xref: /dpdk/doc/guides/testpmd_app_ug/testpmd_funcs.rst (revision 1960be7d32f85f07099e9642e3aba2a3c837b598)
1..  SPDX-License-Identifier: BSD-3-Clause
2    Copyright(c) 2010-2016 Intel Corporation.
3
4.. _testpmd_runtime:
5
6Testpmd Runtime Functions
7=========================
8
9Where the testpmd application is started in interactive mode, (``-i|--interactive``),
10it displays a prompt that can be used to start and stop forwarding,
11configure the application, display statistics (including the extended NIC
12statistics aka xstats) , set the Flow Director and other tasks::
13
14   testpmd>
15
16The testpmd prompt has some, limited, readline support.
17Common bash command-line functions such as ``Ctrl+a`` and ``Ctrl+e`` to go to the start and end of the prompt line are supported
18as well as access to the command history via the up-arrow.
19
20There is also support for tab completion.
21If you type a partial command and hit ``<TAB>`` you get a list of the available completions:
22
23.. code-block:: console
24
25   testpmd> show port <TAB>
26
27       info [Mul-choice STRING]: show|clear port info|stats|xstats|fdir|stat_qmap|dcb_tc|cap X
28       info [Mul-choice STRING]: show|clear port info|stats|xstats|fdir|stat_qmap|dcb_tc|cap all
29       stats [Mul-choice STRING]: show|clear port info|stats|xstats|fdir|stat_qmap|dcb_tc|cap X
30       stats [Mul-choice STRING]: show|clear port info|stats|xstats|fdir|stat_qmap|dcb_tc|cap all
31       ...
32
33
34.. note::
35
36   Some examples in this document are too long to fit on one line are are shown wrapped at `"\\"` for display purposes::
37
38      testpmd> set flow_ctrl rx (on|off) tx (on|off) (high_water) (low_water) \
39               (pause_time) (send_xon) (port_id)
40
41In the real ``testpmd>`` prompt these commands should be on a single line.
42
43Help Functions
44--------------
45
46The testpmd has on-line help for the functions that are available at runtime.
47These are divided into sections and can be accessed using help, help section or help all:
48
49.. code-block:: console
50
51   testpmd> help
52
53       help control    : Start and stop forwarding.
54       help display    : Displaying port, stats and config information.
55       help config     : Configuration information.
56       help ports      : Configuring ports.
57       help registers  : Reading and setting port registers.
58       help filters    : Filters configuration help.
59       help all        : All of the above sections.
60
61
62Command File Functions
63----------------------
64
65To facilitate loading large number of commands or to avoid cutting and pasting where not
66practical or possible testpmd supports alternative methods for executing commands.
67
68* If started with the ``--cmdline-file=FILENAME`` command line argument testpmd
69  will execute all CLI commands contained within the file immediately before
70  starting packet forwarding or entering interactive mode.
71
72.. code-block:: console
73
74   ./testpmd -n4 -r2 ... -- -i --cmdline-file=/home/ubuntu/flow-create-commands.txt
75   Interactive-mode selected
76   CLI commands to be read from /home/ubuntu/flow-create-commands.txt
77   Configuring Port 0 (socket 0)
78   Port 0: 7C:FE:90:CB:74:CE
79   Configuring Port 1 (socket 0)
80   Port 1: 7C:FE:90:CB:74:CA
81   Checking link statuses...
82   Port 0 Link Up - speed 10000 Mbps - full-duplex
83   Port 1 Link Up - speed 10000 Mbps - full-duplex
84   Done
85   Flow rule #0 created
86   Flow rule #1 created
87   ...
88   ...
89   Flow rule #498 created
90   Flow rule #499 created
91   Read all CLI commands from /home/ubuntu/flow-create-commands.txt
92   testpmd>
93
94
95* At run-time additional commands can be loaded in bulk by invoking the ``load FILENAME``
96  command.
97
98.. code-block:: console
99
100   testpmd> load /home/ubuntu/flow-create-commands.txt
101   Flow rule #0 created
102   Flow rule #1 created
103   ...
104   ...
105   Flow rule #498 created
106   Flow rule #499 created
107   Read all CLI commands from /home/ubuntu/flow-create-commands.txt
108   testpmd>
109
110
111In all cases output from any included command will be displayed as standard output.
112Execution will continue until the end of the file is reached regardless of
113whether any errors occur.  The end user must examine the output to determine if
114any failures occurred.
115
116
117Control Functions
118-----------------
119
120start
121~~~~~
122
123Start packet forwarding with current configuration::
124
125   testpmd> start
126
127start tx_first
128~~~~~~~~~~~~~~
129
130Start packet forwarding with current configuration after sending specified number of bursts of packets::
131
132   testpmd> start tx_first (""|burst_num)
133
134The default burst number is 1 when ``burst_num`` not presented.
135
136stop
137~~~~
138
139Stop packet forwarding, and display accumulated statistics::
140
141   testpmd> stop
142
143quit
144~~~~
145
146Quit to prompt::
147
148   testpmd> quit
149
150
151Display Functions
152-----------------
153
154The functions in the following sections are used to display information about the
155testpmd configuration or the NIC status.
156
157show port
158~~~~~~~~~
159
160Display information for a given port or all ports::
161
162   testpmd> show port (info|stats|xstats|fdir|stat_qmap|dcb_tc|cap) (port_id|all)
163
164The available information categories are:
165
166* ``info``: General port information such as MAC address.
167
168* ``stats``: RX/TX statistics.
169
170* ``xstats``: RX/TX extended NIC statistics.
171
172* ``fdir``: Flow Director information and statistics.
173
174* ``stat_qmap``: Queue statistics mapping.
175
176* ``dcb_tc``: DCB information such as TC mapping.
177
178* ``cap``: Supported offload capabilities.
179
180For example:
181
182.. code-block:: console
183
184   testpmd> show port info 0
185
186   ********************* Infos for port 0 *********************
187
188   MAC address: XX:XX:XX:XX:XX:XX
189   Connect to socket: 0
190   memory allocation on the socket: 0
191   Link status: up
192   Link speed: 40000 Mbps
193   Link duplex: full-duplex
194   Promiscuous mode: enabled
195   Allmulticast mode: disabled
196   Maximum number of MAC addresses: 64
197   Maximum number of MAC addresses of hash filtering: 0
198   VLAN offload:
199       strip on
200       filter on
201       qinq(extend) off
202   Redirection table size: 512
203   Supported flow types:
204     ipv4-frag
205     ipv4-tcp
206     ipv4-udp
207     ipv4-sctp
208     ipv4-other
209     ipv6-frag
210     ipv6-tcp
211     ipv6-udp
212     ipv6-sctp
213     ipv6-other
214     l2_payload
215     port
216     vxlan
217     geneve
218     nvgre
219
220show port rss reta
221~~~~~~~~~~~~~~~~~~
222
223Display the rss redirection table entry indicated by masks on port X::
224
225   testpmd> show port (port_id) rss reta (size) (mask0, mask1...)
226
227size is used to indicate the hardware supported reta size
228
229show port rss-hash
230~~~~~~~~~~~~~~~~~~
231
232Display the RSS hash functions and RSS hash key of a port::
233
234   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]
235
236clear port
237~~~~~~~~~~
238
239Clear the port statistics for a given port or for all ports::
240
241   testpmd> clear port (info|stats|xstats|fdir|stat_qmap) (port_id|all)
242
243For example::
244
245   testpmd> clear port stats all
246
247show (rxq|txq)
248~~~~~~~~~~~~~~
249
250Display information for a given port's RX/TX queue::
251
252   testpmd> show (rxq|txq) info (port_id) (queue_id)
253
254show config
255~~~~~~~~~~~
256
257Displays the configuration of the application.
258The configuration comes from the command-line, the runtime or the application defaults::
259
260   testpmd> show config (rxtx|cores|fwd|txpkts)
261
262The available information categories are:
263
264* ``rxtx``: RX/TX configuration items.
265
266* ``cores``: List of forwarding cores.
267
268* ``fwd``: Packet forwarding configuration.
269
270* ``txpkts``: Packets to TX configuration.
271
272For example:
273
274.. code-block:: console
275
276   testpmd> show config rxtx
277
278   io packet forwarding - CRC stripping disabled - packets/burst=16
279   nb forwarding cores=2 - nb forwarding ports=1
280   RX queues=1 - RX desc=128 - RX free threshold=0
281   RX threshold registers: pthresh=8 hthresh=8 wthresh=4
282   TX queues=1 - TX desc=512 - TX free threshold=0
283   TX threshold registers: pthresh=36 hthresh=0 wthresh=0
284   TX RS bit threshold=0 - TXQ flags=0x0
285
286set fwd
287~~~~~~~
288
289Set the packet forwarding mode::
290
291   testpmd> set fwd (io|mac|macswap|flowgen| \
292                     rxonly|txonly|csum|icmpecho) (""|retry)
293
294``retry`` can be specified for forwarding engines except ``rx_only``.
295
296The available information categories are:
297
298* ``io``: Forwards packets "as-is" in I/O mode.
299  This is the fastest possible forwarding operation as it does not access packets data.
300  This is the default mode.
301
302* ``mac``: Changes the source and the destination Ethernet addresses of packets before forwarding them.
303  Default application behaviour is to set source Ethernet address to that of the transmitting interface, and destination
304  address to a dummy value (set during init). The user may specify a target destination Ethernet address via the 'eth-peer' or
305  'eth-peer-configfile' command-line options. It is not currently possible to specify a specific source Ethernet address.
306
307* ``macswap``: MAC swap forwarding mode.
308  Swaps the source and the destination Ethernet addresses of packets before forwarding them.
309
310* ``flowgen``: Multi-flow generation mode.
311  Originates a number of flows (with varying destination IP addresses), and terminate receive traffic.
312
313* ``rxonly``: Receives packets but doesn't transmit them.
314
315* ``txonly``: Generates and transmits packets without receiving any.
316
317* ``csum``: Changes the checksum field with hardware or software methods depending on the offload flags on the packet.
318
319* ``icmpecho``: Receives a burst of packets, lookup for IMCP echo requests and, if any, send back ICMP echo replies.
320
321* ``ieee1588``: Demonstrate L2 IEEE1588 V2 PTP timestamping for RX and TX. Requires ``CONFIG_RTE_LIBRTE_IEEE1588=y``.
322
323* ``tm``: Traffic Management forwarding mode
324  Demonstrates the use of ethdev traffic management APIs and softnic PMD for
325  QoS traffic management. In this mode, 5-level hierarchical QoS scheduler is
326  available as an default option that can be enabled through CLI. The user can
327  also modify the default hierarchy or specify the new hierarchy through CLI for
328  implementing QoS scheduler.  Requires ``CONFIG_RTE_LIBRTE_PMD_SOFTNIC=y`` ``CONFIG_RTE_LIBRTE_SCHED=y``.
329
330Example::
331
332   testpmd> set fwd rxonly
333
334   Set rxonly packet forwarding mode
335
336
337read rxd
338~~~~~~~~
339
340Display an RX descriptor for a port RX queue::
341
342   testpmd> read rxd (port_id) (queue_id) (rxd_id)
343
344For example::
345
346   testpmd> read rxd 0 0 4
347        0x0000000B - 0x001D0180 / 0x0000000B - 0x001D0180
348
349read txd
350~~~~~~~~
351
352Display a TX descriptor for a port TX queue::
353
354   testpmd> read txd (port_id) (queue_id) (txd_id)
355
356For example::
357
358   testpmd> read txd 0 0 4
359        0x00000001 - 0x24C3C440 / 0x000F0000 - 0x2330003C
360
361ddp get list
362~~~~~~~~~~~~
363
364Get loaded dynamic device personalization (DDP) package info list::
365
366   testpmd> ddp get list (port_id)
367
368ddp get info
369~~~~~~~~~~~~
370
371Display information about dynamic device personalization (DDP) profile::
372
373   testpmd> ddp get info (profile_path)
374
375show vf stats
376~~~~~~~~~~~~~
377
378Display VF statistics::
379
380   testpmd> show vf stats (port_id) (vf_id)
381
382clear vf stats
383~~~~~~~~~~~~~~
384
385Reset VF statistics::
386
387   testpmd> clear vf stats (port_id) (vf_id)
388
389show port pctype mapping
390~~~~~~~~~~~~~~~~~~~~~~~~
391
392List all items from the pctype mapping table::
393
394   testpmd> show port (port_id) pctype mapping
395
396show rx offloading capabilities
397~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
398
399List all per queue and per port Rx offloading capabilities of a port::
400
401   testpmd> show port (port_id) rx_offload capabilities
402
403show rx offloading configuration
404~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
405
406List port level and all queue level Rx offloading configuration::
407
408   testpmd> show port (port_id) rx_offload configuration
409
410show tx offloading capabilities
411~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
412
413List all per queue and per port Tx offloading capabilities of a port::
414
415   testpmd> show port (port_id) tx_offload capabilities
416
417show tx offloading configuration
418~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
419
420List port level and all queue level Tx offloading configuration::
421
422   testpmd> show port (port_id) tx_offload configuration
423
424
425Configuration Functions
426-----------------------
427
428The testpmd application can be configured from the runtime as well as from the command-line.
429
430This section details the available configuration functions that are available.
431
432.. note::
433
434   Configuration changes only become active when forwarding is started/restarted.
435
436set default
437~~~~~~~~~~~
438
439Reset forwarding to the default configuration::
440
441   testpmd> set default
442
443set verbose
444~~~~~~~~~~~
445
446Set the debug verbosity level::
447
448   testpmd> set verbose (level)
449
450Currently the only available levels are 0 (silent except for error) and 1 (fully verbose).
451
452set log
453~~~~~~~
454
455Set the log level for a log type::
456
457	testpmd> set log global|(type) (level)
458
459Where:
460
461* ``type`` is the log name.
462
463* ``level`` is the log level.
464
465For example, to change the global log level::
466	testpmd> set log global (level)
467
468Regexes can also be used for type. To change log level of user1, user2 and user3::
469	testpmd> set log user[1-3] (level)
470
471set nbport
472~~~~~~~~~~
473
474Set the number of ports used by the application:
475
476set nbport (num)
477
478This is equivalent to the ``--nb-ports`` command-line option.
479
480set nbcore
481~~~~~~~~~~
482
483Set the number of cores used by the application::
484
485   testpmd> set nbcore (num)
486
487This is equivalent to the ``--nb-cores`` command-line option.
488
489.. note::
490
491   The number of cores used must not be greater than number of ports used multiplied by the number of queues per port.
492
493set coremask
494~~~~~~~~~~~~
495
496Set the forwarding cores hexadecimal mask::
497
498   testpmd> set coremask (mask)
499
500This is equivalent to the ``--coremask`` command-line option.
501
502.. note::
503
504   The master lcore is reserved for command line parsing only and cannot be masked on for packet forwarding.
505
506set portmask
507~~~~~~~~~~~~
508
509Set the forwarding ports hexadecimal mask::
510
511   testpmd> set portmask (mask)
512
513This is equivalent to the ``--portmask`` command-line option.
514
515set burst
516~~~~~~~~~
517
518Set number of packets per burst::
519
520   testpmd> set burst (num)
521
522This is equivalent to the ``--burst command-line`` option.
523
524When retry is enabled, the transmit delay time and number of retries can also be set::
525
526   testpmd> set burst tx delay (microseconds) retry (num)
527
528set txpkts
529~~~~~~~~~~
530
531Set the length of each segment of the TX-ONLY packets or length of packet for FLOWGEN mode::
532
533   testpmd> set txpkts (x[,y]*)
534
535Where x[,y]* represents a CSV list of values, without white space.
536
537set txsplit
538~~~~~~~~~~~
539
540Set the split policy for the TX packets, applicable for TX-ONLY and CSUM forwarding modes::
541
542   testpmd> set txsplit (off|on|rand)
543
544Where:
545
546* ``off`` disable packet copy & split for CSUM mode.
547
548* ``on`` split outgoing packet into multiple segments. Size of each segment
549  and number of segments per packet is determined by ``set txpkts`` command
550  (see above).
551
552* ``rand`` same as 'on', but number of segments per each packet is a random value between 1 and total number of segments.
553
554set corelist
555~~~~~~~~~~~~
556
557Set the list of forwarding cores::
558
559   testpmd> set corelist (x[,y]*)
560
561For example, to change the forwarding cores:
562
563.. code-block:: console
564
565   testpmd> set corelist 3,1
566   testpmd> show config fwd
567
568   io packet forwarding - ports=2 - cores=2 - streams=2 - NUMA support disabled
569   Logical Core 3 (socket 0) forwards packets on 1 streams:
570   RX P=0/Q=0 (socket 0) -> TX P=1/Q=0 (socket 0) peer=02:00:00:00:00:01
571   Logical Core 1 (socket 0) forwards packets on 1 streams:
572   RX P=1/Q=0 (socket 0) -> TX P=0/Q=0 (socket 0) peer=02:00:00:00:00:00
573
574.. note::
575
576   The cores are used in the same order as specified on the command line.
577
578set portlist
579~~~~~~~~~~~~
580
581Set the list of forwarding ports::
582
583   testpmd> set portlist (x[,y]*)
584
585For example, to change the port forwarding:
586
587.. code-block:: console
588
589   testpmd> set portlist 0,2,1,3
590   testpmd> show config fwd
591
592   io packet forwarding - ports=4 - cores=1 - streams=4
593   Logical Core 3 (socket 0) forwards packets on 4 streams:
594   RX P=0/Q=0 (socket 0) -> TX P=2/Q=0 (socket 0) peer=02:00:00:00:00:01
595   RX P=2/Q=0 (socket 0) -> TX P=0/Q=0 (socket 0) peer=02:00:00:00:00:00
596   RX P=1/Q=0 (socket 0) -> TX P=3/Q=0 (socket 0) peer=02:00:00:00:00:03
597   RX P=3/Q=0 (socket 0) -> TX P=1/Q=0 (socket 0) peer=02:00:00:00:00:02
598
599set tx loopback
600~~~~~~~~~~~~~~~
601
602Enable/disable tx loopback::
603
604   testpmd> set tx loopback (port_id) (on|off)
605
606set drop enable
607~~~~~~~~~~~~~~~
608
609set drop enable bit for all queues::
610
611   testpmd> set all queues drop (port_id) (on|off)
612
613set split drop enable (for VF)
614~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
615
616set split drop enable bit for VF from PF::
617
618   testpmd> set vf split drop (port_id) (vf_id) (on|off)
619
620set mac antispoof (for VF)
621~~~~~~~~~~~~~~~~~~~~~~~~~~
622
623Set mac antispoof for a VF from the PF::
624
625   testpmd> set vf mac antispoof  (port_id) (vf_id) (on|off)
626
627set macsec offload
628~~~~~~~~~~~~~~~~~~
629
630Enable/disable MACsec offload::
631
632   testpmd> set macsec offload (port_id) on encrypt (on|off) replay-protect (on|off)
633   testpmd> set macsec offload (port_id) off
634
635set macsec sc
636~~~~~~~~~~~~~
637
638Configure MACsec secure connection (SC)::
639
640   testpmd> set macsec sc (tx|rx) (port_id) (mac) (pi)
641
642.. note::
643
644   The pi argument is ignored for tx.
645   Check the NIC Datasheet for hardware limits.
646
647set macsec sa
648~~~~~~~~~~~~~
649
650Configure MACsec secure association (SA)::
651
652   testpmd> set macsec sa (tx|rx) (port_id) (idx) (an) (pn) (key)
653
654.. note::
655
656   The IDX value must be 0 or 1.
657   Check the NIC Datasheet for hardware limits.
658
659set broadcast mode (for VF)
660~~~~~~~~~~~~~~~~~~~~~~~~~~~
661
662Set broadcast mode for a VF from the PF::
663
664   testpmd> set vf broadcast (port_id) (vf_id) (on|off)
665
666vlan set strip
667~~~~~~~~~~~~~~
668
669Set the VLAN strip on a port::
670
671   testpmd> vlan set strip (on|off) (port_id)
672
673vlan set stripq
674~~~~~~~~~~~~~~~
675
676Set the VLAN strip for a queue on a port::
677
678   testpmd> vlan set stripq (on|off) (port_id,queue_id)
679
680vlan set stripq (for VF)
681~~~~~~~~~~~~~~~~~~~~~~~~
682
683Set VLAN strip for all queues in a pool for a VF from the PF::
684
685   testpmd> set vf vlan stripq (port_id) (vf_id) (on|off)
686
687vlan set insert (for VF)
688~~~~~~~~~~~~~~~~~~~~~~~~
689
690Set VLAN insert for a VF from the PF::
691
692   testpmd> set vf vlan insert (port_id) (vf_id) (vlan_id)
693
694vlan set tag (for VF)
695~~~~~~~~~~~~~~~~~~~~~
696
697Set VLAN tag for a VF from the PF::
698
699   testpmd> set vf vlan tag (port_id) (vf_id) (on|off)
700
701vlan set antispoof (for VF)
702~~~~~~~~~~~~~~~~~~~~~~~~~~~
703
704Set VLAN antispoof for a VF from the PF::
705
706   testpmd> set vf vlan antispoof (port_id) (vf_id) (on|off)
707
708vlan set filter
709~~~~~~~~~~~~~~~
710
711Set the VLAN filter on a port::
712
713   testpmd> vlan set filter (on|off) (port_id)
714
715vlan set qinq
716~~~~~~~~~~~~~
717
718Set the VLAN QinQ (extended queue in queue) on for a port::
719
720   testpmd> vlan set qinq (on|off) (port_id)
721
722vlan set tpid
723~~~~~~~~~~~~~
724
725Set the inner or outer VLAN TPID for packet filtering on a port::
726
727   testpmd> vlan set (inner|outer) tpid (value) (port_id)
728
729.. note::
730
731   TPID value must be a 16-bit number (value <= 65536).
732
733rx_vlan add
734~~~~~~~~~~~
735
736Add a VLAN ID, or all identifiers, to the set of VLAN identifiers filtered by port ID::
737
738   testpmd> rx_vlan add (vlan_id|all) (port_id)
739
740.. note::
741
742   VLAN filter must be set on that port. VLAN ID < 4096.
743   Depending on the NIC used, number of vlan_ids may be limited to the maximum entries
744   in VFTA table. This is important if enabling all vlan_ids.
745
746rx_vlan rm
747~~~~~~~~~~
748
749Remove a VLAN ID, or all identifiers, from the set of VLAN identifiers filtered by port ID::
750
751   testpmd> rx_vlan rm (vlan_id|all) (port_id)
752
753rx_vlan add (for VF)
754~~~~~~~~~~~~~~~~~~~~
755
756Add a VLAN ID, to the set of VLAN identifiers filtered for VF(s) for port ID::
757
758   testpmd> rx_vlan add (vlan_id) port (port_id) vf (vf_mask)
759
760rx_vlan rm (for VF)
761~~~~~~~~~~~~~~~~~~~
762
763Remove a VLAN ID, from the set of VLAN identifiers filtered for VF(s) for port ID::
764
765   testpmd> rx_vlan rm (vlan_id) port (port_id) vf (vf_mask)
766
767tunnel_filter add
768~~~~~~~~~~~~~~~~~
769
770Add a tunnel filter on a port::
771
772   testpmd> tunnel_filter add (port_id) (outer_mac) (inner_mac) (ip_addr) \
773            (inner_vlan) (vxlan|nvgre|ipingre) (imac-ivlan|imac-ivlan-tenid|\
774            imac-tenid|imac|omac-imac-tenid|oip|iip) (tenant_id) (queue_id)
775
776The available information categories are:
777
778* ``vxlan``: Set tunnel type as VXLAN.
779
780* ``nvgre``: Set tunnel type as NVGRE.
781
782* ``ipingre``: Set tunnel type as IP-in-GRE.
783
784* ``imac-ivlan``: Set filter type as Inner MAC and VLAN.
785
786* ``imac-ivlan-tenid``: Set filter type as Inner MAC, VLAN and tenant ID.
787
788* ``imac-tenid``: Set filter type as Inner MAC and tenant ID.
789
790* ``imac``: Set filter type as Inner MAC.
791
792* ``omac-imac-tenid``: Set filter type as Outer MAC, Inner MAC and tenant ID.
793
794* ``oip``: Set filter type as Outer IP.
795
796* ``iip``: Set filter type as Inner IP.
797
798Example::
799
800   testpmd> tunnel_filter add 0 68:05:CA:28:09:82 00:00:00:00:00:00 \
801            192.168.2.2 0 ipingre oip 1 1
802
803   Set an IP-in-GRE tunnel on port 0, and the filter type is Outer IP.
804
805tunnel_filter remove
806~~~~~~~~~~~~~~~~~~~~
807
808Remove a tunnel filter on a port::
809
810   testpmd> tunnel_filter rm (port_id) (outer_mac) (inner_mac) (ip_addr) \
811            (inner_vlan) (vxlan|nvgre|ipingre) (imac-ivlan|imac-ivlan-tenid|\
812            imac-tenid|imac|omac-imac-tenid|oip|iip) (tenant_id) (queue_id)
813
814rx_vxlan_port add
815~~~~~~~~~~~~~~~~~
816
817Add an UDP port for VXLAN packet filter on a port::
818
819   testpmd> rx_vxlan_port add (udp_port) (port_id)
820
821rx_vxlan_port remove
822~~~~~~~~~~~~~~~~~~~~
823
824Remove an UDP port for VXLAN packet filter on a port::
825
826   testpmd> rx_vxlan_port rm (udp_port) (port_id)
827
828tx_vlan set
829~~~~~~~~~~~
830
831Set hardware insertion of VLAN IDs in packets sent on a port::
832
833   testpmd> tx_vlan set (port_id) vlan_id[, vlan_id_outer]
834
835For example, set a single VLAN ID (5) insertion on port 0::
836
837   tx_vlan set 0 5
838
839Or, set double VLAN ID (inner: 2, outer: 3) insertion on port 1::
840
841   tx_vlan set 1 2 3
842
843
844tx_vlan set pvid
845~~~~~~~~~~~~~~~~
846
847Set port based hardware insertion of VLAN ID in packets sent on a port::
848
849   testpmd> tx_vlan set pvid (port_id) (vlan_id) (on|off)
850
851tx_vlan reset
852~~~~~~~~~~~~~
853
854Disable hardware insertion of a VLAN header in packets sent on a port::
855
856   testpmd> tx_vlan reset (port_id)
857
858csum set
859~~~~~~~~
860
861Select hardware or software calculation of the checksum when
862transmitting a packet using the ``csum`` forwarding engine::
863
864   testpmd> csum set (ip|udp|tcp|sctp|outer-ip) (hw|sw) (port_id)
865
866Where:
867
868* ``ip|udp|tcp|sctp`` always relate to  the inner layer.
869
870* ``outer-ip`` relates to the outer IP layer (only for IPv4) in the case where the packet is recognized
871  as a tunnel packet by the forwarding engine (vxlan, gre and ipip are
872  supported). See also the ``csum parse-tunnel`` command.
873
874.. note::
875
876   Check the NIC Datasheet for hardware limits.
877
878RSS queue region
879~~~~~~~~~~~~~~~~
880
881Set RSS queue region span on a port::
882
883   testpmd> set port (port_id) queue-region region_id (value) \
884		queue_start_index (value) queue_num (value)
885
886Set flowtype mapping on a RSS queue region on a port::
887
888   testpmd> set port (port_id) queue-region region_id (value) flowtype (value)
889
890where:
891
892* For the flowtype(pctype) of packet,the specific index for each type has
893  been defined in file i40e_type.h as enum i40e_filter_pctype.
894
895Set user priority mapping on a RSS queue region on a port::
896
897   testpmd> set port (port_id) queue-region UP (value) region_id (value)
898
899Flush all queue region related configuration on a port::
900
901   testpmd> set port (port_id) queue-region flush (on|off)
902
903where:
904
905* "on"is just an enable function which server for other configuration,
906  it is for all configuration about queue region from up layer,
907  at first will only keep in DPDK softwarestored in driver,
908  only after "flush on", it commit all configuration to HW.
909  "off" is just clean all configuration about queue region just now,
910  and restore all to DPDK i40e driver default config when start up.
911
912Show all queue region related configuration info on a port::
913
914   testpmd> show port (port_id) queue-region
915
916.. note::
917
918  Queue region only support on PF by now, so these command is
919  only for configuration of queue region on PF port.
920
921csum parse-tunnel
922~~~~~~~~~~~~~~~~~
923
924Define how tunneled packets should be handled by the csum forward
925engine::
926
927   testpmd> csum parse-tunnel (on|off) (tx_port_id)
928
929If enabled, the csum forward engine will try to recognize supported
930tunnel headers (vxlan, gre, ipip).
931
932If disabled, treat tunnel packets as non-tunneled packets (a inner
933header is handled as a packet payload).
934
935.. note::
936
937   The port argument is the TX port like in the ``csum set`` command.
938
939Example:
940
941Consider a packet in packet like the following::
942
943   eth_out/ipv4_out/udp_out/vxlan/eth_in/ipv4_in/tcp_in
944
945* If parse-tunnel is enabled, the ``ip|udp|tcp|sctp`` parameters of ``csum set``
946  command relate to the inner headers (here ``ipv4_in`` and ``tcp_in``), and the
947  ``outer-ip parameter`` relates to the outer headers (here ``ipv4_out``).
948
949* If parse-tunnel is disabled, the ``ip|udp|tcp|sctp`` parameters of ``csum  set``
950   command relate to the outer headers, here ``ipv4_out`` and ``udp_out``.
951
952csum show
953~~~~~~~~~
954
955Display tx checksum offload configuration::
956
957   testpmd> csum show (port_id)
958
959tso set
960~~~~~~~
961
962Enable TCP Segmentation Offload (TSO) in the ``csum`` forwarding engine::
963
964   testpmd> tso set (segsize) (port_id)
965
966.. note::
967
968   Check the NIC datasheet for hardware limits.
969
970tso show
971~~~~~~~~
972
973Display the status of TCP Segmentation Offload::
974
975   testpmd> tso show (port_id)
976
977set port - gro
978~~~~~~~~~~~~~~
979
980Enable or disable GRO in ``csum`` forwarding engine::
981
982   testpmd> set port <port_id> gro on|off
983
984If enabled, the csum forwarding engine will perform GRO on the TCP/IPv4
985packets received from the given port.
986
987If disabled, packets received from the given port won't be performed
988GRO. By default, GRO is disabled for all ports.
989
990.. note::
991
992   When enable GRO for a port, TCP/IPv4 packets received from the port
993   will be performed GRO. After GRO, all merged packets have bad
994   checksums, since the GRO library doesn't re-calculate checksums for
995   the merged packets. Therefore, if users want the merged packets to
996   have correct checksums, please select HW IP checksum calculation and
997   HW TCP checksum calculation for the port which the merged packets are
998   transmitted to.
999
1000show port - gro
1001~~~~~~~~~~~~~~~
1002
1003Display GRO configuration for a given port::
1004
1005   testpmd> show port <port_id> gro
1006
1007set gro flush
1008~~~~~~~~~~~~~
1009
1010Set the cycle to flush the GROed packets from reassembly tables::
1011
1012   testpmd> set gro flush <cycles>
1013
1014When enable GRO, the csum forwarding engine performs GRO on received
1015packets, and the GROed packets are stored in reassembly tables. Users
1016can use this command to determine when the GROed packets are flushed
1017from the reassembly tables.
1018
1019The ``cycles`` is measured in GRO operation times. The csum forwarding
1020engine flushes the GROed packets from the tables every ``cycles`` GRO
1021operations.
1022
1023By default, the value of ``cycles`` is 1, which means flush GROed packets
1024from the reassembly tables as soon as one GRO operation finishes. The value
1025of ``cycles`` should be in the range of 1 to ``GRO_MAX_FLUSH_CYCLES``.
1026
1027Please note that the large value of ``cycles`` may cause the poor TCP/IP
1028stack performance. Because the GROed packets are delayed to arrive the
1029stack, thus causing more duplicated ACKs and TCP retransmissions.
1030
1031set port - gso
1032~~~~~~~~~~~~~~
1033
1034Toggle per-port GSO support in ``csum`` forwarding engine::
1035
1036   testpmd> set port <port_id> gso on|off
1037
1038If enabled, the csum forwarding engine will perform GSO on supported IPv4
1039packets, transmitted on the given port.
1040
1041If disabled, packets transmitted on the given port will not undergo GSO.
1042By default, GSO is disabled for all ports.
1043
1044.. note::
1045
1046   When GSO is enabled on a port, supported IPv4 packets transmitted on that
1047   port undergo GSO. Afterwards, the segmented packets are represented by
1048   multi-segment mbufs; however, the csum forwarding engine doesn't calculation
1049   of checksums for GSO'd segments in SW. As a result, if users want correct
1050   checksums in GSO segments, they should enable HW checksum calculation for
1051   GSO-enabled ports.
1052
1053   For example, HW checksum calculation for VxLAN GSO'd packets may be enabled
1054   by setting the following options in the csum forwarding engine:
1055
1056   testpmd> csum set outer_ip hw <port_id>
1057
1058   testpmd> csum set ip hw <port_id>
1059
1060   testpmd> csum set tcp hw <port_id>
1061
1062   UDP GSO is the same as IP fragmentation, which treats the UDP header
1063   as the payload and does not modify it during segmentation. That is,
1064   after UDP GSO, only the first output fragment has the original UDP
1065   header. Therefore, users need to enable HW IP checksum calculation
1066   and SW UDP checksum calculation for GSO-enabled ports, if they want
1067   correct checksums for UDP/IPv4 packets.
1068
1069set gso segsz
1070~~~~~~~~~~~~~
1071
1072Set the maximum GSO segment size (measured in bytes), which includes the
1073packet header and the packet payload for GSO-enabled ports (global)::
1074
1075   testpmd> set gso segsz <length>
1076
1077show port - gso
1078~~~~~~~~~~~~~~~
1079
1080Display the status of Generic Segmentation Offload for a given port::
1081
1082   testpmd> show port <port_id> gso
1083
1084mac_addr add
1085~~~~~~~~~~~~
1086
1087Add an alternative MAC address to a port::
1088
1089   testpmd> mac_addr add (port_id) (XX:XX:XX:XX:XX:XX)
1090
1091mac_addr remove
1092~~~~~~~~~~~~~~~
1093
1094Remove a MAC address from a port::
1095
1096   testpmd> mac_addr remove (port_id) (XX:XX:XX:XX:XX:XX)
1097
1098mac_addr add (for VF)
1099~~~~~~~~~~~~~~~~~~~~~
1100
1101Add an alternative MAC address for a VF to a port::
1102
1103   testpmd> mac_add add port (port_id) vf (vf_id) (XX:XX:XX:XX:XX:XX)
1104
1105mac_addr set
1106~~~~~~~~~~~~
1107
1108Set the default MAC address for a port::
1109
1110   testpmd> mac_addr set (port_id) (XX:XX:XX:XX:XX:XX)
1111
1112mac_addr set (for VF)
1113~~~~~~~~~~~~~~~~~~~~~
1114
1115Set the MAC address for a VF from the PF::
1116
1117   testpmd> set vf mac addr (port_id) (vf_id) (XX:XX:XX:XX:XX:XX)
1118
1119set eth-peer
1120~~~~~~~~~~~~
1121
1122Set the forwarding peer address for certain port::
1123
1124   testpmd> set eth-peer (port_id) (perr_addr)
1125
1126This is equivalent to the ``--eth-peer`` command-line option.
1127
1128set port-uta
1129~~~~~~~~~~~~
1130
1131Set the unicast hash filter(s) on/off for a port::
1132
1133   testpmd> set port (port_id) uta (XX:XX:XX:XX:XX:XX|all) (on|off)
1134
1135set promisc
1136~~~~~~~~~~~
1137
1138Set the promiscuous mode on for a port or for all ports.
1139In promiscuous mode packets are not dropped if they aren't for the specified MAC address::
1140
1141   testpmd> set promisc (port_id|all) (on|off)
1142
1143set allmulti
1144~~~~~~~~~~~~
1145
1146Set the allmulti mode for a port or for all ports::
1147
1148   testpmd> set allmulti (port_id|all) (on|off)
1149
1150Same as the ifconfig (8) option. Controls how multicast packets are handled.
1151
1152set promisc (for VF)
1153~~~~~~~~~~~~~~~~~~~~
1154
1155Set the unicast promiscuous mode for a VF from PF.
1156It's supported by Intel i40e NICs now.
1157In promiscuous mode packets are not dropped if they aren't for the specified MAC address::
1158
1159   testpmd> set vf promisc (port_id) (vf_id) (on|off)
1160
1161set allmulticast (for VF)
1162~~~~~~~~~~~~~~~~~~~~~~~~~
1163
1164Set the multicast promiscuous mode for a VF from PF.
1165It's supported by Intel i40e NICs now.
1166In promiscuous mode packets are not dropped if they aren't for the specified MAC address::
1167
1168   testpmd> set vf allmulti (port_id) (vf_id) (on|off)
1169
1170set tx max bandwidth (for VF)
1171~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
1172
1173Set TX max absolute bandwidth (Mbps) for a VF from PF::
1174
1175   testpmd> set vf tx max-bandwidth (port_id) (vf_id) (max_bandwidth)
1176
1177set tc tx min bandwidth (for VF)
1178~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
1179
1180Set all TCs' TX min relative bandwidth (%) for a VF from PF::
1181
1182   testpmd> set vf tc tx min-bandwidth (port_id) (vf_id) (bw1, bw2, ...)
1183
1184set tc tx max bandwidth (for VF)
1185~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
1186
1187Set a TC's TX max absolute bandwidth (Mbps) for a VF from PF::
1188
1189   testpmd> set vf tc tx max-bandwidth (port_id) (vf_id) (tc_no) (max_bandwidth)
1190
1191set tc strict link priority mode
1192~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
1193
1194Set some TCs' strict link priority mode on a physical port::
1195
1196   testpmd> set tx strict-link-priority (port_id) (tc_bitmap)
1197
1198set tc tx min bandwidth
1199~~~~~~~~~~~~~~~~~~~~~~~
1200
1201Set all TCs' TX min relative bandwidth (%) globally for all PF and VFs::
1202
1203   testpmd> set tc tx min-bandwidth (port_id) (bw1, bw2, ...)
1204
1205set flow_ctrl rx
1206~~~~~~~~~~~~~~~~
1207
1208Set the link flow control parameter on a port::
1209
1210   testpmd> set flow_ctrl rx (on|off) tx (on|off) (high_water) (low_water) \
1211            (pause_time) (send_xon) mac_ctrl_frame_fwd (on|off) \
1212	    autoneg (on|off) (port_id)
1213
1214Where:
1215
1216* ``high_water`` (integer): High threshold value to trigger XOFF.
1217
1218* ``low_water`` (integer): Low threshold value to trigger XON.
1219
1220* ``pause_time`` (integer): Pause quota in the Pause frame.
1221
1222* ``send_xon`` (0/1): Send XON frame.
1223
1224* ``mac_ctrl_frame_fwd``: Enable receiving MAC control frames.
1225
1226* ``autoneg``: Change the auto-negotiation parameter.
1227
1228set pfc_ctrl rx
1229~~~~~~~~~~~~~~~
1230
1231Set the priority flow control parameter on a port::
1232
1233   testpmd> set pfc_ctrl rx (on|off) tx (on|off) (high_water) (low_water) \
1234            (pause_time) (priority) (port_id)
1235
1236Where:
1237
1238* ``high_water`` (integer): High threshold value.
1239
1240* ``low_water`` (integer): Low threshold value.
1241
1242* ``pause_time`` (integer): Pause quota in the Pause frame.
1243
1244* ``priority`` (0-7): VLAN User Priority.
1245
1246set stat_qmap
1247~~~~~~~~~~~~~
1248
1249Set statistics mapping (qmapping 0..15) for RX/TX queue on port::
1250
1251   testpmd> set stat_qmap (tx|rx) (port_id) (queue_id) (qmapping)
1252
1253For example, to set rx queue 2 on port 0 to mapping 5::
1254
1255   testpmd>set stat_qmap rx 0 2 5
1256
1257set xstats-hide-zero
1258~~~~~~~~~~~~~~~~~~~~
1259
1260Set the option to hide zero values for xstats display::
1261
1262	testpmd> set xstats-hide-zero on|off
1263
1264.. note::
1265
1266	By default, the zero values are displayed for xstats.
1267
1268set port - rx/tx (for VF)
1269~~~~~~~~~~~~~~~~~~~~~~~~~
1270
1271Set VF receive/transmit from a port::
1272
1273   testpmd> set port (port_id) vf (vf_id) (rx|tx) (on|off)
1274
1275set port - mac address filter (for VF)
1276~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
1277
1278Add/Remove unicast or multicast MAC addr filter for a VF::
1279
1280   testpmd> set port (port_id) vf (vf_id) (mac_addr) \
1281            (exact-mac|exact-mac-vlan|hashmac|hashmac-vlan) (on|off)
1282
1283set port - rx mode(for VF)
1284~~~~~~~~~~~~~~~~~~~~~~~~~~
1285
1286Set the VF receive mode of a port::
1287
1288   testpmd> set port (port_id) vf (vf_id) \
1289            rxmode (AUPE|ROPE|BAM|MPE) (on|off)
1290
1291The available receive modes are:
1292
1293* ``AUPE``: Accepts untagged VLAN.
1294
1295* ``ROPE``: Accepts unicast hash.
1296
1297* ``BAM``: Accepts broadcast packets.
1298
1299* ``MPE``: Accepts all multicast packets.
1300
1301set port - tx_rate (for Queue)
1302~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
1303
1304Set TX rate limitation for a queue on a port::
1305
1306   testpmd> set port (port_id) queue (queue_id) rate (rate_value)
1307
1308set port - tx_rate (for VF)
1309~~~~~~~~~~~~~~~~~~~~~~~~~~~
1310
1311Set TX rate limitation for queues in VF on a port::
1312
1313   testpmd> set port (port_id) vf (vf_id) rate (rate_value) queue_mask (queue_mask)
1314
1315set port - mirror rule
1316~~~~~~~~~~~~~~~~~~~~~~
1317
1318Set pool or vlan type mirror rule for a port::
1319
1320   testpmd> set port (port_id) mirror-rule (rule_id) \
1321            (pool-mirror-up|pool-mirror-down|vlan-mirror) \
1322            (poolmask|vlanid[,vlanid]*) dst-pool (pool_id) (on|off)
1323
1324Set link mirror rule for a port::
1325
1326   testpmd> set port (port_id) mirror-rule (rule_id) \
1327           (uplink-mirror|downlink-mirror) dst-pool (pool_id) (on|off)
1328
1329For example to enable mirror traffic with vlan 0,1 to pool 0::
1330
1331   set port 0 mirror-rule 0 vlan-mirror 0,1 dst-pool 0 on
1332
1333reset port - mirror rule
1334~~~~~~~~~~~~~~~~~~~~~~~~
1335
1336Reset a mirror rule for a port::
1337
1338   testpmd> reset port (port_id) mirror-rule (rule_id)
1339
1340set flush_rx
1341~~~~~~~~~~~~
1342
1343Set the flush on RX streams before forwarding.
1344The default is flush ``on``.
1345Mainly used with PCAP drivers to turn off the default behavior of flushing the first 512 packets on RX streams::
1346
1347   testpmd> set flush_rx off
1348
1349set bypass mode
1350~~~~~~~~~~~~~~~
1351
1352Set the bypass mode for the lowest port on bypass enabled NIC::
1353
1354   testpmd> set bypass mode (normal|bypass|isolate) (port_id)
1355
1356set bypass event
1357~~~~~~~~~~~~~~~~
1358
1359Set the event required to initiate specified bypass mode for the lowest port on a bypass enabled::
1360
1361   testpmd> set bypass event (timeout|os_on|os_off|power_on|power_off) \
1362            mode (normal|bypass|isolate) (port_id)
1363
1364Where:
1365
1366* ``timeout``: Enable bypass after watchdog timeout.
1367
1368* ``os_on``: Enable bypass when OS/board is powered on.
1369
1370* ``os_off``: Enable bypass when OS/board is powered off.
1371
1372* ``power_on``: Enable bypass when power supply is turned on.
1373
1374* ``power_off``: Enable bypass when power supply is turned off.
1375
1376
1377set bypass timeout
1378~~~~~~~~~~~~~~~~~~
1379
1380Set the bypass watchdog timeout to ``n`` seconds where 0 = instant::
1381
1382   testpmd> set bypass timeout (0|1.5|2|3|4|8|16|32)
1383
1384show bypass config
1385~~~~~~~~~~~~~~~~~~
1386
1387Show the bypass configuration for a bypass enabled NIC using the lowest port on the NIC::
1388
1389   testpmd> show bypass config (port_id)
1390
1391set link up
1392~~~~~~~~~~~
1393
1394Set link up for a port::
1395
1396   testpmd> set link-up port (port id)
1397
1398set link down
1399~~~~~~~~~~~~~
1400
1401Set link down for a port::
1402
1403   testpmd> set link-down port (port id)
1404
1405E-tag set
1406~~~~~~~~~
1407
1408Enable E-tag insertion for a VF on a port::
1409
1410   testpmd> E-tag set insertion on port-tag-id (value) port (port_id) vf (vf_id)
1411
1412Disable E-tag insertion for a VF on a port::
1413
1414   testpmd> E-tag set insertion off port (port_id) vf (vf_id)
1415
1416Enable/disable E-tag stripping on a port::
1417
1418   testpmd> E-tag set stripping (on|off) port (port_id)
1419
1420Enable/disable E-tag based forwarding on a port::
1421
1422   testpmd> E-tag set forwarding (on|off) port (port_id)
1423
1424Add an E-tag forwarding filter on a port::
1425
1426   testpmd> E-tag set filter add e-tag-id (value) dst-pool (pool_id) port (port_id)
1427
1428Delete an E-tag forwarding filter on a port::
1429   testpmd> E-tag set filter del e-tag-id (value) port (port_id)
1430
1431ddp add
1432~~~~~~~
1433
1434Load a dynamic device personalization (DDP) profile and store backup profile::
1435
1436   testpmd> ddp add (port_id) (profile_path[,backup_profile_path])
1437
1438ddp del
1439~~~~~~~
1440
1441Delete a dynamic device personalization profile and restore backup profile::
1442
1443   testpmd> ddp del (port_id) (backup_profile_path)
1444
1445ptype mapping
1446~~~~~~~~~~~~~
1447
1448List all items from the ptype mapping table::
1449
1450   testpmd> ptype mapping get (port_id) (valid_only)
1451
1452Where:
1453
1454* ``valid_only``: A flag indicates if only list valid items(=1) or all itemss(=0).
1455
1456Replace a specific or a group of software defined ptype with a new one::
1457
1458   testpmd> ptype mapping replace  (port_id) (target) (mask) (pkt_type)
1459
1460where:
1461
1462* ``target``: A specific software ptype or a mask to represent a group of software ptypes.
1463
1464* ``mask``: A flag indicate if "target" is a specific software ptype(=0) or a ptype mask(=1).
1465
1466* ``pkt_type``: The new software ptype to replace the old ones.
1467
1468Update hardware defined ptype to software defined packet type mapping table::
1469
1470   testpmd> ptype mapping update (port_id) (hw_ptype) (sw_ptype)
1471
1472where:
1473
1474* ``hw_ptype``: hardware ptype as the index of the ptype mapping table.
1475
1476* ``sw_ptype``: software ptype as the value of the ptype mapping table.
1477
1478Reset ptype mapping table::
1479
1480   testpmd> ptype mapping reset (port_id)
1481
1482config per port Rx offloading
1483~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
1484
1485Enable or disable a per port Rx offloading on all Rx queues of a port::
1486
1487   testpmd> port config (port_id) rx_offload (offloading) on|off
1488
1489* ``offloading``: can be any of these offloading capability:
1490                  vlan_strip, ipv4_cksum, udp_cksum, tcp_cksum, tcp_lro,
1491                  qinq_strip, outer_ipv4_cksum, macsec_strip,
1492                  header_split, vlan_filter, vlan_extend, jumbo_frame,
1493                  crc_strip, scatter, timestamp, security
1494
1495This command should be run when the port is stopped, or else it will fail.
1496
1497config per queue Rx offloading
1498~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
1499
1500Enable or disable a per queue Rx offloading only on a specific Rx queue::
1501
1502   testpmd> port (port_id) rxq (queue_id) rx_offload (offloading) on|off
1503
1504* ``offloading``: can be any of these offloading capability:
1505                  vlan_strip, ipv4_cksum, udp_cksum, tcp_cksum, tcp_lro,
1506                  qinq_strip, outer_ipv4_cksum, macsec_strip,
1507                  header_split, vlan_filter, vlan_extend, jumbo_frame,
1508                  crc_strip, scatter, timestamp, security
1509
1510This command should be run when the port is stopped, or else it will fail.
1511
1512config per port Tx offloading
1513~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
1514
1515Enable or disable a per port Tx offloading on all Tx queues of a port::
1516
1517   testpmd> port config (port_id) tx_offload (offloading) on|off
1518
1519* ``offloading``: can be any of these offloading capability:
1520                  vlan_insert, ipv4_cksum, udp_cksum, tcp_cksum,
1521                  sctp_cksum, tcp_tso, udp_tso, outer_ipv4_cksum,
1522                  qinq_insert, vxlan_tnl_tso, gre_tnl_tso,
1523                  ipip_tnl_tso, geneve_tnl_tso, macsec_insert,
1524                  mt_lockfree, multi_segs, fast_free, security
1525
1526This command should be run when the port is stopped, or else it will fail.
1527
1528config per queue Tx offloading
1529~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
1530
1531Enable or disable a per queue Tx offloading only on a specific Tx queue::
1532
1533   testpmd> port (port_id) txq (queue_id) tx_offload (offloading) on|off
1534
1535* ``offloading``: can be any of these offloading capability:
1536                  vlan_insert, ipv4_cksum, udp_cksum, tcp_cksum,
1537                  sctp_cksum, tcp_tso, udp_tso, outer_ipv4_cksum,
1538                  qinq_insert, vxlan_tnl_tso, gre_tnl_tso,
1539                  ipip_tnl_tso, geneve_tnl_tso, macsec_insert,
1540                  mt_lockfree, multi_segs, fast_free, security
1541
1542This command should be run when the port is stopped, or else it will fail.
1543
1544Config VXLAN Encap outer layers
1545~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
1546
1547Configure the outer layer to encapsulate a packet inside a VXLAN tunnel::
1548
1549 set vxlan ip-version (ipv4|ipv6) vni (vni) udp-src (udp-src) \
1550 udp-dst (udp-dst) ip-src (ip-src) ip-dst (ip-dst) eth-src (eth-src) \
1551 eth-dst (eth-dst)
1552
1553 set vxlan-with-vlan ip-version (ipv4|ipv6) vni (vni) udp-src (udp-src) \
1554 udp-dst (udp-dst) ip-src (ip-src) ip-dst (ip-dst) vlan-tci (vlan-tci) \
1555 eth-src (eth-src) eth-dst (eth-dst)
1556
1557Those command will set an internal configuration inside testpmd, any following
1558flow rule using the action vxlan_encap will use the last configuration set.
1559To have a different encapsulation header, one of those commands must be called
1560before the flow rule creation.
1561
1562Port Functions
1563--------------
1564
1565The following sections show functions for configuring ports.
1566
1567.. note::
1568
1569   Port configuration changes only become active when forwarding is started/restarted.
1570
1571port attach
1572~~~~~~~~~~~
1573
1574Attach a port specified by pci address or virtual device args::
1575
1576   testpmd> port attach (identifier)
1577
1578To attach a new pci device, the device should be recognized by kernel first.
1579Then it should be moved under DPDK management.
1580Finally the port can be attached to testpmd.
1581
1582For example, to move a pci device using ixgbe under DPDK management:
1583
1584.. code-block:: console
1585
1586   # Check the status of the available devices.
1587   ./usertools/dpdk-devbind.py --status
1588
1589   Network devices using DPDK-compatible driver
1590   ============================================
1591   <none>
1592
1593   Network devices using kernel driver
1594   ===================================
1595   0000:0a:00.0 '82599ES 10-Gigabit' if=eth2 drv=ixgbe unused=
1596
1597
1598   # Bind the device to igb_uio.
1599   sudo ./usertools/dpdk-devbind.py -b igb_uio 0000:0a:00.0
1600
1601
1602   # Recheck the status of the devices.
1603   ./usertools/dpdk-devbind.py --status
1604   Network devices using DPDK-compatible driver
1605   ============================================
1606   0000:0a:00.0 '82599ES 10-Gigabit' drv=igb_uio unused=
1607
1608To attach a port created by virtual device, above steps are not needed.
1609
1610For example, to attach a port whose pci address is 0000:0a:00.0.
1611
1612.. code-block:: console
1613
1614   testpmd> port attach 0000:0a:00.0
1615   Attaching a new port...
1616   EAL: PCI device 0000:0a:00.0 on NUMA socket -1
1617   EAL:   probe driver: 8086:10fb rte_ixgbe_pmd
1618   EAL:   PCI memory mapped at 0x7f83bfa00000
1619   EAL:   PCI memory mapped at 0x7f83bfa80000
1620   PMD: eth_ixgbe_dev_init(): MAC: 2, PHY: 18, SFP+: 5
1621   PMD: eth_ixgbe_dev_init(): port 0 vendorID=0x8086 deviceID=0x10fb
1622   Port 0 is attached. Now total ports is 1
1623   Done
1624
1625For example, to attach a port created by pcap PMD.
1626
1627.. code-block:: console
1628
1629   testpmd> port attach net_pcap0
1630   Attaching a new port...
1631   PMD: Initializing pmd_pcap for net_pcap0
1632   PMD: Creating pcap-backed ethdev on numa socket 0
1633   Port 0 is attached. Now total ports is 1
1634   Done
1635
1636In this case, identifier is ``net_pcap0``.
1637This identifier format is the same as ``--vdev`` format of DPDK applications.
1638
1639For example, to re-attach a bonded port which has been previously detached,
1640the mode and slave parameters must be given.
1641
1642.. code-block:: console
1643
1644   testpmd> port attach net_bond_0,mode=0,slave=1
1645   Attaching a new port...
1646   EAL: Initializing pmd_bond for net_bond_0
1647   EAL: Create bonded device net_bond_0 on port 0 in mode 0 on socket 0.
1648   Port 0 is attached. Now total ports is 1
1649   Done
1650
1651
1652port detach
1653~~~~~~~~~~~
1654
1655Detach a specific port::
1656
1657   testpmd> port detach (port_id)
1658
1659Before detaching a port, the port should be stopped and closed.
1660
1661For example, to detach a pci device port 0.
1662
1663.. code-block:: console
1664
1665   testpmd> port stop 0
1666   Stopping ports...
1667   Done
1668   testpmd> port close 0
1669   Closing ports...
1670   Done
1671
1672   testpmd> port detach 0
1673   Detaching a port...
1674   EAL: PCI device 0000:0a:00.0 on NUMA socket -1
1675   EAL:   remove driver: 8086:10fb rte_ixgbe_pmd
1676   EAL:   PCI memory unmapped at 0x7f83bfa00000
1677   EAL:   PCI memory unmapped at 0x7f83bfa80000
1678   Done
1679
1680
1681For example, to detach a virtual device port 0.
1682
1683.. code-block:: console
1684
1685   testpmd> port stop 0
1686   Stopping ports...
1687   Done
1688   testpmd> port close 0
1689   Closing ports...
1690   Done
1691
1692   testpmd> port detach 0
1693   Detaching a port...
1694   PMD: Closing pcap ethdev on numa socket 0
1695   Port 'net_pcap0' is detached. Now total ports is 0
1696   Done
1697
1698To remove a pci device completely from the system, first detach the port from testpmd.
1699Then the device should be moved under kernel management.
1700Finally the device can be removed using kernel pci hotplug functionality.
1701
1702For example, to move a pci device under kernel management:
1703
1704.. code-block:: console
1705
1706   sudo ./usertools/dpdk-devbind.py -b ixgbe 0000:0a:00.0
1707
1708   ./usertools/dpdk-devbind.py --status
1709
1710   Network devices using DPDK-compatible driver
1711   ============================================
1712   <none>
1713
1714   Network devices using kernel driver
1715   ===================================
1716   0000:0a:00.0 '82599ES 10-Gigabit' if=eth2 drv=ixgbe unused=igb_uio
1717
1718To remove a port created by a virtual device, above steps are not needed.
1719
1720port start
1721~~~~~~~~~~
1722
1723Start all ports or a specific port::
1724
1725   testpmd> port start (port_id|all)
1726
1727port stop
1728~~~~~~~~~
1729
1730Stop all ports or a specific port::
1731
1732   testpmd> port stop (port_id|all)
1733
1734port close
1735~~~~~~~~~~
1736
1737Close all ports or a specific port::
1738
1739   testpmd> port close (port_id|all)
1740
1741port config - queue ring size
1742~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
1743
1744Configure a rx/tx queue ring size::
1745
1746   testpmd> port (port_id) (rxq|txq) (queue_id) ring_size (value)
1747
1748Only take effect after command that (re-)start the port or command that setup specific queue.
1749
1750port start/stop queue
1751~~~~~~~~~~~~~~~~~~~~~
1752
1753Start/stop a rx/tx queue on a specific port::
1754
1755   testpmd> port (port_id) (rxq|txq) (queue_id) (start|stop)
1756
1757port setup queue
1758~~~~~~~~~~~~~~~~~~~~~
1759
1760Setup a rx/tx queue on a specific port::
1761
1762   testpmd> port (port_id) (rxq|txq) (queue_id) setup
1763
1764Only take effect when port is started.
1765
1766port config - speed
1767~~~~~~~~~~~~~~~~~~~
1768
1769Set the speed and duplex mode for all ports or a specific port::
1770
1771   testpmd> port config (port_id|all) speed (10|100|1000|10000|25000|40000|50000|100000|auto) \
1772            duplex (half|full|auto)
1773
1774port config - queues/descriptors
1775~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
1776
1777Set number of queues/descriptors for rxq, txq, rxd and txd::
1778
1779   testpmd> port config all (rxq|txq|rxd|txd) (value)
1780
1781This is equivalent to the ``--rxq``, ``--txq``, ``--rxd`` and ``--txd`` command-line options.
1782
1783port config - max-pkt-len
1784~~~~~~~~~~~~~~~~~~~~~~~~~
1785
1786Set the maximum packet length::
1787
1788   testpmd> port config all max-pkt-len (value)
1789
1790This is equivalent to the ``--max-pkt-len`` command-line option.
1791
1792port config - CRC Strip
1793~~~~~~~~~~~~~~~~~~~~~~~
1794
1795Set hardware CRC stripping on or off for all ports::
1796
1797   testpmd> port config all crc-strip (on|off)
1798
1799CRC stripping is on by default.
1800
1801The ``off`` option is equivalent to the ``--disable-crc-strip`` command-line option.
1802
1803port config - scatter
1804~~~~~~~~~~~~~~~~~~~~~~~
1805
1806Set RX scatter mode on or off for all ports::
1807
1808   testpmd> port config all scatter (on|off)
1809
1810RX scatter mode is off by default.
1811
1812The ``on`` option is equivalent to the ``--enable-scatter`` command-line option.
1813
1814port config - RX Checksum
1815~~~~~~~~~~~~~~~~~~~~~~~~~
1816
1817Set hardware RX checksum offload to on or off for all ports::
1818
1819   testpmd> port config all rx-cksum (on|off)
1820
1821Checksum offload is off by default.
1822
1823The ``on`` option is equivalent to the ``--enable-rx-cksum`` command-line option.
1824
1825port config - VLAN
1826~~~~~~~~~~~~~~~~~~
1827
1828Set hardware VLAN on or off for all ports::
1829
1830   testpmd> port config all hw-vlan (on|off)
1831
1832Hardware VLAN is off by default.
1833
1834The ``on`` option is equivalent to the ``--enable-hw-vlan`` command-line option.
1835
1836port config - VLAN filter
1837~~~~~~~~~~~~~~~~~~~~~~~~~
1838
1839Set hardware VLAN filter on or off for all ports::
1840
1841   testpmd> port config all hw-vlan-filter (on|off)
1842
1843Hardware VLAN filter is off by default.
1844
1845The ``on`` option is equivalent to the ``--enable-hw-vlan-filter`` command-line option.
1846
1847port config - VLAN strip
1848~~~~~~~~~~~~~~~~~~~~~~~~
1849
1850Set hardware VLAN strip on or off for all ports::
1851
1852   testpmd> port config all hw-vlan-strip (on|off)
1853
1854Hardware VLAN strip is off by default.
1855
1856The ``on`` option is equivalent to the ``--enable-hw-vlan-strip`` command-line option.
1857
1858port config - VLAN extend
1859~~~~~~~~~~~~~~~~~~~~~~~~~
1860
1861Set hardware VLAN extend on or off for all ports::
1862
1863   testpmd> port config all hw-vlan-extend (on|off)
1864
1865Hardware VLAN extend is off by default.
1866
1867The ``on`` option is equivalent to the ``--enable-hw-vlan-extend`` command-line option.
1868
1869port config - Drop Packets
1870~~~~~~~~~~~~~~~~~~~~~~~~~~
1871
1872Set packet drop for packets with no descriptors on or off for all ports::
1873
1874   testpmd> port config all drop-en (on|off)
1875
1876Packet dropping for packets with no descriptors is off by default.
1877
1878The ``on`` option is equivalent to the ``--enable-drop-en`` command-line option.
1879
1880port config - RSS
1881~~~~~~~~~~~~~~~~~
1882
1883Set the RSS (Receive Side Scaling) mode on or off::
1884
1885   testpmd> port config all rss (all|default|ip|tcp|udp|sctp|ether|port|vxlan|geneve|nvgre|none)
1886
1887RSS is on by default.
1888
1889The ``all`` option is equivalent to ip|tcp|udp|sctp|ether.
1890The ``default`` option enables all supported RSS types reported by device info.
1891The ``none`` option is equivalent to the ``--disable-rss`` command-line option.
1892
1893port config - RSS Reta
1894~~~~~~~~~~~~~~~~~~~~~~
1895
1896Set the RSS (Receive Side Scaling) redirection table::
1897
1898   testpmd> port config all rss reta (hash,queue)[,(hash,queue)]
1899
1900port config - DCB
1901~~~~~~~~~~~~~~~~~
1902
1903Set the DCB mode for an individual port::
1904
1905   testpmd> port config (port_id) dcb vt (on|off) (traffic_class) pfc (on|off)
1906
1907The traffic class should be 4 or 8.
1908
1909port config - Burst
1910~~~~~~~~~~~~~~~~~~~
1911
1912Set the number of packets per burst::
1913
1914   testpmd> port config all burst (value)
1915
1916This is equivalent to the ``--burst`` command-line option.
1917
1918port config - Threshold
1919~~~~~~~~~~~~~~~~~~~~~~~
1920
1921Set thresholds for TX/RX queues::
1922
1923   testpmd> port config all (threshold) (value)
1924
1925Where the threshold type can be:
1926
1927* ``txpt:`` Set the prefetch threshold register of the TX rings, 0 <= value <= 255.
1928
1929* ``txht:`` Set the host threshold register of the TX rings, 0 <= value <= 255.
1930
1931* ``txwt:`` Set the write-back threshold register of the TX rings, 0 <= value <= 255.
1932
1933* ``rxpt:`` Set the prefetch threshold register of the RX rings, 0 <= value <= 255.
1934
1935* ``rxht:`` Set the host threshold register of the RX rings, 0 <= value <= 255.
1936
1937* ``rxwt:`` Set the write-back threshold register of the RX rings, 0 <= value <= 255.
1938
1939* ``txfreet:`` Set the transmit free threshold of the TX rings, 0 <= value <= txd.
1940
1941* ``rxfreet:`` Set the transmit free threshold of the RX rings, 0 <= value <= rxd.
1942
1943* ``txrst:`` Set the transmit RS bit threshold of TX rings, 0 <= value <= txd.
1944
1945These threshold options are also available from the command-line.
1946
1947port config - E-tag
1948~~~~~~~~~~~~~~~~~~~
1949
1950Set the value of ether-type for E-tag::
1951
1952   testpmd> port config (port_id|all) l2-tunnel E-tag ether-type (value)
1953
1954Enable/disable the E-tag support::
1955
1956   testpmd> port config (port_id|all) l2-tunnel E-tag (enable|disable)
1957
1958port config pctype mapping
1959~~~~~~~~~~~~~~~~~~~~~~~~~~
1960
1961Reset pctype mapping table::
1962
1963   testpmd> port config (port_id) pctype mapping reset
1964
1965Update hardware defined pctype to software defined flow type mapping table::
1966
1967   testpmd> port config (port_id) pctype mapping update (pctype_id_0[,pctype_id_1]*) (flow_type_id)
1968
1969where:
1970
1971* ``pctype_id_x``: hardware pctype id as index of bit in bitmask value of the pctype mapping table.
1972
1973* ``flow_type_id``: software flow type id as the index of the pctype mapping table.
1974
1975port config input set
1976~~~~~~~~~~~~~~~~~~~~~
1977
1978Config RSS/FDIR/FDIR flexible payload input set for some pctype::
1979   testpmd> port config (port_id) pctype (pctype_id) \
1980            (hash_inset|fdir_inset|fdir_flx_inset) \
1981	    (get|set|clear) field (field_idx)
1982
1983Clear RSS/FDIR/FDIR flexible payload input set for some pctype::
1984   testpmd> port config (port_id) pctype (pctype_id) \
1985            (hash_inset|fdir_inset|fdir_flx_inset) clear all
1986
1987where:
1988
1989* ``pctype_id``: hardware packet classification types.
1990* ``field_idx``: hardware field index.
1991
1992port config udp_tunnel_port
1993~~~~~~~~~~~~~~~~~~~~~~~~~~~
1994
1995Add/remove UDP tunnel port for VXLAN/GENEVE tunneling protocols::
1996    testpmd> port config (port_id) udp_tunnel_port add|rm vxlan|geneve (udp_port)
1997
1998Link Bonding Functions
1999----------------------
2000
2001The Link Bonding functions make it possible to dynamically create and
2002manage link bonding devices from within testpmd interactive prompt.
2003
2004create bonded device
2005~~~~~~~~~~~~~~~~~~~~
2006
2007Create a new bonding device::
2008
2009   testpmd> create bonded device (mode) (socket)
2010
2011For example, to create a bonded device in mode 1 on socket 0::
2012
2013   testpmd> create bonded 1 0
2014   created new bonded device (port X)
2015
2016add bonding slave
2017~~~~~~~~~~~~~~~~~
2018
2019Adds Ethernet device to a Link Bonding device::
2020
2021   testpmd> add bonding slave (slave id) (port id)
2022
2023For example, to add Ethernet device (port 6) to a Link Bonding device (port 10)::
2024
2025   testpmd> add bonding slave 6 10
2026
2027
2028remove bonding slave
2029~~~~~~~~~~~~~~~~~~~~
2030
2031Removes an Ethernet slave device from a Link Bonding device::
2032
2033   testpmd> remove bonding slave (slave id) (port id)
2034
2035For example, to remove Ethernet slave device (port 6) to a Link Bonding device (port 10)::
2036
2037   testpmd> remove bonding slave 6 10
2038
2039set bonding mode
2040~~~~~~~~~~~~~~~~
2041
2042Set the Link Bonding mode of a Link Bonding device::
2043
2044   testpmd> set bonding mode (value) (port id)
2045
2046For example, to set the bonding mode of a Link Bonding device (port 10) to broadcast (mode 3)::
2047
2048   testpmd> set bonding mode 3 10
2049
2050set bonding primary
2051~~~~~~~~~~~~~~~~~~~
2052
2053Set an Ethernet slave device as the primary device on a Link Bonding device::
2054
2055   testpmd> set bonding primary (slave id) (port id)
2056
2057For example, to set the Ethernet slave device (port 6) as the primary port of a Link Bonding device (port 10)::
2058
2059   testpmd> set bonding primary 6 10
2060
2061set bonding mac
2062~~~~~~~~~~~~~~~
2063
2064Set the MAC address of a Link Bonding device::
2065
2066   testpmd> set bonding mac (port id) (mac)
2067
2068For example, to set the MAC address of a Link Bonding device (port 10) to 00:00:00:00:00:01::
2069
2070   testpmd> set bonding mac 10 00:00:00:00:00:01
2071
2072set bonding xmit_balance_policy
2073~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
2074
2075Set the transmission policy for a Link Bonding device when it is in Balance XOR mode::
2076
2077   testpmd> set bonding xmit_balance_policy (port_id) (l2|l23|l34)
2078
2079For example, set a Link Bonding device (port 10) to use a balance policy of layer 3+4 (IP addresses & UDP ports)::
2080
2081   testpmd> set bonding xmit_balance_policy 10 l34
2082
2083
2084set bonding mon_period
2085~~~~~~~~~~~~~~~~~~~~~~
2086
2087Set the link status monitoring polling period in milliseconds for a bonding device.
2088
2089This adds support for PMD slave devices which do not support link status interrupts.
2090When the mon_period is set to a value greater than 0 then all PMD's which do not support
2091link status ISR will be queried every polling interval to check if their link status has changed::
2092
2093   testpmd> set bonding mon_period (port_id) (value)
2094
2095For example, to set the link status monitoring polling period of bonded device (port 5) to 150ms::
2096
2097   testpmd> set bonding mon_period 5 150
2098
2099
2100set bonding lacp dedicated_queue
2101~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
2102
2103Enable dedicated tx/rx queues on bonding devices slaves to handle LACP control plane traffic
2104when in mode 4 (link-aggregration-802.3ad)::
2105
2106   testpmd> set bonding lacp dedicated_queues (port_id) (enable|disable)
2107
2108
2109set bonding agg_mode
2110~~~~~~~~~~~~~~~~~~~~
2111
2112Enable one of the specific aggregators mode when in mode 4 (link-aggregration-802.3ad)::
2113
2114   testpmd> set bonding agg_mode (port_id) (bandwidth|count|stable)
2115
2116
2117show bonding config
2118~~~~~~~~~~~~~~~~~~~
2119
2120Show the current configuration of a Link Bonding device::
2121
2122   testpmd> show bonding config (port id)
2123
2124For example,
2125to show the configuration a Link Bonding device (port 9) with 3 slave devices (1, 3, 4)
2126in balance mode with a transmission policy of layer 2+3::
2127
2128   testpmd> show bonding config 9
2129        Bonding mode: 2
2130        Balance Xmit Policy: BALANCE_XMIT_POLICY_LAYER23
2131        Slaves (3): [1 3 4]
2132        Active Slaves (3): [1 3 4]
2133        Primary: [3]
2134
2135
2136Register Functions
2137------------------
2138
2139The Register Functions can be used to read from and write to registers on the network card referenced by a port number.
2140This is mainly useful for debugging purposes.
2141Reference should be made to the appropriate datasheet for the network card for details on the register addresses
2142and fields that can be accessed.
2143
2144read reg
2145~~~~~~~~
2146
2147Display the value of a port register::
2148
2149   testpmd> read reg (port_id) (address)
2150
2151For example, to examine the Flow Director control register (FDIRCTL, 0x0000EE000) on an Intel 82599 10 GbE Controller::
2152
2153   testpmd> read reg 0 0xEE00
2154   port 0 PCI register at offset 0xEE00: 0x4A060029 (1241907241)
2155
2156read regfield
2157~~~~~~~~~~~~~
2158
2159Display a port register bit field::
2160
2161   testpmd> read regfield (port_id) (address) (bit_x) (bit_y)
2162
2163For example, reading the lowest two bits from the register in the example above::
2164
2165   testpmd> read regfield 0 0xEE00 0 1
2166   port 0 PCI register at offset 0xEE00: bits[0, 1]=0x1 (1)
2167
2168read regbit
2169~~~~~~~~~~~
2170
2171Display a single port register bit::
2172
2173   testpmd> read regbit (port_id) (address) (bit_x)
2174
2175For example, reading the lowest bit from the register in the example above::
2176
2177   testpmd> read regbit 0 0xEE00 0
2178   port 0 PCI register at offset 0xEE00: bit 0=1
2179
2180write reg
2181~~~~~~~~~
2182
2183Set the value of a port register::
2184
2185   testpmd> write reg (port_id) (address) (value)
2186
2187For example, to clear a register::
2188
2189   testpmd> write reg 0 0xEE00 0x0
2190   port 0 PCI register at offset 0xEE00: 0x00000000 (0)
2191
2192write regfield
2193~~~~~~~~~~~~~~
2194
2195Set bit field of a port register::
2196
2197   testpmd> write regfield (port_id) (address) (bit_x) (bit_y) (value)
2198
2199For example, writing to the register cleared in the example above::
2200
2201   testpmd> write regfield 0 0xEE00 0 1 2
2202   port 0 PCI register at offset 0xEE00: 0x00000002 (2)
2203
2204write regbit
2205~~~~~~~~~~~~
2206
2207Set single bit value of a port register::
2208
2209   testpmd> write regbit (port_id) (address) (bit_x) (value)
2210
2211For example, to set the high bit in the register from the example above::
2212
2213   testpmd> write regbit 0 0xEE00 31 1
2214   port 0 PCI register at offset 0xEE00: 0x8000000A (2147483658)
2215
2216Traffic Metering and Policing
2217-----------------------------
2218
2219The following section shows functions for configuring traffic metering and
2220policing on the ethernet device through the use of generic ethdev API.
2221
2222show port traffic management capability
2223~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
2224
2225Show traffic metering and policing capability of the port::
2226
2227   testpmd> show port meter cap (port_id)
2228
2229add port meter profile (srTCM rfc2967)
2230~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
2231
2232Add meter profile (srTCM rfc2697) to the ethernet device::
2233
2234   testpmd> add port meter profile srtcm_rfc2697 (port_id) (profile_id) \
2235   (cir) (cbs) (ebs)
2236
2237where:
2238
2239* ``profile_id``: ID for the meter profile.
2240* ``cir``: Committed Information Rate (CIR) (bytes/second).
2241* ``cbs``: Committed Burst Size (CBS) (bytes).
2242* ``ebs``: Excess Burst Size (EBS) (bytes).
2243
2244add port meter profile (trTCM rfc2968)
2245~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
2246
2247Add meter profile (srTCM rfc2698) to the ethernet device::
2248
2249   testpmd> add port meter profile trtcm_rfc2698 (port_id) (profile_id) \
2250   (cir) (pir) (cbs) (pbs)
2251
2252where:
2253
2254* ``profile_id``: ID for the meter profile.
2255* ``cir``: Committed information rate (bytes/second).
2256* ``pir``: Peak information rate (bytes/second).
2257* ``cbs``: Committed burst size (bytes).
2258* ``pbs``: Peak burst size (bytes).
2259
2260add port meter profile (trTCM rfc4115)
2261~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
2262
2263Add meter profile (trTCM rfc4115) to the ethernet device::
2264
2265   testpmd> add port meter profile trtcm_rfc4115 (port_id) (profile_id) \
2266   (cir) (eir) (cbs) (ebs)
2267
2268where:
2269
2270* ``profile_id``: ID for the meter profile.
2271* ``cir``: Committed information rate (bytes/second).
2272* ``eir``: Excess information rate (bytes/second).
2273* ``cbs``: Committed burst size (bytes).
2274* ``ebs``: Excess burst size (bytes).
2275
2276delete port meter profile
2277~~~~~~~~~~~~~~~~~~~~~~~~~
2278
2279Delete meter profile from the ethernet device::
2280
2281   testpmd> del port meter profile (port_id) (profile_id)
2282
2283create port meter
2284~~~~~~~~~~~~~~~~~
2285
2286Create new meter object for the ethernet device::
2287
2288   testpmd> create port meter (port_id) (mtr_id) (profile_id) \
2289   (meter_enable) (g_action) (y_action) (r_action) (stats_mask) (shared) \
2290   (use_pre_meter_color) [(dscp_tbl_entry0) (dscp_tbl_entry1)...\
2291   (dscp_tbl_entry63)]
2292
2293where:
2294
2295* ``mtr_id``: meter object ID.
2296* ``profile_id``: ID for the meter profile.
2297* ``meter_enable``: When this parameter has a non-zero value, the meter object
2298  gets enabled at the time of creation, otherwise remains disabled.
2299* ``g_action``: Policer action for the packet with green color.
2300* ``y_action``: Policer action for the packet with yellow color.
2301* ``r_action``: Policer action for the packet with red color.
2302* ``stats_mask``: Mask of statistics counter types to be enabled for the
2303  meter object.
2304* ``shared``:  When this parameter has a non-zero value, the meter object is
2305  shared by multiple flows. Otherwise, meter object is used by single flow.
2306* ``use_pre_meter_color``: When this parameter has a non-zero value, the
2307  input color for the current meter object is determined by the latest meter
2308  object in the same flow. Otherwise, the current meter object uses the
2309  *dscp_table* to determine the input color.
2310* ``dscp_tbl_entryx``: DSCP table entry x providing meter providing input
2311  color, 0 <= x <= 63.
2312
2313enable port meter
2314~~~~~~~~~~~~~~~~~
2315
2316Enable meter for the ethernet device::
2317
2318   testpmd> enable port meter (port_id) (mtr_id)
2319
2320disable port meter
2321~~~~~~~~~~~~~~~~~~
2322
2323Disable meter for the ethernet device::
2324
2325   testpmd> disable port meter (port_id) (mtr_id)
2326
2327delete port meter
2328~~~~~~~~~~~~~~~~~
2329
2330Delete meter for the ethernet device::
2331
2332   testpmd> del port meter (port_id) (mtr_id)
2333
2334Set port meter profile
2335~~~~~~~~~~~~~~~~~~~~~~
2336
2337Set meter profile for the ethernet device::
2338
2339   testpmd> set port meter profile (port_id) (mtr_id) (profile_id)
2340
2341set port meter dscp table
2342~~~~~~~~~~~~~~~~~~~~~~~~~
2343
2344Set meter dscp table for the ethernet device::
2345
2346   testpmd> set port meter dscp table (port_id) (mtr_id) [(dscp_tbl_entry0) \
2347   (dscp_tbl_entry1)...(dscp_tbl_entry63)]
2348
2349set port meter policer action
2350~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
2351
2352Set meter policer action for the ethernet device::
2353
2354   testpmd> set port meter policer action (port_id) (mtr_id) (action_mask) \
2355   (action0) [(action1) (action1)]
2356
2357where:
2358
2359* ``action_mask``: Bit mask indicating which policer actions need to be
2360  updated. One or more policer actions can be updated in a single function
2361  invocation. To update the policer action associated with color C, bit
2362  (1 << C) needs to be set in *action_mask* and element at position C
2363  in the *actions* array needs to be valid.
2364* ``actionx``: Policer action for the color x,
2365  RTE_MTR_GREEN <= x < RTE_MTR_COLORS
2366
2367set port meter stats mask
2368~~~~~~~~~~~~~~~~~~~~~~~~~
2369
2370Set meter stats mask for the ethernet device::
2371
2372   testpmd> set port meter stats mask (port_id) (mtr_id) (stats_mask)
2373
2374where:
2375
2376* ``stats_mask``: Bit mask indicating statistics counter types to be enabled.
2377
2378show port meter stats
2379~~~~~~~~~~~~~~~~~~~~~
2380
2381Show meter stats of the ethernet device::
2382
2383   testpmd> show port meter stats (port_id) (mtr_id) (clear)
2384
2385where:
2386
2387* ``clear``: Flag that indicates whether the statistics counters should
2388  be cleared (i.e. set to zero) immediately after they have been read or not.
2389
2390Traffic Management
2391------------------
2392
2393The following section shows functions for configuring traffic management on
2394on the ethernet device through the use of generic TM API.
2395
2396show port traffic management capability
2397~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
2398
2399Show traffic management capability of the port::
2400
2401   testpmd> show port tm cap (port_id)
2402
2403show port traffic management capability (hierarchy level)
2404~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
2405
2406Show traffic management hierarchy level capability of the port::
2407
2408   testpmd> show port tm level cap (port_id) (level_id)
2409
2410show port traffic management capability (hierarchy node level)
2411~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
2412
2413Show the traffic management hierarchy node capability of the port::
2414
2415   testpmd> show port tm node cap (port_id) (node_id)
2416
2417show port traffic management hierarchy node type
2418~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
2419
2420Show the port traffic management hierarchy node type::
2421
2422   testpmd> show port tm node type (port_id) (node_id)
2423
2424show port traffic management hierarchy node stats
2425~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
2426
2427Show the port traffic management hierarchy node statistics::
2428
2429   testpmd> show port tm node stats (port_id) (node_id) (clear)
2430
2431where:
2432
2433* ``clear``: When this parameter has a non-zero value, the statistics counters
2434  are cleared (i.e. set to zero) immediately after they have been read,
2435  otherwise the statistics counters are left untouched.
2436
2437Add port traffic management private shaper profile
2438~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
2439
2440Add the port traffic management private shaper profile::
2441
2442   testpmd> add port tm node shaper profile (port_id) (shaper_profile_id) \
2443   (tb_rate) (tb_size) (packet_length_adjust)
2444
2445where:
2446
2447* ``shaper_profile id``: Shaper profile ID for the new profile.
2448* ``tb_rate``: Token bucket rate (bytes per second).
2449* ``tb_size``: Token bucket size (bytes).
2450* ``packet_length_adjust``: The value (bytes) to be added to the length of
2451  each packet for the purpose of shaping. This parameter value can be used to
2452  correct the packet length with the framing overhead bytes that are consumed
2453  on the wire.
2454
2455Delete port traffic management private shaper profile
2456~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
2457
2458Delete the port traffic management private shaper::
2459
2460   testpmd> del port tm node shaper profile (port_id) (shaper_profile_id)
2461
2462where:
2463
2464* ``shaper_profile id``: Shaper profile ID that needs to be deleted.
2465
2466Add port traffic management shared shaper
2467~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
2468
2469Create the port traffic management shared shaper::
2470
2471   testpmd> add port tm node shared shaper (port_id) (shared_shaper_id) \
2472   (shaper_profile_id)
2473
2474where:
2475
2476* ``shared_shaper_id``: Shared shaper ID to be created.
2477* ``shaper_profile id``: Shaper profile ID for shared shaper.
2478
2479Set port traffic management shared shaper
2480~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
2481
2482Update the port traffic management shared shaper::
2483
2484   testpmd> set port tm node shared shaper (port_id) (shared_shaper_id) \
2485   (shaper_profile_id)
2486
2487where:
2488
2489* ``shared_shaper_id``: Shared shaper ID to be update.
2490* ``shaper_profile id``: Shaper profile ID for shared shaper.
2491
2492Delete port traffic management shared shaper
2493~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
2494
2495Delete the port traffic management shared shaper::
2496
2497   testpmd> del port tm node shared shaper (port_id) (shared_shaper_id)
2498
2499where:
2500
2501* ``shared_shaper_id``: Shared shaper ID to be deleted.
2502
2503Set port traffic management hiearchy node private shaper
2504~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
2505
2506set the port traffic management hierarchy node private shaper::
2507
2508   testpmd> set port tm node shaper profile (port_id) (node_id) \
2509   (shaper_profile_id)
2510
2511where:
2512
2513* ``shaper_profile id``: Private shaper profile ID to be enabled on the
2514  hierarchy node.
2515
2516Add port traffic management WRED profile
2517~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
2518
2519Create a new WRED profile::
2520
2521   testpmd> add port tm node wred profile (port_id) (wred_profile_id) \
2522   (color_g) (min_th_g) (max_th_g) (maxp_inv_g) (wq_log2_g) \
2523   (color_y) (min_th_y) (max_th_y) (maxp_inv_y) (wq_log2_y) \
2524   (color_r) (min_th_r) (max_th_r) (maxp_inv_r) (wq_log2_r)
2525
2526where:
2527
2528* ``wred_profile id``: Identifier for the newly create WRED profile
2529* ``color_g``: Packet color (green)
2530* ``min_th_g``: Minimum queue threshold for packet with green color
2531* ``max_th_g``: Minimum queue threshold for packet with green color
2532* ``maxp_inv_g``: Inverse of packet marking probability maximum value (maxp)
2533* ``wq_log2_g``: Negated log2 of queue weight (wq)
2534* ``color_y``: Packet color (yellow)
2535* ``min_th_y``: Minimum queue threshold for packet with yellow color
2536* ``max_th_y``: Minimum queue threshold for packet with yellow color
2537* ``maxp_inv_y``: Inverse of packet marking probability maximum value (maxp)
2538* ``wq_log2_y``: Negated log2 of queue weight (wq)
2539* ``color_r``: Packet color (red)
2540* ``min_th_r``: Minimum queue threshold for packet with yellow color
2541* ``max_th_r``: Minimum queue threshold for packet with yellow color
2542* ``maxp_inv_r``: Inverse of packet marking probability maximum value (maxp)
2543* ``wq_log2_r``: Negated log2 of queue weight (wq)
2544
2545Delete port traffic management WRED profile
2546~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
2547
2548Delete the WRED profile::
2549
2550   testpmd> del port tm node wred profile (port_id) (wred_profile_id)
2551
2552Add port traffic management hierarchy nonleaf node
2553~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
2554
2555Add nonleaf node to port traffic management hiearchy::
2556
2557   testpmd> add port tm nonleaf node (port_id) (node_id) (parent_node_id) \
2558   (priority) (weight) (level_id) (shaper_profile_id) \
2559   (n_sp_priorities) (stats_mask) (n_shared_shapers) \
2560   [(shared_shaper_0) (shared_shaper_1) ...] \
2561
2562where:
2563
2564* ``parent_node_id``: Node ID of the parent.
2565* ``priority``: Node priority (highest node priority is zero). This is used by
2566  the SP algorithm running on the parent node for scheduling this node.
2567* ``weight``: Node weight (lowest weight is one). The node weight is relative
2568  to the weight sum of all siblings that have the same priority. It is used by
2569  the WFQ algorithm running on the parent node for scheduling this node.
2570* ``level_id``: Hiearchy level of the node.
2571* ``shaper_profile_id``: Shaper profile ID of the private shaper to be used by
2572  the node.
2573* ``n_sp_priorities``: Number of strict priorities.
2574* ``stats_mask``: Mask of statistics counter types to be enabled for this node.
2575* ``n_shared_shapers``: Number of shared shapers.
2576* ``shared_shaper_id``: Shared shaper id.
2577
2578Add port traffic management hierarchy leaf node
2579~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
2580
2581Add leaf node to port traffic management hiearchy::
2582
2583   testpmd> add port tm leaf node (port_id) (node_id) (parent_node_id) \
2584   (priority) (weight) (level_id) (shaper_profile_id) \
2585   (cman_mode) (wred_profile_id) (stats_mask) (n_shared_shapers) \
2586   [(shared_shaper_id) (shared_shaper_id) ...] \
2587
2588where:
2589
2590* ``parent_node_id``: Node ID of the parent.
2591* ``priority``: Node priority (highest node priority is zero). This is used by
2592  the SP algorithm running on the parent node for scheduling this node.
2593* ``weight``: Node weight (lowest weight is one). The node weight is relative
2594  to the weight sum of all siblings that have the same priority. It is used by
2595  the WFQ algorithm running on the parent node for scheduling this node.
2596* ``level_id``: Hiearchy level of the node.
2597* ``shaper_profile_id``: Shaper profile ID of the private shaper to be used by
2598  the node.
2599* ``cman_mode``: Congestion management mode to be enabled for this node.
2600* ``wred_profile_id``: WRED profile id to be enabled for this node.
2601* ``stats_mask``: Mask of statistics counter types to be enabled for this node.
2602* ``n_shared_shapers``: Number of shared shapers.
2603* ``shared_shaper_id``: Shared shaper id.
2604
2605Delete port traffic management hierarchy node
2606~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
2607
2608Delete node from port traffic management hiearchy::
2609
2610   testpmd> del port tm node (port_id) (node_id)
2611
2612Update port traffic management hierarchy parent node
2613~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
2614
2615Update port traffic management hierarchy parent node::
2616
2617   testpmd> set port tm node parent (port_id) (node_id) (parent_node_id) \
2618   (priority) (weight)
2619
2620This function can only be called after the hierarchy commit invocation. Its
2621success depends on the port support for this operation, as advertised through
2622the port capability set. This function is valid for all nodes of the traffic
2623management hierarchy except root node.
2624
2625Suspend port traffic management hierarchy node
2626~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
2627
2628   testpmd> suspend port tm node (port_id) (node_id)
2629
2630Resume port traffic management hierarchy node
2631~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
2632
2633   testpmd> resume port tm node (port_id) (node_id)
2634
2635Commit port traffic management hierarchy
2636~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
2637
2638Commit the traffic management hierarchy on the port::
2639
2640   testpmd> port tm hierarchy commit (port_id) (clean_on_fail)
2641
2642where:
2643
2644* ``clean_on_fail``: When set to non-zero, hierarchy is cleared on function
2645  call failure. On the other hand, hierarchy is preserved when this parameter
2646  is equal to zero.
2647
2648Set port traffic management default hierarchy (tm forwarding mode)
2649~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
2650
2651set the traffic management default hierarchy on the port::
2652
2653   testpmd> set port tm hierarchy default (port_id)
2654
2655Filter Functions
2656----------------
2657
2658This section details the available filter functions that are available.
2659
2660Note these functions interface the deprecated legacy filtering framework,
2661superseded by *rte_flow*. See `Flow rules management`_.
2662
2663ethertype_filter
2664~~~~~~~~~~~~~~~~~~~~
2665
2666Add or delete a L2 Ethertype filter, which identify packets by their L2 Ethertype mainly assign them to a receive queue::
2667
2668   ethertype_filter (port_id) (add|del) (mac_addr|mac_ignr) (mac_address) \
2669                    ethertype (ether_type) (drop|fwd) queue (queue_id)
2670
2671The available information parameters are:
2672
2673* ``port_id``: The port which the Ethertype filter assigned on.
2674
2675* ``mac_addr``: Compare destination mac address.
2676
2677* ``mac_ignr``: Ignore destination mac address match.
2678
2679* ``mac_address``: Destination mac address to match.
2680
2681* ``ether_type``: The EtherType value want to match,
2682  for example 0x0806 for ARP packet. 0x0800 (IPv4) and 0x86DD (IPv6) are invalid.
2683
2684* ``queue_id``: The receive queue associated with this EtherType filter.
2685  It is meaningless when deleting or dropping.
2686
2687Example, to add/remove an ethertype filter rule::
2688
2689   testpmd> ethertype_filter 0 add mac_ignr 00:11:22:33:44:55 \
2690                             ethertype 0x0806 fwd queue 3
2691
2692   testpmd> ethertype_filter 0 del mac_ignr 00:11:22:33:44:55 \
2693                             ethertype 0x0806 fwd queue 3
2694
26952tuple_filter
2696~~~~~~~~~~~~~~~~~
2697
2698Add or delete a 2-tuple filter,
2699which identifies packets by specific protocol and destination TCP/UDP port
2700and forwards packets into one of the receive queues::
2701
2702   2tuple_filter (port_id) (add|del) dst_port (dst_port_value) \
2703                 protocol (protocol_value) mask (mask_value) \
2704                 tcp_flags (tcp_flags_value) priority (prio_value) \
2705                 queue (queue_id)
2706
2707The available information parameters are:
2708
2709* ``port_id``: The port which the 2-tuple filter assigned on.
2710
2711* ``dst_port_value``: Destination port in L4.
2712
2713* ``protocol_value``: IP L4 protocol.
2714
2715* ``mask_value``: Participates in the match or not by bit for field above, 1b means participate.
2716
2717* ``tcp_flags_value``: TCP control bits. The non-zero value is invalid, when the pro_value is not set to 0x06 (TCP).
2718
2719* ``prio_value``: Priority of this filter.
2720
2721* ``queue_id``: The receive queue associated with this 2-tuple filter.
2722
2723Example, to add/remove an 2tuple filter rule::
2724
2725   testpmd> 2tuple_filter 0 add dst_port 32 protocol 0x06 mask 0x03 \
2726                          tcp_flags 0x02 priority 3 queue 3
2727
2728   testpmd> 2tuple_filter 0 del dst_port 32 protocol 0x06 mask 0x03 \
2729                          tcp_flags 0x02 priority 3 queue 3
2730
27315tuple_filter
2732~~~~~~~~~~~~~~~~~
2733
2734Add or delete a 5-tuple filter,
2735which consists of a 5-tuple (protocol, source and destination IP addresses, source and destination TCP/UDP/SCTP port)
2736and routes packets into one of the receive queues::
2737
2738   5tuple_filter (port_id) (add|del) dst_ip (dst_address) src_ip \
2739                 (src_address) dst_port (dst_port_value) \
2740                 src_port (src_port_value) protocol (protocol_value) \
2741                 mask (mask_value) tcp_flags (tcp_flags_value) \
2742                 priority (prio_value) queue (queue_id)
2743
2744The available information parameters are:
2745
2746* ``port_id``: The port which the 5-tuple filter assigned on.
2747
2748* ``dst_address``: Destination IP address.
2749
2750* ``src_address``: Source IP address.
2751
2752* ``dst_port_value``: TCP/UDP destination port.
2753
2754* ``src_port_value``: TCP/UDP source port.
2755
2756* ``protocol_value``: L4 protocol.
2757
2758* ``mask_value``: Participates in the match or not by bit for field above, 1b means participate
2759
2760* ``tcp_flags_value``: TCP control bits. The non-zero value is invalid, when the protocol_value is not set to 0x06 (TCP).
2761
2762* ``prio_value``: The priority of this filter.
2763
2764* ``queue_id``: The receive queue associated with this 5-tuple filter.
2765
2766Example, to add/remove an 5tuple filter rule::
2767
2768   testpmd> 5tuple_filter 0 add dst_ip 2.2.2.5 src_ip 2.2.2.4 \
2769            dst_port 64 src_port 32 protocol 0x06 mask 0x1F \
2770            flags 0x0 priority 3 queue 3
2771
2772   testpmd> 5tuple_filter 0 del dst_ip 2.2.2.5 src_ip 2.2.2.4 \
2773            dst_port 64 src_port 32 protocol 0x06 mask 0x1F \
2774            flags 0x0 priority 3 queue 3
2775
2776syn_filter
2777~~~~~~~~~~
2778
2779Using the  SYN filter, TCP packets whose *SYN* flag is set can be forwarded to a separate queue::
2780
2781   syn_filter (port_id) (add|del) priority (high|low) queue (queue_id)
2782
2783The available information parameters are:
2784
2785* ``port_id``: The port which the SYN filter assigned on.
2786
2787* ``high``: This SYN filter has higher priority than other filters.
2788
2789* ``low``: This SYN filter has lower priority than other filters.
2790
2791* ``queue_id``: The receive queue associated with this SYN filter
2792
2793Example::
2794
2795   testpmd> syn_filter 0 add priority high queue 3
2796
2797flex_filter
2798~~~~~~~~~~~
2799
2800With flex filter, packets can be recognized by any arbitrary pattern within the first 128 bytes of the packet
2801and routed into one of the receive queues::
2802
2803   flex_filter (port_id) (add|del) len (len_value) bytes (bytes_value) \
2804               mask (mask_value) priority (prio_value) queue (queue_id)
2805
2806The available information parameters are:
2807
2808* ``port_id``: The port which the Flex filter is assigned on.
2809
2810* ``len_value``: Filter length in bytes, no greater than 128.
2811
2812* ``bytes_value``: A string in hexadecimal, means the value the flex filter needs to match.
2813
2814* ``mask_value``: A string in hexadecimal, bit 1 means corresponding byte participates in the match.
2815
2816* ``prio_value``: The priority of this filter.
2817
2818* ``queue_id``: The receive queue associated with this Flex filter.
2819
2820Example::
2821
2822   testpmd> flex_filter 0 add len 16 bytes 0x00000000000000000000000008060000 \
2823                          mask 000C priority 3 queue 3
2824
2825   testpmd> flex_filter 0 del len 16 bytes 0x00000000000000000000000008060000 \
2826                          mask 000C priority 3 queue 3
2827
2828
2829.. _testpmd_flow_director:
2830
2831flow_director_filter
2832~~~~~~~~~~~~~~~~~~~~
2833
2834The Flow Director works in receive mode to identify specific flows or sets of flows and route them to specific queues.
2835
2836Four types of filtering are supported which are referred to as Perfect Match, Signature, Perfect-mac-vlan and
2837Perfect-tunnel filters, the match mode is set by the ``--pkt-filter-mode`` command-line parameter:
2838
2839* Perfect match filters.
2840  The hardware checks a match between the masked fields of the received packets and the programmed filters.
2841  The masked fields are for IP flow.
2842
2843* Signature filters.
2844  The hardware checks a match between a hash-based signature of the masked fields of the received packet.
2845
2846* Perfect-mac-vlan match filters.
2847  The hardware checks a match between the masked fields of the received packets and the programmed filters.
2848  The masked fields are for MAC VLAN flow.
2849
2850* Perfect-tunnel match filters.
2851  The hardware checks a match between the masked fields of the received packets and the programmed filters.
2852  The masked fields are for tunnel flow.
2853
2854* Perfect-raw-flow-type match filters.
2855  The hardware checks a match between the masked fields of the received packets and pre-loaded raw (template) packet.
2856  The masked fields are specified by input sets.
2857
2858The Flow Director filters can match the different fields for different type of packet: flow type, specific input set
2859per flow type and the flexible payload.
2860
2861The Flow Director can also mask out parts of all of these fields so that filters
2862are only applied to certain fields or parts of the fields.
2863
2864Note that for raw flow type mode the source and destination fields in the
2865raw packet buffer need to be presented in a reversed order with respect
2866to the expected received packets.
2867For example: IP source and destination addresses or TCP/UDP/SCTP
2868source and destination ports
2869
2870Different NICs may have different capabilities, command show port fdir (port_id) can be used to acquire the information.
2871
2872# Commands to add flow director filters of different flow types::
2873
2874   flow_director_filter (port_id) mode IP (add|del|update) \
2875                        flow (ipv4-other|ipv4-frag|ipv6-other|ipv6-frag) \
2876                        src (src_ip_address) dst (dst_ip_address) \
2877                        tos (tos_value) proto (proto_value) ttl (ttl_value) \
2878                        vlan (vlan_value) flexbytes (flexbytes_value) \
2879                        (drop|fwd) pf|vf(vf_id) queue (queue_id) \
2880                        fd_id (fd_id_value)
2881
2882   flow_director_filter (port_id) mode IP (add|del|update) \
2883                        flow (ipv4-tcp|ipv4-udp|ipv6-tcp|ipv6-udp) \
2884                        src (src_ip_address) (src_port) \
2885                        dst (dst_ip_address) (dst_port) \
2886                        tos (tos_value) ttl (ttl_value) \
2887                        vlan (vlan_value) flexbytes (flexbytes_value) \
2888                        (drop|fwd) queue pf|vf(vf_id) (queue_id) \
2889                        fd_id (fd_id_value)
2890
2891   flow_director_filter (port_id) mode IP (add|del|update) \
2892                        flow (ipv4-sctp|ipv6-sctp) \
2893                        src (src_ip_address) (src_port) \
2894                        dst (dst_ip_address) (dst_port) \
2895                        tos (tos_value) ttl (ttl_value) \
2896                        tag (verification_tag) vlan (vlan_value) \
2897                        flexbytes (flexbytes_value) (drop|fwd) \
2898                        pf|vf(vf_id) queue (queue_id) fd_id (fd_id_value)
2899
2900   flow_director_filter (port_id) mode IP (add|del|update) flow l2_payload \
2901                        ether (ethertype) flexbytes (flexbytes_value) \
2902                        (drop|fwd) pf|vf(vf_id) queue (queue_id)
2903                        fd_id (fd_id_value)
2904
2905   flow_director_filter (port_id) mode MAC-VLAN (add|del|update) \
2906                        mac (mac_address) vlan (vlan_value) \
2907                        flexbytes (flexbytes_value) (drop|fwd) \
2908                        queue (queue_id) fd_id (fd_id_value)
2909
2910   flow_director_filter (port_id) mode Tunnel (add|del|update) \
2911                        mac (mac_address) vlan (vlan_value) \
2912                        tunnel (NVGRE|VxLAN) tunnel-id (tunnel_id_value) \
2913                        flexbytes (flexbytes_value) (drop|fwd) \
2914                        queue (queue_id) fd_id (fd_id_value)
2915
2916   flow_director_filter (port_id) mode raw (add|del|update) flow (flow_id) \
2917                        (drop|fwd) queue (queue_id) fd_id (fd_id_value) \
2918                        packet (packet file name)
2919
2920For example, to add an ipv4-udp flow type filter::
2921
2922   testpmd> flow_director_filter 0 mode IP add flow ipv4-udp src 2.2.2.3 32 \
2923            dst 2.2.2.5 33 tos 2 ttl 40 vlan 0x1 flexbytes (0x88,0x48) \
2924            fwd pf queue 1 fd_id 1
2925
2926For example, add an ipv4-other flow type filter::
2927
2928   testpmd> flow_director_filter 0 mode IP add flow ipv4-other src 2.2.2.3 \
2929             dst 2.2.2.5 tos 2 proto 20 ttl 40 vlan 0x1 \
2930             flexbytes (0x88,0x48) fwd pf queue 1 fd_id 1
2931
2932flush_flow_director
2933~~~~~~~~~~~~~~~~~~~
2934
2935Flush all flow director filters on a device::
2936
2937   testpmd> flush_flow_director (port_id)
2938
2939Example, to flush all flow director filter on port 0::
2940
2941   testpmd> flush_flow_director 0
2942
2943flow_director_mask
2944~~~~~~~~~~~~~~~~~~
2945
2946Set flow director's input masks::
2947
2948   flow_director_mask (port_id) mode IP vlan (vlan_value) \
2949                      src_mask (ipv4_src) (ipv6_src) (src_port) \
2950                      dst_mask (ipv4_dst) (ipv6_dst) (dst_port)
2951
2952   flow_director_mask (port_id) mode MAC-VLAN vlan (vlan_value)
2953
2954   flow_director_mask (port_id) mode Tunnel vlan (vlan_value) \
2955                      mac (mac_value) tunnel-type (tunnel_type_value) \
2956                      tunnel-id (tunnel_id_value)
2957
2958Example, to set flow director mask on port 0::
2959
2960   testpmd> flow_director_mask 0 mode IP vlan 0xefff \
2961            src_mask 255.255.255.255 \
2962                FFFF:FFFF:FFFF:FFFF:FFFF:FFFF:FFFF:FFFF 0xFFFF \
2963            dst_mask 255.255.255.255 \
2964                FFFF:FFFF:FFFF:FFFF:FFFF:FFFF:FFFF:FFFF 0xFFFF
2965
2966flow_director_flex_mask
2967~~~~~~~~~~~~~~~~~~~~~~~
2968
2969set masks of flow director's flexible payload based on certain flow type::
2970
2971   testpmd> flow_director_flex_mask (port_id) \
2972            flow (none|ipv4-other|ipv4-frag|ipv4-tcp|ipv4-udp|ipv4-sctp| \
2973                  ipv6-other|ipv6-frag|ipv6-tcp|ipv6-udp|ipv6-sctp| \
2974                  l2_payload|all) (mask)
2975
2976Example, to set flow director's flex mask for all flow type on port 0::
2977
2978   testpmd> flow_director_flex_mask 0 flow all \
2979            (0xff,0xff,0,0,0,0,0,0,0,0,0,0,0,0,0,0)
2980
2981
2982flow_director_flex_payload
2983~~~~~~~~~~~~~~~~~~~~~~~~~~
2984
2985Configure flexible payload selection::
2986
2987   flow_director_flex_payload (port_id) (raw|l2|l3|l4) (config)
2988
2989For example, to select the first 16 bytes from the offset 4 (bytes) of packet's payload as flexible payload::
2990
2991   testpmd> flow_director_flex_payload 0 l4 \
2992            (4,5,6,7,8,9,10,11,12,13,14,15,16,17,18,19)
2993
2994get_sym_hash_ena_per_port
2995~~~~~~~~~~~~~~~~~~~~~~~~~
2996
2997Get symmetric hash enable configuration per port::
2998
2999   get_sym_hash_ena_per_port (port_id)
3000
3001For example, to get symmetric hash enable configuration of port 1::
3002
3003   testpmd> get_sym_hash_ena_per_port 1
3004
3005set_sym_hash_ena_per_port
3006~~~~~~~~~~~~~~~~~~~~~~~~~
3007
3008Set symmetric hash enable configuration per port to enable or disable::
3009
3010   set_sym_hash_ena_per_port (port_id) (enable|disable)
3011
3012For example, to set symmetric hash enable configuration of port 1 to enable::
3013
3014   testpmd> set_sym_hash_ena_per_port 1 enable
3015
3016get_hash_global_config
3017~~~~~~~~~~~~~~~~~~~~~~
3018
3019Get the global configurations of hash filters::
3020
3021   get_hash_global_config (port_id)
3022
3023For example, to get the global configurations of hash filters of port 1::
3024
3025   testpmd> get_hash_global_config 1
3026
3027set_hash_global_config
3028~~~~~~~~~~~~~~~~~~~~~~
3029
3030Set the global configurations of hash filters::
3031
3032   set_hash_global_config (port_id) (toeplitz|simple_xor|default) \
3033   (ipv4|ipv4-frag|ipv4-tcp|ipv4-udp|ipv4-sctp|ipv4-other|ipv6|ipv6-frag| \
3034   ipv6-tcp|ipv6-udp|ipv6-sctp|ipv6-other|l2_payload|<flow_id>) \
3035   (enable|disable)
3036
3037For example, to enable simple_xor for flow type of ipv6 on port 2::
3038
3039   testpmd> set_hash_global_config 2 simple_xor ipv6 enable
3040
3041set_hash_input_set
3042~~~~~~~~~~~~~~~~~~
3043
3044Set the input set for hash::
3045
3046   set_hash_input_set (port_id) (ipv4-frag|ipv4-tcp|ipv4-udp|ipv4-sctp| \
3047   ipv4-other|ipv6-frag|ipv6-tcp|ipv6-udp|ipv6-sctp|ipv6-other| \
3048   l2_payload|<flow_id>) (ovlan|ivlan|src-ipv4|dst-ipv4|src-ipv6|dst-ipv6| \
3049   ipv4-tos|ipv4-proto|ipv6-tc|ipv6-next-header|udp-src-port|udp-dst-port| \
3050   tcp-src-port|tcp-dst-port|sctp-src-port|sctp-dst-port|sctp-veri-tag| \
3051   udp-key|gre-key|fld-1st|fld-2nd|fld-3rd|fld-4th|fld-5th|fld-6th|fld-7th| \
3052   fld-8th|none) (select|add)
3053
3054For example, to add source IP to hash input set for flow type of ipv4-udp on port 0::
3055
3056   testpmd> set_hash_input_set 0 ipv4-udp src-ipv4 add
3057
3058set_fdir_input_set
3059~~~~~~~~~~~~~~~~~~
3060
3061The Flow Director filters can match the different fields for different type of packet, i.e. specific input set
3062on per flow type and the flexible payload. This command can be used to change input set for each flow type.
3063
3064Set the input set for flow director::
3065
3066   set_fdir_input_set (port_id) (ipv4-frag|ipv4-tcp|ipv4-udp|ipv4-sctp| \
3067   ipv4-other|ipv6|ipv6-frag|ipv6-tcp|ipv6-udp|ipv6-sctp|ipv6-other| \
3068   l2_payload|<flow_id>) (ivlan|ethertype|src-ipv4|dst-ipv4|src-ipv6|dst-ipv6| \
3069   ipv4-tos|ipv4-proto|ipv4-ttl|ipv6-tc|ipv6-next-header|ipv6-hop-limits| \
3070   tudp-src-port|udp-dst-port|cp-src-port|tcp-dst-port|sctp-src-port| \
3071   sctp-dst-port|sctp-veri-tag|none) (select|add)
3072
3073For example to add source IP to FD input set for flow type of ipv4-udp on port 0::
3074
3075   testpmd> set_fdir_input_set 0 ipv4-udp src-ipv4 add
3076
3077global_config
3078~~~~~~~~~~~~~
3079
3080Set different GRE key length for input set::
3081
3082   global_config (port_id) gre-key-len (number in bytes)
3083
3084For example to set GRE key length for input set to 4 bytes on port 0::
3085
3086   testpmd> global_config 0 gre-key-len 4
3087
3088
3089.. _testpmd_rte_flow:
3090
3091Flow rules management
3092---------------------
3093
3094Control of the generic flow API (*rte_flow*) is fully exposed through the
3095``flow`` command (validation, creation, destruction, queries and operation
3096modes).
3097
3098Considering *rte_flow* overlaps with all `Filter Functions`_, using both
3099features simultaneously may cause undefined side-effects and is therefore
3100not recommended.
3101
3102``flow`` syntax
3103~~~~~~~~~~~~~~~
3104
3105Because the ``flow`` command uses dynamic tokens to handle the large number
3106of possible flow rules combinations, its behavior differs slightly from
3107other commands, in particular:
3108
3109- Pressing *?* or the *<tab>* key displays contextual help for the current
3110  token, not that of the entire command.
3111
3112- Optional and repeated parameters are supported (provided they are listed
3113  in the contextual help).
3114
3115The first parameter stands for the operation mode. Possible operations and
3116their general syntax are described below. They are covered in detail in the
3117following sections.
3118
3119- Check whether a flow rule can be created::
3120
3121   flow validate {port_id}
3122       [group {group_id}] [priority {level}] [ingress] [egress] [transfer]
3123       pattern {item} [/ {item} [...]] / end
3124       actions {action} [/ {action} [...]] / end
3125
3126- Create a flow rule::
3127
3128   flow create {port_id}
3129       [group {group_id}] [priority {level}] [ingress] [egress] [transfer]
3130       pattern {item} [/ {item} [...]] / end
3131       actions {action} [/ {action} [...]] / end
3132
3133- Destroy specific flow rules::
3134
3135   flow destroy {port_id} rule {rule_id} [...]
3136
3137- Destroy all flow rules::
3138
3139   flow flush {port_id}
3140
3141- Query an existing flow rule::
3142
3143   flow query {port_id} {rule_id} {action}
3144
3145- List existing flow rules sorted by priority, filtered by group
3146  identifiers::
3147
3148   flow list {port_id} [group {group_id}] [...]
3149
3150- Restrict ingress traffic to the defined flow rules::
3151
3152   flow isolate {port_id} {boolean}
3153
3154Validating flow rules
3155~~~~~~~~~~~~~~~~~~~~~
3156
3157``flow validate`` reports whether a flow rule would be accepted by the
3158underlying device in its current state but stops short of creating it. It is
3159bound to ``rte_flow_validate()``::
3160
3161   flow validate {port_id}
3162      [group {group_id}] [priority {level}] [ingress] [egress] [transfer]
3163      pattern {item} [/ {item} [...]] / end
3164      actions {action} [/ {action} [...]] / end
3165
3166If successful, it will show::
3167
3168   Flow rule validated
3169
3170Otherwise it will show an error message of the form::
3171
3172   Caught error type [...] ([...]): [...]
3173
3174This command uses the same parameters as ``flow create``, their format is
3175described in `Creating flow rules`_.
3176
3177Check whether redirecting any Ethernet packet received on port 0 to RX queue
3178index 6 is supported::
3179
3180   testpmd> flow validate 0 ingress pattern eth / end
3181      actions queue index 6 / end
3182   Flow rule validated
3183   testpmd>
3184
3185Port 0 does not support TCPv6 rules::
3186
3187   testpmd> flow validate 0 ingress pattern eth / ipv6 / tcp / end
3188      actions drop / end
3189   Caught error type 9 (specific pattern item): Invalid argument
3190   testpmd>
3191
3192Creating flow rules
3193~~~~~~~~~~~~~~~~~~~
3194
3195``flow create`` validates and creates the specified flow rule. It is bound
3196to ``rte_flow_create()``::
3197
3198   flow create {port_id}
3199      [group {group_id}] [priority {level}] [ingress] [egress] [transfer]
3200      pattern {item} [/ {item} [...]] / end
3201      actions {action} [/ {action} [...]] / end
3202
3203If successful, it will return a flow rule ID usable with other commands::
3204
3205   Flow rule #[...] created
3206
3207Otherwise it will show an error message of the form::
3208
3209   Caught error type [...] ([...]): [...]
3210
3211Parameters describe in the following order:
3212
3213- Attributes (*group*, *priority*, *ingress*, *egress*, *transfer* tokens).
3214- A matching pattern, starting with the *pattern* token and terminated by an
3215  *end* pattern item.
3216- Actions, starting with the *actions* token and terminated by an *end*
3217  action.
3218
3219These translate directly to *rte_flow* objects provided as-is to the
3220underlying functions.
3221
3222The shortest valid definition only comprises mandatory tokens::
3223
3224   testpmd> flow create 0 pattern end actions end
3225
3226Note that PMDs may refuse rules that essentially do nothing such as this
3227one.
3228
3229**All unspecified object values are automatically initialized to 0.**
3230
3231Attributes
3232^^^^^^^^^^
3233
3234These tokens affect flow rule attributes (``struct rte_flow_attr``) and are
3235specified before the ``pattern`` token.
3236
3237- ``group {group id}``: priority group.
3238- ``priority {level}``: priority level within group.
3239- ``ingress``: rule applies to ingress traffic.
3240- ``egress``: rule applies to egress traffic.
3241- ``transfer``: apply rule directly to endpoints found in pattern.
3242
3243Each instance of an attribute specified several times overrides the previous
3244value as shown below (group 4 is used)::
3245
3246   testpmd> flow create 0 group 42 group 24 group 4 [...]
3247
3248Note that once enabled, ``ingress`` and ``egress`` cannot be disabled.
3249
3250While not specifying a direction is an error, some rules may allow both
3251simultaneously.
3252
3253Most rules affect RX therefore contain the ``ingress`` token::
3254
3255   testpmd> flow create 0 ingress pattern [...]
3256
3257Matching pattern
3258^^^^^^^^^^^^^^^^
3259
3260A matching pattern starts after the ``pattern`` token. It is made of pattern
3261items and is terminated by a mandatory ``end`` item.
3262
3263Items are named after their type (*RTE_FLOW_ITEM_TYPE_* from ``enum
3264rte_flow_item_type``).
3265
3266The ``/`` token is used as a separator between pattern items as shown
3267below::
3268
3269   testpmd> flow create 0 ingress pattern eth / ipv4 / udp / end [...]
3270
3271Note that protocol items like these must be stacked from lowest to highest
3272layer to make sense. For instance, the following rule is either invalid or
3273unlikely to match any packet::
3274
3275   testpmd> flow create 0 ingress pattern eth / udp / ipv4 / end [...]
3276
3277More information on these restrictions can be found in the *rte_flow*
3278documentation.
3279
3280Several items support additional specification structures, for example
3281``ipv4`` allows specifying source and destination addresses as follows::
3282
3283   testpmd> flow create 0 ingress pattern eth / ipv4 src is 10.1.1.1
3284      dst is 10.2.0.0 / end [...]
3285
3286This rule matches all IPv4 traffic with the specified properties.
3287
3288In this example, ``src`` and ``dst`` are field names of the underlying
3289``struct rte_flow_item_ipv4`` object. All item properties can be specified
3290in a similar fashion.
3291
3292The ``is`` token means that the subsequent value must be matched exactly,
3293and assigns ``spec`` and ``mask`` fields in ``struct rte_flow_item``
3294accordingly. Possible assignment tokens are:
3295
3296- ``is``: match value perfectly (with full bit-mask).
3297- ``spec``: match value according to configured bit-mask.
3298- ``last``: specify upper bound to establish a range.
3299- ``mask``: specify bit-mask with relevant bits set to one.
3300- ``prefix``: generate bit-mask from a prefix length.
3301
3302These yield identical results::
3303
3304   ipv4 src is 10.1.1.1
3305
3306::
3307
3308   ipv4 src spec 10.1.1.1 src mask 255.255.255.255
3309
3310::
3311
3312   ipv4 src spec 10.1.1.1 src prefix 32
3313
3314::
3315
3316   ipv4 src is 10.1.1.1 src last 10.1.1.1 # range with a single value
3317
3318::
3319
3320   ipv4 src is 10.1.1.1 src last 0 # 0 disables range
3321
3322Inclusive ranges can be defined with ``last``::
3323
3324   ipv4 src is 10.1.1.1 src last 10.2.3.4 # 10.1.1.1 to 10.2.3.4
3325
3326Note that ``mask`` affects both ``spec`` and ``last``::
3327
3328   ipv4 src is 10.1.1.1 src last 10.2.3.4 src mask 255.255.0.0
3329      # matches 10.1.0.0 to 10.2.255.255
3330
3331Properties can be modified multiple times::
3332
3333   ipv4 src is 10.1.1.1 src is 10.1.2.3 src is 10.2.3.4 # matches 10.2.3.4
3334
3335::
3336
3337   ipv4 src is 10.1.1.1 src prefix 24 src prefix 16 # matches 10.1.0.0/16
3338
3339Pattern items
3340^^^^^^^^^^^^^
3341
3342This section lists supported pattern items and their attributes, if any.
3343
3344- ``end``: end list of pattern items.
3345
3346- ``void``: no-op pattern item.
3347
3348- ``invert``: perform actions when pattern does not match.
3349
3350- ``any``: match any protocol for the current layer.
3351
3352  - ``num {unsigned}``: number of layers covered.
3353
3354- ``pf``: match traffic from/to the physical function.
3355
3356- ``vf``: match traffic from/to a virtual function ID.
3357
3358  - ``id {unsigned}``: VF ID.
3359
3360- ``phy_port``: match traffic from/to a specific physical port.
3361
3362  - ``index {unsigned}``: physical port index.
3363
3364- ``port_id``: match traffic from/to a given DPDK port ID.
3365
3366  - ``id {unsigned}``: DPDK port ID.
3367
3368- ``mark``: match value set in previously matched flow rule using the mark action.
3369
3370  - ``id {unsigned}``: arbitrary integer value.
3371
3372- ``raw``: match an arbitrary byte string.
3373
3374  - ``relative {boolean}``: look for pattern after the previous item.
3375  - ``search {boolean}``: search pattern from offset (see also limit).
3376  - ``offset {integer}``: absolute or relative offset for pattern.
3377  - ``limit {unsigned}``: search area limit for start of pattern.
3378  - ``pattern {string}``: byte string to look for.
3379
3380- ``eth``: match Ethernet header.
3381
3382  - ``dst {MAC-48}``: destination MAC.
3383  - ``src {MAC-48}``: source MAC.
3384  - ``type {unsigned}``: EtherType or TPID.
3385
3386- ``vlan``: match 802.1Q/ad VLAN tag.
3387
3388  - ``tci {unsigned}``: tag control information.
3389  - ``pcp {unsigned}``: priority code point.
3390  - ``dei {unsigned}``: drop eligible indicator.
3391  - ``vid {unsigned}``: VLAN identifier.
3392  - ``inner_type {unsigned}``: inner EtherType or TPID.
3393
3394- ``ipv4``: match IPv4 header.
3395
3396  - ``tos {unsigned}``: type of service.
3397  - ``ttl {unsigned}``: time to live.
3398  - ``proto {unsigned}``: next protocol ID.
3399  - ``src {ipv4 address}``: source address.
3400  - ``dst {ipv4 address}``: destination address.
3401
3402- ``ipv6``: match IPv6 header.
3403
3404  - ``tc {unsigned}``: traffic class.
3405  - ``flow {unsigned}``: flow label.
3406  - ``proto {unsigned}``: protocol (next header).
3407  - ``hop {unsigned}``: hop limit.
3408  - ``src {ipv6 address}``: source address.
3409  - ``dst {ipv6 address}``: destination address.
3410
3411- ``icmp``: match ICMP header.
3412
3413  - ``type {unsigned}``: ICMP packet type.
3414  - ``code {unsigned}``: ICMP packet code.
3415
3416- ``udp``: match UDP header.
3417
3418  - ``src {unsigned}``: UDP source port.
3419  - ``dst {unsigned}``: UDP destination port.
3420
3421- ``tcp``: match TCP header.
3422
3423  - ``src {unsigned}``: TCP source port.
3424  - ``dst {unsigned}``: TCP destination port.
3425
3426- ``sctp``: match SCTP header.
3427
3428  - ``src {unsigned}``: SCTP source port.
3429  - ``dst {unsigned}``: SCTP destination port.
3430  - ``tag {unsigned}``: validation tag.
3431  - ``cksum {unsigned}``: checksum.
3432
3433- ``vxlan``: match VXLAN header.
3434
3435  - ``vni {unsigned}``: VXLAN identifier.
3436
3437- ``e_tag``: match IEEE 802.1BR E-Tag header.
3438
3439  - ``grp_ecid_b {unsigned}``: GRP and E-CID base.
3440
3441- ``nvgre``: match NVGRE header.
3442
3443  - ``tni {unsigned}``: virtual subnet ID.
3444
3445- ``mpls``: match MPLS header.
3446
3447  - ``label {unsigned}``: MPLS label.
3448
3449- ``gre``: match GRE header.
3450
3451  - ``protocol {unsigned}``: protocol type.
3452
3453- ``fuzzy``: fuzzy pattern match, expect faster than default.
3454
3455  - ``thresh {unsigned}``: accuracy threshold.
3456
3457- ``gtp``, ``gtpc``, ``gtpu``: match GTPv1 header.
3458
3459  - ``teid {unsigned}``: tunnel endpoint identifier.
3460
3461- ``geneve``: match GENEVE header.
3462
3463  - ``vni {unsigned}``: virtual network identifier.
3464  - ``protocol {unsigned}``: protocol type.
3465
3466- ``vxlan-gpe``: match VXLAN-GPE header.
3467
3468  - ``vni {unsigned}``: VXLAN-GPE identifier.
3469
3470- ``arp_eth_ipv4``: match ARP header for Ethernet/IPv4.
3471
3472  - ``sha {MAC-48}``: sender hardware address.
3473  - ``spa {ipv4 address}``: sender IPv4 address.
3474  - ``tha {MAC-48}``: target hardware address.
3475  - ``tpa {ipv4 address}``: target IPv4 address.
3476
3477- ``ipv6_ext``: match presence of any IPv6 extension header.
3478
3479  - ``next_hdr {unsigned}``: next header.
3480
3481- ``icmp6``: match any ICMPv6 header.
3482
3483  - ``type {unsigned}``: ICMPv6 type.
3484  - ``code {unsigned}``: ICMPv6 code.
3485
3486- ``icmp6_nd_ns``: match ICMPv6 neighbor discovery solicitation.
3487
3488  - ``target_addr {ipv6 address}``: target address.
3489
3490- ``icmp6_nd_na``: match ICMPv6 neighbor discovery advertisement.
3491
3492  - ``target_addr {ipv6 address}``: target address.
3493
3494- ``icmp6_nd_opt``: match presence of any ICMPv6 neighbor discovery option.
3495
3496  - ``type {unsigned}``: ND option type.
3497
3498- ``icmp6_nd_opt_sla_eth``: match ICMPv6 neighbor discovery source Ethernet
3499  link-layer address option.
3500
3501  - ``sla {MAC-48}``: source Ethernet LLA.
3502
3503- ``icmp6_nd_opt_sla_eth``: match ICMPv6 neighbor discovery target Ethernet
3504  link-layer address option.
3505
3506  - ``tla {MAC-48}``: target Ethernet LLA.
3507
3508Actions list
3509^^^^^^^^^^^^
3510
3511A list of actions starts after the ``actions`` token in the same fashion as
3512`Matching pattern`_; actions are separated by ``/`` tokens and the list is
3513terminated by a mandatory ``end`` action.
3514
3515Actions are named after their type (*RTE_FLOW_ACTION_TYPE_* from ``enum
3516rte_flow_action_type``).
3517
3518Dropping all incoming UDPv4 packets can be expressed as follows::
3519
3520   testpmd> flow create 0 ingress pattern eth / ipv4 / udp / end
3521      actions drop / end
3522
3523Several actions have configurable properties which must be specified when
3524there is no valid default value. For example, ``queue`` requires a target
3525queue index.
3526
3527This rule redirects incoming UDPv4 traffic to queue index 6::
3528
3529   testpmd> flow create 0 ingress pattern eth / ipv4 / udp / end
3530      actions queue index 6 / end
3531
3532While this one could be rejected by PMDs (unspecified queue index)::
3533
3534   testpmd> flow create 0 ingress pattern eth / ipv4 / udp / end
3535      actions queue / end
3536
3537As defined by *rte_flow*, the list is not ordered, all actions of a given
3538rule are performed simultaneously. These are equivalent::
3539
3540   queue index 6 / void / mark id 42 / end
3541
3542::
3543
3544   void / mark id 42 / queue index 6 / end
3545
3546All actions in a list should have different types, otherwise only the last
3547action of a given type is taken into account::
3548
3549   queue index 4 / queue index 5 / queue index 6 / end # will use queue 6
3550
3551::
3552
3553   drop / drop / drop / end # drop is performed only once
3554
3555::
3556
3557   mark id 42 / queue index 3 / mark id 24 / end # mark will be 24
3558
3559Considering they are performed simultaneously, opposite and overlapping
3560actions can sometimes be combined when the end result is unambiguous::
3561
3562   drop / queue index 6 / end # drop has no effect
3563
3564::
3565
3566   queue index 6 / rss queues 6 7 8 / end # queue has no effect
3567
3568::
3569
3570   drop / passthru / end # drop has no effect
3571
3572Note that PMDs may still refuse such combinations.
3573
3574Actions
3575^^^^^^^
3576
3577This section lists supported actions and their attributes, if any.
3578
3579- ``end``: end list of actions.
3580
3581- ``void``: no-op action.
3582
3583- ``passthru``: let subsequent rule process matched packets.
3584
3585- ``jump``: redirect traffic to group on device.
3586
3587  - ``group {unsigned}``: group to redirect to.
3588
3589- ``mark``: attach 32 bit value to packets.
3590
3591  - ``id {unsigned}``: 32 bit value to return with packets.
3592
3593- ``flag``: flag packets.
3594
3595- ``queue``: assign packets to a given queue index.
3596
3597  - ``index {unsigned}``: queue index to use.
3598
3599- ``drop``: drop packets (note: passthru has priority).
3600
3601- ``count``: enable counters for this rule.
3602
3603- ``rss``: spread packets among several queues.
3604
3605  - ``func {hash function}``: RSS hash function to apply, allowed tokens are
3606    the same as `set_hash_global_config`_.
3607
3608  - ``level {unsigned}``: encapsulation level for ``types``.
3609
3610  - ``types [{RSS hash type} [...]] end``: specific RSS hash types, allowed
3611    tokens are the same as `set_hash_input_set`_, except that an empty list
3612    does not disable RSS but instead requests unspecified "best-effort"
3613    settings.
3614
3615  - ``key {string}``: RSS hash key, overrides ``key_len``.
3616
3617  - ``key_len {unsigned}``: RSS hash key length in bytes, can be used in
3618    conjunction with ``key`` to pad or truncate it.
3619
3620  - ``queues [{unsigned} [...]] end``: queue indices to use.
3621
3622- ``pf``: direct traffic to physical function.
3623
3624- ``vf``: direct traffic to a virtual function ID.
3625
3626  - ``original {boolean}``: use original VF ID if possible.
3627  - ``id {unsigned}``: VF ID.
3628
3629- ``phy_port``: direct packets to physical port index.
3630
3631  - ``original {boolean}``: use original port index if possible.
3632  - ``index {unsigned}``: physical port index.
3633
3634- ``port_id``: direct matching traffic to a given DPDK port ID.
3635
3636  - ``original {boolean}``: use original DPDK port ID if possible.
3637  - ``id {unsigned}``: DPDK port ID.
3638
3639- ``of_set_mpls_ttl``: OpenFlow's ``OFPAT_SET_MPLS_TTL``.
3640
3641  - ``mpls_ttl``: MPLS TTL.
3642
3643- ``of_dec_mpls_ttl``: OpenFlow's ``OFPAT_DEC_MPLS_TTL``.
3644
3645- ``of_set_nw_ttl``: OpenFlow's ``OFPAT_SET_NW_TTL``.
3646
3647  - ``nw_ttl``: IP TTL.
3648
3649- ``of_dec_nw_ttl``: OpenFlow's ``OFPAT_DEC_NW_TTL``.
3650
3651- ``of_copy_ttl_out``: OpenFlow's ``OFPAT_COPY_TTL_OUT``.
3652
3653- ``of_copy_ttl_in``: OpenFlow's ``OFPAT_COPY_TTL_IN``.
3654
3655- ``of_pop_vlan``: OpenFlow's ``OFPAT_POP_VLAN``.
3656
3657- ``of_push_vlan``: OpenFlow's ``OFPAT_PUSH_VLAN``.
3658
3659  - ``ethertype``: Ethertype.
3660
3661- ``of_set_vlan_vid``: OpenFlow's ``OFPAT_SET_VLAN_VID``.
3662
3663  - ``vlan_vid``: VLAN id.
3664
3665- ``of_set_vlan_pcp``: OpenFlow's ``OFPAT_SET_VLAN_PCP``.
3666
3667  - ``vlan_pcp``: VLAN priority.
3668
3669- ``of_pop_mpls``: OpenFlow's ``OFPAT_POP_MPLS``.
3670
3671  - ``ethertype``: Ethertype.
3672
3673- ``of_push_mpls``: OpenFlow's ``OFPAT_PUSH_MPLS``.
3674
3675  - ``ethertype``: Ethertype.
3676
3677- ``vxlan_encap``: Performs a VXLAN encapsulation, outer layer configuration
3678  is done through `Config VXLAN Encap outer layers`_.
3679
3680- ``vxlan_decap``: Performs a decapsulation action by stripping all headers of
3681  the VXLAN tunnel network overlay from the matched flow.
3682
3683Destroying flow rules
3684~~~~~~~~~~~~~~~~~~~~~
3685
3686``flow destroy`` destroys one or more rules from their rule ID (as returned
3687by ``flow create``), this command calls ``rte_flow_destroy()`` as many
3688times as necessary::
3689
3690   flow destroy {port_id} rule {rule_id} [...]
3691
3692If successful, it will show::
3693
3694   Flow rule #[...] destroyed
3695
3696It does not report anything for rule IDs that do not exist. The usual error
3697message is shown when a rule cannot be destroyed::
3698
3699   Caught error type [...] ([...]): [...]
3700
3701``flow flush`` destroys all rules on a device and does not take extra
3702arguments. It is bound to ``rte_flow_flush()``::
3703
3704   flow flush {port_id}
3705
3706Any errors are reported as above.
3707
3708Creating several rules and destroying them::
3709
3710   testpmd> flow create 0 ingress pattern eth / ipv6 / end
3711      actions queue index 2 / end
3712   Flow rule #0 created
3713   testpmd> flow create 0 ingress pattern eth / ipv4 / end
3714      actions queue index 3 / end
3715   Flow rule #1 created
3716   testpmd> flow destroy 0 rule 0 rule 1
3717   Flow rule #1 destroyed
3718   Flow rule #0 destroyed
3719   testpmd>
3720
3721The same result can be achieved using ``flow flush``::
3722
3723   testpmd> flow create 0 ingress pattern eth / ipv6 / end
3724      actions queue index 2 / end
3725   Flow rule #0 created
3726   testpmd> flow create 0 ingress pattern eth / ipv4 / end
3727      actions queue index 3 / end
3728   Flow rule #1 created
3729   testpmd> flow flush 0
3730   testpmd>
3731
3732Non-existent rule IDs are ignored::
3733
3734   testpmd> flow create 0 ingress pattern eth / ipv6 / end
3735      actions queue index 2 / end
3736   Flow rule #0 created
3737   testpmd> flow create 0 ingress pattern eth / ipv4 / end
3738      actions queue index 3 / end
3739   Flow rule #1 created
3740   testpmd> flow destroy 0 rule 42 rule 10 rule 2
3741   testpmd>
3742   testpmd> flow destroy 0 rule 0
3743   Flow rule #0 destroyed
3744   testpmd>
3745
3746Querying flow rules
3747~~~~~~~~~~~~~~~~~~~
3748
3749``flow query`` queries a specific action of a flow rule having that
3750ability. Such actions collect information that can be reported using this
3751command. It is bound to ``rte_flow_query()``::
3752
3753   flow query {port_id} {rule_id} {action}
3754
3755If successful, it will display either the retrieved data for known actions
3756or the following message::
3757
3758   Cannot display result for action type [...] ([...])
3759
3760Otherwise, it will complain either that the rule does not exist or that some
3761error occurred::
3762
3763   Flow rule #[...] not found
3764
3765::
3766
3767   Caught error type [...] ([...]): [...]
3768
3769Currently only the ``count`` action is supported. This action reports the
3770number of packets that hit the flow rule and the total number of bytes. Its
3771output has the following format::
3772
3773   count:
3774    hits_set: [...] # whether "hits" contains a valid value
3775    bytes_set: [...] # whether "bytes" contains a valid value
3776    hits: [...] # number of packets
3777    bytes: [...] # number of bytes
3778
3779Querying counters for TCPv6 packets redirected to queue 6::
3780
3781   testpmd> flow create 0 ingress pattern eth / ipv6 / tcp / end
3782      actions queue index 6 / count / end
3783   Flow rule #4 created
3784   testpmd> flow query 0 4 count
3785   count:
3786    hits_set: 1
3787    bytes_set: 0
3788    hits: 386446
3789    bytes: 0
3790   testpmd>
3791
3792Listing flow rules
3793~~~~~~~~~~~~~~~~~~
3794
3795``flow list`` lists existing flow rules sorted by priority and optionally
3796filtered by group identifiers::
3797
3798   flow list {port_id} [group {group_id}] [...]
3799
3800This command only fails with the following message if the device does not
3801exist::
3802
3803   Invalid port [...]
3804
3805Output consists of a header line followed by a short description of each
3806flow rule, one per line. There is no output at all when no flow rules are
3807configured on the device::
3808
3809   ID      Group   Prio    Attr    Rule
3810   [...]   [...]   [...]   [...]   [...]
3811
3812``Attr`` column flags:
3813
3814- ``i`` for ``ingress``.
3815- ``e`` for ``egress``.
3816
3817Creating several flow rules and listing them::
3818
3819   testpmd> flow create 0 ingress pattern eth / ipv4 / end
3820      actions queue index 6 / end
3821   Flow rule #0 created
3822   testpmd> flow create 0 ingress pattern eth / ipv6 / end
3823      actions queue index 2 / end
3824   Flow rule #1 created
3825   testpmd> flow create 0 priority 5 ingress pattern eth / ipv4 / udp / end
3826      actions rss queues 6 7 8 end / end
3827   Flow rule #2 created
3828   testpmd> flow list 0
3829   ID      Group   Prio    Attr    Rule
3830   0       0       0       i-      ETH IPV4 => QUEUE
3831   1       0       0       i-      ETH IPV6 => QUEUE
3832   2       0       5       i-      ETH IPV4 UDP => RSS
3833   testpmd>
3834
3835Rules are sorted by priority (i.e. group ID first, then priority level)::
3836
3837   testpmd> flow list 1
3838   ID      Group   Prio    Attr    Rule
3839   0       0       0       i-      ETH => COUNT
3840   6       0       500     i-      ETH IPV6 TCP => DROP COUNT
3841   5       0       1000    i-      ETH IPV6 ICMP => QUEUE
3842   1       24      0       i-      ETH IPV4 UDP => QUEUE
3843   4       24      10      i-      ETH IPV4 TCP => DROP
3844   3       24      20      i-      ETH IPV4 => DROP
3845   2       24      42      i-      ETH IPV4 UDP => QUEUE
3846   7       63      0       i-      ETH IPV6 UDP VXLAN => MARK QUEUE
3847   testpmd>
3848
3849Output can be limited to specific groups::
3850
3851   testpmd> flow list 1 group 0 group 63
3852   ID      Group   Prio    Attr    Rule
3853   0       0       0       i-      ETH => COUNT
3854   6       0       500     i-      ETH IPV6 TCP => DROP COUNT
3855   5       0       1000    i-      ETH IPV6 ICMP => QUEUE
3856   7       63      0       i-      ETH IPV6 UDP VXLAN => MARK QUEUE
3857   testpmd>
3858
3859Toggling isolated mode
3860~~~~~~~~~~~~~~~~~~~~~~
3861
3862``flow isolate`` can be used to tell the underlying PMD that ingress traffic
3863must only be injected from the defined flow rules; that no default traffic
3864is expected outside those rules and the driver is free to assign more
3865resources to handle them. It is bound to ``rte_flow_isolate()``::
3866
3867 flow isolate {port_id} {boolean}
3868
3869If successful, enabling or disabling isolated mode shows either::
3870
3871 Ingress traffic on port [...]
3872    is now restricted to the defined flow rules
3873
3874Or::
3875
3876 Ingress traffic on port [...]
3877    is not restricted anymore to the defined flow rules
3878
3879Otherwise, in case of error::
3880
3881   Caught error type [...] ([...]): [...]
3882
3883Mainly due to its side effects, PMDs supporting this mode may not have the
3884ability to toggle it more than once without reinitializing affected ports
3885first (e.g. by exiting testpmd).
3886
3887Enabling isolated mode::
3888
3889 testpmd> flow isolate 0 true
3890 Ingress traffic on port 0 is now restricted to the defined flow rules
3891 testpmd>
3892
3893Disabling isolated mode::
3894
3895 testpmd> flow isolate 0 false
3896 Ingress traffic on port 0 is not restricted anymore to the defined flow rules
3897 testpmd>
3898
3899Sample QinQ flow rules
3900~~~~~~~~~~~~~~~~~~~~~~
3901
3902Before creating QinQ rule(s) the following commands should be issued to enable QinQ::
3903
3904   testpmd> port stop 0
3905   testpmd> vlan set qinq on 0
3906
3907The above command sets the inner and outer TPID's to 0x8100.
3908
3909To change the TPID's the following commands should be used::
3910
3911   testpmd> vlan set outer tpid 0xa100 0
3912   testpmd> vlan set inner tpid 0x9100 0
3913   testpmd> port start 0
3914
3915Validate and create a QinQ rule on port 0 to steer traffic to a VF queue in a VM.
3916
3917::
3918
3919   testpmd> flow validate 0 ingress pattern eth / vlan tci is 123 /
3920       vlan tci is 456 / end actions vf id 1 / queue index 0 / end
3921   Flow rule #0 validated
3922
3923   testpmd> flow create 0 ingress pattern eth / vlan tci is 4 /
3924       vlan tci is 456 / end actions vf id 123 / queue index 0 / end
3925   Flow rule #0 created
3926
3927   testpmd> flow list 0
3928   ID      Group   Prio    Attr    Rule
3929   0       0       0       i-      ETH VLAN VLAN=>VF QUEUE
3930
3931Validate and create a QinQ rule on port 0 to steer traffic to a queue on the host.
3932
3933::
3934
3935   testpmd> flow validate 0 ingress pattern eth / vlan tci is 321 /
3936        vlan tci is 654 / end actions pf / queue index 0 / end
3937   Flow rule #1 validated
3938
3939   testpmd> flow create 0 ingress pattern eth / vlan tci is 321 /
3940        vlan tci is 654 / end actions pf / queue index 1 / end
3941   Flow rule #1 created
3942
3943   testpmd> flow list 0
3944   ID      Group   Prio    Attr    Rule
3945   0       0       0       i-      ETH VLAN VLAN=>VF QUEUE
3946   1       0       0       i-      ETH VLAN VLAN=>PF QUEUE
3947
3948Sample VXLAN encapsulation rule
3949~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
3950
3951VXLAN encapsulation outer layer has default value pre-configured in testpmd
3952source code, those can be changed by using the following commands
3953
3954IPv4 VXLAN outer header::
3955
3956 testpmd> set vxlan ip-version ipv4 vni 4 udp-src 4 udp-dst 4 ip-src 127.0.0.1
3957        ip-dst 128.0.0.1 eth-src 11:11:11:11:11:11 eth-dst 22:22:22:22:22:22
3958 testpmd> flow create 0 ingress pattern end actions vxlan_encap /
3959        queue index 0 / end
3960
3961 testpmd> set vxlan-with-vlan ip-version ipv4 vni 4 udp-src 4 udp-dst 4 ip-src
3962         127.0.0.1 ip-dst 128.0.0.1 vlan-tci 34 eth-src 11:11:11:11:11:11
3963         eth-dst 22:22:22:22:22:22
3964 testpmd> flow create 0 ingress pattern end actions vxlan_encap /
3965         queue index 0 / end
3966
3967IPv6 VXLAN outer header::
3968
3969 testpmd> set vxlan ip-version ipv6 vni 4 udp-src 4 udp-dst 4 ip-src ::1
3970        ip-dst ::2222 eth-src 11:11:11:11:11:11 eth-dst 22:22:22:22:22:22
3971 testpmd> flow create 0 ingress pattern end actions vxlan_encap /
3972         queue index 0 / end
3973
3974 testpmd> set vxlan-with-vlan ip-version ipv6 vni 4 udp-src 4 udp-dst 4
3975         ip-src ::1 ip-dst ::2222 vlan-tci 34 eth-src 11:11:11:11:11:11
3976         eth-dst 22:22:22:22:22:22
3977 testpmd> flow create 0 ingress pattern end actions vxlan_encap /
3978         queue index 0 / end
3979
3980BPF Functions
3981--------------
3982
3983The following sections show functions to load/unload eBPF based filters.
3984
3985bpf-load
3986~~~~~~~~
3987
3988Load an eBPF program as a callback for partciular RX/TX queue::
3989
3990   testpmd> bpf-load rx|tx (portid) (queueid) (load-flags) (bpf-prog-filename)
3991
3992The available load-flags are:
3993
3994* ``J``: use JIT generated native code, otherwise BPF interpreter will be used.
3995
3996* ``M``: assume input parameter is a pointer to rte_mbuf, otherwise assume it is a pointer to first segment's data.
3997
3998* ``-``: none.
3999
4000.. note::
4001
4002   You'll need clang v3.7 or above to build bpf program you'd like to load
4003
4004For example:
4005
4006.. code-block:: console
4007
4008   cd test/bpf
4009   clang -O2 -target bpf -c t1.c
4010
4011Then to load (and JIT compile) t1.o at RX queue 0, port 1::
4012
4013.. code-block:: console
4014
4015   testpmd> bpf-load rx 1 0 J ./dpdk.org/test/bpf/t1.o
4016
4017To load (not JITed) t1.o at TX queue 0, port 0::
4018
4019.. code-block:: console
4020
4021   testpmd> bpf-load tx 0 0 - ./dpdk.org/test/bpf/t1.o
4022
4023bpf-unload
4024~~~~~~~~~~
4025
4026Unload previously loaded eBPF program for partciular RX/TX queue::
4027
4028   testpmd> bpf-unload rx|tx (portid) (queueid)
4029
4030For example to unload BPF filter from TX queue 0, port 0:
4031
4032.. code-block:: console
4033
4034   testpmd> bpf-load tx 0 0 - ./dpdk.org/test/bpf/t1.o
4035