:FFFF 0xFFFF
1982
1983flow_director_flex_mask
1984~~~~~~~~~~~~~~~~~~~~~~~
1985
1986set masks of flow director's flexible payload based on certain flow type::
1987
1988   testpmd> flow_director_flex_mask (port_id) \
1989            flow (none|ipv4-other|ipv4-frag|ipv4-tcp|ipv4-udp|ipv4-sctp| \
1990                  ipv6-other|ipv6-frag|ipv6-tcp|ipv6-udp|ipv6-sctp| \
1991                  l2_payload|all) (mask)
1992
1993Example, to set flow director's flex mask for all flow type on port 0::
1994
1995   testpmd> flow_director_flex_mask 0 flow all \
1996            (0xff,0xff,0,0,0,0,0,0,0,0,0,0,0,0,0,0)
1997
1998
1999flow_director_flex_payload
2000~~~~~~~~~~~~~~~~~~~~~~~~~~
2001
2002Configure flexible payload selection::
2003
2004   flow_director_flex_payload (port_id) (raw|l2|l3|l4) (config)
2005
2006For example, to select the first 16 bytes from the offset 4 (bytes) of packet's payload as flexible payload::
2007
2008   testpmd> flow_director_flex_payload 0 l4 \
2009            (4,5,6,7,8,9,10,11,12,13,14,15,16,17,18,19)
2010
2011get_sym_hash_ena_per_port
2012~~~~~~~~~~~~~~~~~~~~~~~~~
2013
2014Get symmetric hash enable configuration per port::
2015
2016   get_sym_hash_ena_per_port (port_id)
2017
2018For example, to get symmetric hash enable configuration of port 1::
2019
2020   testpmd> get_sym_hash_ena_per_port 1
2021
2022set_sym_hash_ena_per_port
2023~~~~~~~~~~~~~~~~~~~~~~~~~
2024
2025Set symmetric hash enable configuration per port to enable or disable::
2026
2027   set_sym_hash_ena_per_port (port_id) (enable|disable)
2028
2029For example, to set symmetric hash enable configuration of port 1 to enable::
2030
2031   testpmd> set_sym_hash_ena_per_port 1 enable
2032
2033get_hash_global_config
2034~~~~~~~~~~~~~~~~~~~~~~
2035
2036Get the global configurations of hash filters::
2037
2038   get_hash_global_config (port_id)
2039
2040For example, to get the global configurations of hash filters of port 1::
2041
2042   testpmd> get_hash_global_config 1
2043
2044set_hash_global_config
2045~~~~~~~~~~~~~~~~~~~~~~
2046
2047Set the global configurations of hash filters::
2048
2049   set_hash_global_config (port_id) (toeplitz|simple_xor|default) \
2050   (ipv4|ipv4-frag|ipv4-tcp|ipv4-udp|ipv4-sctp|ipv4-other|ipv6|ipv6-frag| \
2051   ipv6-tcp|ipv6-udp|ipv6-sctp|ipv6-other|l2_payload) \
2052   (enable|disable)
2053
2054For example, to enable simple_xor for flow type of ipv6 on port 2::
2055
2056   testpmd> set_hash_global_config 2 simple_xor ipv6 enable
2057
2058set_hash_input_set
2059~~~~~~~~~~~~~~~~~~
2060
2061Set the input set for hash::
2062
2063   set_hash_input_set (port_id) (ipv4-frag|ipv4-tcp|ipv4-udp|ipv4-sctp| \
2064   ipv4-other|ipv6-frag|ipv6-tcp|ipv6-udp|ipv6-sctp|ipv6-other| \
2065   l2_payload) (ovlan|ivlan|src-ipv4|dst-ipv4|src-ipv6|dst-ipv6|ipv4-tos| \
2066   ipv4-proto|ipv6-tc|ipv6-next-header|udp-src-port|udp-dst-port| \
2067   tcp-src-port|tcp-dst-port|sctp-src-port|sctp-dst-port|sctp-veri-tag| \
2068   udp-key|gre-key|fld-1st|fld-2nd|fld-3rd|fld-4th|fld-5th|fld-6th|fld-7th| \
2069   fld-8th|none) (select|add)
2070
2071For example, to add source IP to hash input set for flow type of ipv4-udp on port 0::
2072
2073   testpmd> set_hash_input_set 0 ipv4-udp src-ipv4 add
2074
2075set_fdir_input_set
2076~~~~~~~~~~~~~~~~~~
2077
2078The Flow Director filters can match the different fields for different type of packet, i.e. specific input set
2079on per flow type and the flexible payload. This command can be used to change input set for each flow type.
2080
2081Set the input set for flow director::
2082
2083   set_fdir_input_set (port_id) (ipv4-frag|ipv4-tcp|ipv4-udp|ipv4-sctp| \
2084   ipv4-other|ipv6|ipv6-frag|ipv6-tcp|ipv6-udp|ipv6-sctp|ipv6-other| \
2085   l2_payload) (ivlan|ethertype|src-ipv4|dst-ipv4|src-ipv6|dst-ipv6|ipv4-tos| \
2086   ipv4-proto|ipv4-ttl|ipv6-tc|ipv6-next-header|ipv6-hop-limits| \
2087   tudp-src-port|udp-dst-port|cp-src-port|tcp-dst-port|sctp-src-port| \
2088   sctp-dst-port|sctp-veri-tag|none) (select|add)
2089
2090For example to add source IP to FD input set for flow type of ipv4-udp on port 0::
2091
2092   testpmd> set_fdir_input_set 0 ipv4-udp src-ipv4 add
2093
2094global_config
2095~~~~~~~~~~~~~
2096
2097Set different GRE key length for input set::
2098
2099   global_config (port_id) gre-key-len (number in bytes)
2100
2101For example to set GRE key length for input set to 4 bytes on port 0::
2102
2103   testpmd> global_config 0 gre-key-len 4
2104
2105
2106.. _testpmd_rte_flow:
2107
2108Flow rules management
2109---------------------
2110
2111Control of the generic flow API (*rte_flow*) is fully exposed through the
2112``flow`` command (validation, creation, destruction and queries).
2113
2114Considering *rte_flow* overlaps with all `Filter Functions`_, using both
2115features simultaneously may cause undefined side-effects and is therefore
2116not recommended.
2117
2118``flow`` syntax
2119~~~~~~~~~~~~~~~
2120
2121Because the ``flow`` command uses dynamic tokens to handle the large number
2122of possible flow rules combinations, its behavior differs slightly from
2123other commands, in particular:
2124
2125- Pressing *?* or the *<tab>* key displays contextual help for the current
2126  token, not that of the entire command.
2127
2128- Optional and repeated parameters are supported (provided they are listed
2129  in the contextual help).
2130
2131The first parameter stands for the operation mode. Possible operations and
2132their general syntax are described below. They are covered in detail in the
2133following sections.
2134
2135- Check whether a flow rule can be created::
2136
2137   flow validate {port_id}
2138       [group {group_id}] [priority {level}] [ingress] [egress]
2139       pattern {item} [/ {item} [...]] / end
2140       actions {action} [/ {action} [...]] / end
2141
2142- Create a flow rule::
2143
2144   flow create {port_id}
2145       [group {group_id}] [priority {level}] [ingress] [egress]
2146       pattern {item} [/ {item} [...]] / end
2147       actions {action} [/ {action} [...]] / end
2148
2149- Destroy specific flow rules::
2150
2151   flow destroy {port_id} rule {rule_id} [...]
2152
2153- Destroy all flow rules::
2154
2155   flow flush {port_id}
2156
2157- Query an existing flow rule::
2158
2159   flow query {port_id} {rule_id} {action}
2160
2161- List existing flow rules sorted by priority, filtered by group
2162  identifiers::
2163
2164   flow list {port_id} [group {group_id}] [...]
2165
2166Validating flow rules
2167~~~~~~~~~~~~~~~~~~~~~
2168
2169``flow validate`` reports whether a flow rule would be accepted by the
2170underlying device in its current state but stops short of creating it. It is
2171bound to ``rte_flow_validate()``::
2172
2173   flow validate {port_id}
2174      [group {group_id}] [priority {level}] [ingress] [egress]
2175      pattern {item} [/ {item} [...]] / end
2176      actions {action} [/ {action} [...]] / end
2177
2178If successful, it will show::
2179
2180   Flow rule validated
2181
2182Otherwise it will show an error message of the form::
2183
2184   Caught error type [...] ([...]): [...]
2185
2186This command uses the same parameters as ``flow create``, their format is
2187described in `Creating flow rules`_.
2188
2189Check whether redirecting any Ethernet packet received on port 0 to RX queue
2190index 6 is supported::
2191
2192   testpmd> flow validate 0 ingress pattern eth / end
2193      actions queue index 6 / end
2194   Flow rule validated
2195   testpmd>
2196
2197Port 0 does not support TCPv6 rules::
2198
2199   testpmd> flow validate 0 ingress pattern eth / ipv6 / tcp / end
2200      actions drop / end
2201   Caught error type 9 (specific pattern item): Invalid argument
2202   testpmd>
2203
2204Creating flow rules
2205~~~~~~~~~~~~~~~~~~~
2206
2207``flow create`` validates and creates the specified flow rule. It is bound
2208to ``rte_flow_create()``::
2209
2210   flow create {port_id}
2211      [group {group_id}] [priority {level}] [ingress] [egress]
2212      pattern {item} [/ {item} [...]] / end
2213      actions {action} [/ {action} [...]] / end
2214
2215If successful, it will return a flow rule ID usable with other commands::
2216
2217   Flow rule #[...] created
2218
2219Otherwise it will show an error message of the form::
2220
2221   Caught error type [...] ([...]): [...]
2222
2223Parameters describe in the following order:
2224
2225- Attributes (*group*, *priority*, *ingress*, *egress* tokens).
2226- A matching pattern, starting with the *pattern* token and terminated by an
2227  *end* pattern item.
2228- Actions, starting with the *actions* token and terminated by an *end*
2229  action.
2230
2231These translate directly to *rte_flow* objects provided as-is to the
2232underlying functions.
2233
2234The shortest valid definition only comprises mandatory tokens::
2235
2236   testpmd> flow create 0 pattern end actions end
2237
2238Note that PMDs may refuse rules that essentially do nothing such as this
2239one.
2240
2241**All unspecified object values are automatically initialized to 0.**
2242
2243Attributes
2244^^^^^^^^^^
2245
2246These tokens affect flow rule attributes (``struct rte_flow_attr``) and are
2247specified before the ``pattern`` token.
2248
2249- ``group {group id}``: priority group.
2250- ``priority {level}``: priority level within group.
2251- ``ingress``: rule applies to ingress traffic.
2252- ``egress``: rule applies to egress traffic.
2253
2254Each instance of an attribute specified several times overrides the previous
2255value as shown below (group 4 is used)::
2256
2257   testpmd> flow create 0 group 42 group 24 group 4 [...]
2258
2259Note that once enabled, ``ingress`` and ``egress`` cannot be disabled.
2260
2261While not specifying a direction is an error, some rules may allow both
2262simultaneously.
2263
2264Most rules affect RX therefore contain the ``ingress`` token::
2265
2266   testpmd> flow create 0 ingress pattern [...]
2267
2268Matching pattern
2269^^^^^^^^^^^^^^^^
2270
2271A matching pattern starts after the ``pattern`` token. It is made of pattern
2272items and is terminated by a mandatory ``end`` item.
2273
2274Items are named after their type (*RTE_FLOW_ITEM_TYPE_* from ``enum
2275rte_flow_item_type``).
2276
2277The ``/`` token is used as a separator between pattern items as shown
2278below::
2279
2280   testpmd> flow create 0 ingress pattern eth / ipv4 / udp / end [...]
2281
2282Note that protocol items like these must be stacked from lowest to highest
2283layer to make sense. For instance, the following rule is either invalid or
2284unlikely to match any packet::
2285
2286   testpmd> flow create 0 ingress pattern eth / udp / ipv4 / end [...]
2287
2288More information on these restrictions can be found in the *rte_flow*
2289documentation.
2290
2291Several items support additional specification structures, for example
2292``ipv4`` allows specifying source and destination addresses as follows::
2293
2294   testpmd> flow create 0 ingress pattern eth / ipv4 src is 10.1.1.1
2295      dst is 10.2.0.0 / end [...]
2296
2297This rule matches all IPv4 traffic with the specified properties.
2298
2299In this example, ``src`` and ``dst`` are field names of the underlying
2300``struct rte_flow_item_ipv4`` object. All item properties can be specified
2301in a similar fashion.
2302
2303The ``is`` token means that the subsequent value must be matched exactly,
2304and assigns ``spec`` and ``mask`` fields in ``struct rte_flow_item``
2305accordingly. Possible assignment tokens are:
2306
2307- ``is``: match value perfectly (with full bit-mask).
2308- ``spec``: match value according to configured bit-mask.
2309- ``last``: specify upper bound to establish a range.
2310- ``mask``: specify bit-mask with relevant bits set to one.
2311- ``prefix``: generate bit-mask from a prefix length.
2312
2313These yield identical results::
2314
2315   ipv4 src is 10.1.1.1
2316
2317::
2318
2319   ipv4 src spec 10.1.1.1 src mask 255.255.255.255
2320
2321::
2322
2323   ipv4 src spec 10.1.1.1 src prefix 32
2324
2325::
2326
2327   ipv4 src is 10.1.1.1 src last 10.1.1.1 # range with a single value
2328
2329::
2330
2331   ipv4 src is 10.1.1.1 src last 0 # 0 disables range
2332
2333Inclusive ranges can be defined with ``last``::
2334
2335   ipv4 src is 10.1.1.1 src last 10.2.3.4 # 10.1.1.1 to 10.2.3.4
2336
2337Note that ``mask`` affects both ``spec`` and ``last``::
2338
2339   ipv4 src is 10.1.1.1 src last 10.2.3.4 src mask 255.255.0.0
2340      # matches 10.1.0.0 to 10.2.255.255
2341
2342Properties can be modified multiple times::
2343
2344   ipv4 src is 10.1.1.1 src is 10.1.2.3 src is 10.2.3.4 # matches 10.2.3.4
2345
2346::
2347
2348   ipv4 src is 10.1.1.1 src prefix 24 src prefix 16 # matches 10.1.0.0/16
2349
2350Pattern items
2351^^^^^^^^^^^^^
2352
2353This section lists supported pattern items and their attributes, if any.
2354
2355- ``end``: end list of pattern items.
2356
2357- ``void``: no-op pattern item.
2358
2359- ``invert``: perform actions when pattern does not match.
2360
2361- ``any``: match any protocol for the current layer.
2362
2363  - ``num {unsigned}``: number of layers covered.
2364
2365- ``pf``: match packets addressed to the physical function.
2366
2367- ``vf``: match packets addressed to a virtual function ID.
2368
2369  - ``id {unsigned}``: destination VF ID.
2370
2371- ``port``: device-specific physical port index to use.
2372
2373  - ``index {unsigned}``: physical port index.
2374
2375- ``raw``: match an arbitrary byte string.
2376
2377  - ``relative {boolean}``: look for pattern after the previous item.
2378  - ``search {boolean}``: search pattern from offset (see also limit).
2379  - ``offset {integer}``: absolute or relative offset for pattern.
2380  - ``limit {unsigned}``: search area limit for start of pattern.
2381  - ``pattern {string}``: byte string to look for.
2382
2383- ``eth``: match Ethernet header.
2384
2385  - ``dst {MAC-48}``: destination MAC.
2386  - ``src {MAC-48}``: source MAC.
2387  - ``type {unsigned}``: EtherType.
2388
2389- ``vlan``: match 802.1Q/ad VLAN tag.
2390
2391  - ``tpid {unsigned}``: tag protocol identifier.
2392  - ``tci {unsigned}``: tag control information.
2393  - ``pcp {unsigned}``: priority code point.
2394  - ``dei {unsigned}``: drop eligible indicator.
2395  - ``vid {unsigned}``: VLAN identifier.
2396
2397- ``ipv4``: match IPv4 header.
2398
2399  - ``tos {unsigned}``: type of service.
2400  - ``ttl {unsigned}``: time to live.
2401  - ``proto {unsigned}``: next protocol ID.
2402  - ``src {ipv4 address}``: source address.
2403  - ``dst {ipv4 address}``: destination address.
2404
2405- ``ipv6``: match IPv6 header.
2406
2407  - ``tc {unsigned}``: traffic class.
2408  - ``flow {unsigned}``: flow label.
2409  - ``proto {unsigned}``: protocol (next header).
2410  - ``hop {unsigned}``: hop limit.
2411  - ``src {ipv6 address}``: source address.
2412  - ``dst {ipv6 address}``: destination address.
2413
2414- ``icmp``: match ICMP header.
2415
2416  - ``type {unsigned}``: ICMP packet type.
2417  - ``code {unsigned}``: ICMP packet code.
2418
2419- ``udp``: match UDP header.
2420
2421  - ``src {unsigned}``: UDP source port.
2422  - ``dst {unsigned}``: UDP destination port.
2423
2424- ``tcp``: match TCP header.
2425
2426  - ``src {unsigned}``: TCP source port.
2427  - ``dst {unsigned}``: TCP destination port.
2428
2429- ``sctp``: match SCTP header.
2430
2431  - ``src {unsigned}``: SCTP source port.
2432  - ``dst {unsigned}``: SCTP destination port.
2433  - ``tag {unsigned}``: validation tag.
2434  - ``cksum {unsigned}``: checksum.
2435
2436- ``vxlan``: match VXLAN header.
2437
2438  - ``vni {unsigned}``: VXLAN identifier.
2439
2440Actions list
2441^^^^^^^^^^^^
2442
2443A list of actions starts after the ``actions`` token in the same fashion as
2444`Matching pattern`_; actions are separated by ``/`` tokens and the list is
2445terminated by a mandatory ``end`` action.
2446
2447Actions are named after their type (*RTE_FLOW_ACTION_TYPE_* from ``enum
2448rte_flow_action_type``).
2449
2450Dropping all incoming UDPv4 packets can be expressed as follows::
2451
2452   testpmd> flow create 0 ingress pattern eth / ipv4 / udp / end
2453      actions drop / end
2454
2455Several actions have configurable properties which must be specified when
2456there is no valid default value. For example, ``queue`` requires a target
2457queue index.
2458
2459This rule redirects incoming UDPv4 traffic to queue index 6::
2460
2461   testpmd> flow create 0 ingress pattern eth / ipv4 / udp / end
2462      actions queue index 6 / end
2463
2464While this one could be rejected by PMDs (unspecified queue index)::
2465
2466   testpmd> flow create 0 ingress pattern eth / ipv4 / udp / end
2467      actions queue / end
2468
2469As defined by *rte_flow*, the list is not ordered, all actions of a given
2470rule are performed simultaneously. These are equivalent::
2471
2472   queue index 6 / void / mark id 42 / end
2473
2474::
2475
2476   void / mark id 42 / queue index 6 / end
2477
2478All actions in a list should have different types, otherwise only the last
2479action of a given type is taken into account::
2480
2481   queue index 4 / queue index 5 / queue index 6 / end # will use queue 6
2482
2483::
2484
2485   drop / drop / drop / end # drop is performed only once
2486
2487::
2488
2489   mark id 42 / queue index 3 / mark id 24 / end # mark will be 24
2490
2491Considering they are performed simultaneously, opposite and overlapping
2492actions can sometimes be combined when the end result is unambiguous::
2493
2494   drop / queue index 6 / end # drop has no effect
2495
2496::
2497
2498   drop / dup index 6 / end # same as above
2499
2500::
2501
2502   queue index 6 / rss queues 6 7 8 / end # queue has no effect
2503
2504::
2505
2506   drop / passthru / end # drop has no effect
2507
2508Note that PMDs may still refuse such combinations.
2509
2510Actions
2511^^^^^^^
2512
2513This section lists supported actions and their attributes, if any.
2514
2515- ``end``: end list of actions.
2516
2517- ``void``: no-op action.
2518
2519- ``passthru``: let subsequent rule process matched packets.
2520
2521- ``mark``: attach 32 bit value to packets.
2522
2523  - ``id {unsigned}``: 32 bit value to return with packets.
2524
2525- ``flag``: flag packets.
2526
2527- ``queue``: assign packets to a given queue index.
2528
2529  - ``index {unsigned}``: queue index to use.
2530
2531- ``drop``: drop packets (note: passthru has priority).
2532
2533- ``count``: enable counters for this rule.
2534
2535- ``dup``: duplicate packets to a given queue index.
2536
2537  - ``index {unsigned}``: queue index to duplicate packets to.
2538
2539- ``rss``: spread packets among several queues.
2540
2541  - ``queues [{unsigned} [...]] end``: queue indices to use.
2542
2543- ``pf``: redirect packets to physical device function.
2544
2545- ``vf``: redirect packets to virtual device function.
2546
2547  - ``original {boolean}``: use original VF ID if possible.
2548  - ``id {unsigned}``: VF ID to redirect packets to.
2549
2550Destroying flow rules
2551~~~~~~~~~~~~~~~~~~~~~
2552
2553``flow destroy`` destroys one or more rules from their rule ID (as returned
2554by ``flow create``), this command calls ``rte_flow_destroy()`` as many
2555times as necessary::
2556
2557   flow destroy {port_id} rule {rule_id} [...]
2558
2559If successful, it will show::
2560
2561   Flow rule #[...] destroyed
2562
2563It does not report anything for rule IDs that do not exist. The usual error
2564message is shown when a rule cannot be destroyed::
2565
2566   Caught error type [...] ([...]): [...]
2567
2568``flow flush`` destroys all rules on a device and does not take extra
2569arguments. It is bound to ``rte_flow_flush()``::
2570
2571   flow flush {port_id}
2572
2573Any errors are reported as above.
2574
2575Creating several rules and destroying them::
2576
2577   testpmd> flow create 0 ingress pattern eth / ipv6 / end
2578      actions queue index 2 / end
2579   Flow rule #0 created
2580   testpmd> flow create 0 ingress pattern eth / ipv4 / end
2581      actions queue index 3 / end
2582   Flow rule #1 created
2583   testpmd> flow destroy 0 rule 0 rule 1
2584   Flow rule #1 destroyed
2585   Flow rule #0 destroyed
2586   testpmd>
2587
2588The same result can be achieved using ``flow flush``::
2589
2590   testpmd> flow create 0 ingress pattern eth / ipv6 / end
2591      actions queue index 2 / end
2592   Flow rule #0 created
2593   testpmd> flow create 0 ingress pattern eth / ipv4 / end
2594      actions queue index 3 / end
2595   Flow rule #1 created
2596   testpmd> flow flush 0
2597   testpmd>
2598
2599Non-existent rule IDs are ignored::
2600
2601   testpmd> flow create 0 ingress pattern eth / ipv6 / end
2602      actions queue index 2 / end
2603   Flow rule #0 created
2604   testpmd> flow create 0 ingress pattern eth / ipv4 / end
2605      actions queue index 3 / end
2606   Flow rule #1 created
2607   testpmd> flow destroy 0 rule 42 rule 10 rule 2
2608   testpmd>
2609   testpmd> flow destroy 0 rule 0
2610   Flow rule #0 destroyed
2611   testpmd>
2612
2613Querying flow rules
2614~~~~~~~~~~~~~~~~~~~
2615
2616``flow query`` queries a specific action of a flow rule having that
2617ability. Such actions collect information that can be reported using this
2618command. It is bound to ``rte_flow_query()``::
2619
2620   flow query {port_id} {rule_id} {action}
2621
2622If successful, it will display either the retrieved data for known actions
2623or the following message::
2624
2625   Cannot display result for action type [...] ([...])
2626
2627Otherwise, it will complain either that the rule does not exist or that some
2628error occurred::
2629
2630   Flow rule #[...] not found
2631
2632::
2633
2634   Caught error type [...] ([...]): [...]
2635
2636Currently only the ``count`` action is supported. This action reports the
2637number of packets that hit the flow rule and the total number of bytes. Its
2638output has the following format::
2639
2640   count:
2641    hits_set: [...] # whether "hits" contains a valid value
2642    bytes_set: [...] # whether "bytes" contains a valid value
2643    hits: [...] # number of packets
2644    bytes: [...] # number of bytes
2645
2646Querying counters for TCPv6 packets redirected to queue 6::
2647
2648   testpmd> flow create 0 ingress pattern eth / ipv6 / tcp / end
2649      actions queue index 6 / count / end
2650   Flow rule #4 created
2651   testpmd> flow query 0 4 count
2652   count:
2653    hits_set: 1
2654    bytes_set: 0
2655    hits: 386446
2656    bytes: 0
2657   testpmd>
2658
2659Listing flow rules
2660~~~~~~~~~~~~~~~~~~
2661
2662``flow list`` lists existing flow rules sorted by priority and optionally
2663filtered by group identifiers::
2664
2665   flow list {port_id} [group {group_id}] [...]
2666
2667This command only fails with the following message if the device does not
2668exist::
2669
2670   Invalid port [...]
2671
2672Output consists of a header line followed by a short description of each
2673flow rule, one per line. There is no output at all when no flow rules are
2674configured on the device::
2675
2676   ID      Group   Prio    Attr    Rule
2677   [...]   [...]   [...]   [...]   [...]
2678
2679``Attr`` column flags:
2680
2681- ``i`` for ``ingress``.
2682- ``e`` for ``egress``.
2683
2684Creating several flow rules and listing them::
2685
2686   testpmd> flow create 0 ingress pattern eth / ipv4 / end
2687      actions queue index 6 / end
2688   Flow rule #0 created
2689   testpmd> flow create 0 ingress pattern eth / ipv6 / end
2690      actions queue index 2 / end
2691   Flow rule #1 created
2692   testpmd> flow create 0 priority 5 ingress pattern eth / ipv4 / udp / end
2693      actions rss queues 6 7 8 end / end
2694   Flow rule #2 created
2695   testpmd> flow list 0
2696   ID      Group   Prio    Attr    Rule
2697   0       0       0       i-      ETH IPV4 => QUEUE
2698   1       0       0       i-      ETH IPV6 => QUEUE
2699   2       0       5       i-      ETH IPV4 UDP => RSS
2700   testpmd>
2701
2702Rules are sorted by priority (i.e. group ID first, then priority level)::
2703
2704   testpmd> flow list 1
2705   ID      Group   Prio    Attr    Rule
2706   0       0       0       i-      ETH => COUNT
2707   6       0       500     i-      ETH IPV6 TCP => DROP COUNT
2708   5       0       1000    i-      ETH IPV6 ICMP => QUEUE
2709   1       24      0       i-      ETH IPV4 UDP => QUEUE
2710   4       24      10      i-      ETH IPV4 TCP => DROP
2711   3       24      20      i-      ETH IPV4 => DROP
2712   2       24      42      i-      ETH IPV4 UDP => QUEUE
2713   7       63      0       i-      ETH IPV6 UDP VXLAN => MARK QUEUE
2714   testpmd>
2715
2716Output can be limited to specific groups::
2717
2718   testpmd> flow list 1 group 0 group 63
2719   ID      Group   Prio    Attr    Rule
2720   0       0       0       i-      ETH => COUNT
2721   6       0       500     i-      ETH IPV6 TCP => DROP COUNT
2722   5       0       1000    i-      ETH IPV6 ICMP => QUEUE
2723   7       63      0       i-      ETH IPV6 UDP VXLAN => MARK QUEUE
2724   testpmd>
2725