xref: /dpdk/doc/guides/testpmd_app_ug/testpmd_funcs.rst (revision d326e671663d60c9adad87f0629b1667676733b8)
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
396
397Configuration Functions
398-----------------------
399
400The testpmd application can be configured from the runtime as well as from the command-line.
401
402This section details the available configuration functions that are available.
403
404.. note::
405
406   Configuration changes only become active when forwarding is started/restarted.
407
408set default
409~~~~~~~~~~~
410
411Reset forwarding to the default configuration::
412
413   testpmd> set default
414
415set verbose
416~~~~~~~~~~~
417
418Set the debug verbosity level::
419
420   testpmd> set verbose (level)
421
422Currently the only available levels are 0 (silent except for error) and 1 (fully verbose).
423
424set log
425~~~~~~~
426
427Set the log level for a log type::
428
429	testpmd> set log global|(type) (level)
430
431Where:
432
433* ``type`` is the log name.
434
435* ``level`` is the log level.
436
437For example, to change the global log level::
438	testpmd> set log global (level)
439
440Regexes can also be used for type. To change log level of user1, user2 and user3::
441	testpmd> set log user[1-3] (level)
442
443set nbport
444~~~~~~~~~~
445
446Set the number of ports used by the application:
447
448set nbport (num)
449
450This is equivalent to the ``--nb-ports`` command-line option.
451
452set nbcore
453~~~~~~~~~~
454
455Set the number of cores used by the application::
456
457   testpmd> set nbcore (num)
458
459This is equivalent to the ``--nb-cores`` command-line option.
460
461.. note::
462
463   The number of cores used must not be greater than number of ports used multiplied by the number of queues per port.
464
465set coremask
466~~~~~~~~~~~~
467
468Set the forwarding cores hexadecimal mask::
469
470   testpmd> set coremask (mask)
471
472This is equivalent to the ``--coremask`` command-line option.
473
474.. note::
475
476   The master lcore is reserved for command line parsing only and cannot be masked on for packet forwarding.
477
478set portmask
479~~~~~~~~~~~~
480
481Set the forwarding ports hexadecimal mask::
482
483   testpmd> set portmask (mask)
484
485This is equivalent to the ``--portmask`` command-line option.
486
487set burst
488~~~~~~~~~
489
490Set number of packets per burst::
491
492   testpmd> set burst (num)
493
494This is equivalent to the ``--burst command-line`` option.
495
496When retry is enabled, the transmit delay time and number of retries can also be set::
497
498   testpmd> set burst tx delay (microseconds) retry (num)
499
500set txpkts
501~~~~~~~~~~
502
503Set the length of each segment of the TX-ONLY packets or length of packet for FLOWGEN mode::
504
505   testpmd> set txpkts (x[,y]*)
506
507Where x[,y]* represents a CSV list of values, without white space.
508
509set txsplit
510~~~~~~~~~~~
511
512Set the split policy for the TX packets, applicable for TX-ONLY and CSUM forwarding modes::
513
514   testpmd> set txsplit (off|on|rand)
515
516Where:
517
518* ``off`` disable packet copy & split for CSUM mode.
519
520* ``on`` split outgoing packet into multiple segments. Size of each segment
521  and number of segments per packet is determined by ``set txpkts`` command
522  (see above).
523
524* ``rand`` same as 'on', but number of segments per each packet is a random value between 1 and total number of segments.
525
526set corelist
527~~~~~~~~~~~~
528
529Set the list of forwarding cores::
530
531   testpmd> set corelist (x[,y]*)
532
533For example, to change the forwarding cores:
534
535.. code-block:: console
536
537   testpmd> set corelist 3,1
538   testpmd> show config fwd
539
540   io packet forwarding - ports=2 - cores=2 - streams=2 - NUMA support disabled
541   Logical Core 3 (socket 0) forwards packets on 1 streams:
542   RX P=0/Q=0 (socket 0) -> TX P=1/Q=0 (socket 0) peer=02:00:00:00:00:01
543   Logical Core 1 (socket 0) forwards packets on 1 streams:
544   RX P=1/Q=0 (socket 0) -> TX P=0/Q=0 (socket 0) peer=02:00:00:00:00:00
545
546.. note::
547
548   The cores are used in the same order as specified on the command line.
549
550set portlist
551~~~~~~~~~~~~
552
553Set the list of forwarding ports::
554
555   testpmd> set portlist (x[,y]*)
556
557For example, to change the port forwarding:
558
559.. code-block:: console
560
561   testpmd> set portlist 0,2,1,3
562   testpmd> show config fwd
563
564   io packet forwarding - ports=4 - cores=1 - streams=4
565   Logical Core 3 (socket 0) forwards packets on 4 streams:
566   RX P=0/Q=0 (socket 0) -> TX P=2/Q=0 (socket 0) peer=02:00:00:00:00:01
567   RX P=2/Q=0 (socket 0) -> TX P=0/Q=0 (socket 0) peer=02:00:00:00:00:00
568   RX P=1/Q=0 (socket 0) -> TX P=3/Q=0 (socket 0) peer=02:00:00:00:00:03
569   RX P=3/Q=0 (socket 0) -> TX P=1/Q=0 (socket 0) peer=02:00:00:00:00:02
570
571set tx loopback
572~~~~~~~~~~~~~~~
573
574Enable/disable tx loopback::
575
576   testpmd> set tx loopback (port_id) (on|off)
577
578set drop enable
579~~~~~~~~~~~~~~~
580
581set drop enable bit for all queues::
582
583   testpmd> set all queues drop (port_id) (on|off)
584
585set split drop enable (for VF)
586~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
587
588set split drop enable bit for VF from PF::
589
590   testpmd> set vf split drop (port_id) (vf_id) (on|off)
591
592set mac antispoof (for VF)
593~~~~~~~~~~~~~~~~~~~~~~~~~~
594
595Set mac antispoof for a VF from the PF::
596
597   testpmd> set vf mac antispoof  (port_id) (vf_id) (on|off)
598
599set macsec offload
600~~~~~~~~~~~~~~~~~~
601
602Enable/disable MACsec offload::
603
604   testpmd> set macsec offload (port_id) on encrypt (on|off) replay-protect (on|off)
605   testpmd> set macsec offload (port_id) off
606
607set macsec sc
608~~~~~~~~~~~~~
609
610Configure MACsec secure connection (SC)::
611
612   testpmd> set macsec sc (tx|rx) (port_id) (mac) (pi)
613
614.. note::
615
616   The pi argument is ignored for tx.
617   Check the NIC Datasheet for hardware limits.
618
619set macsec sa
620~~~~~~~~~~~~~
621
622Configure MACsec secure association (SA)::
623
624   testpmd> set macsec sa (tx|rx) (port_id) (idx) (an) (pn) (key)
625
626.. note::
627
628   The IDX value must be 0 or 1.
629   Check the NIC Datasheet for hardware limits.
630
631set broadcast mode (for VF)
632~~~~~~~~~~~~~~~~~~~~~~~~~~~
633
634Set broadcast mode for a VF from the PF::
635
636   testpmd> set vf broadcast (port_id) (vf_id) (on|off)
637
638vlan set strip
639~~~~~~~~~~~~~~
640
641Set the VLAN strip on a port::
642
643   testpmd> vlan set strip (on|off) (port_id)
644
645vlan set stripq
646~~~~~~~~~~~~~~~
647
648Set the VLAN strip for a queue on a port::
649
650   testpmd> vlan set stripq (on|off) (port_id,queue_id)
651
652vlan set stripq (for VF)
653~~~~~~~~~~~~~~~~~~~~~~~~
654
655Set VLAN strip for all queues in a pool for a VF from the PF::
656
657   testpmd> set vf vlan stripq (port_id) (vf_id) (on|off)
658
659vlan set insert (for VF)
660~~~~~~~~~~~~~~~~~~~~~~~~
661
662Set VLAN insert for a VF from the PF::
663
664   testpmd> set vf vlan insert (port_id) (vf_id) (vlan_id)
665
666vlan set tag (for VF)
667~~~~~~~~~~~~~~~~~~~~~
668
669Set VLAN tag for a VF from the PF::
670
671   testpmd> set vf vlan tag (port_id) (vf_id) (on|off)
672
673vlan set antispoof (for VF)
674~~~~~~~~~~~~~~~~~~~~~~~~~~~
675
676Set VLAN antispoof for a VF from the PF::
677
678   testpmd> set vf vlan antispoof (port_id) (vf_id) (on|off)
679
680vlan set filter
681~~~~~~~~~~~~~~~
682
683Set the VLAN filter on a port::
684
685   testpmd> vlan set filter (on|off) (port_id)
686
687vlan set qinq
688~~~~~~~~~~~~~
689
690Set the VLAN QinQ (extended queue in queue) on for a port::
691
692   testpmd> vlan set qinq (on|off) (port_id)
693
694vlan set tpid
695~~~~~~~~~~~~~
696
697Set the inner or outer VLAN TPID for packet filtering on a port::
698
699   testpmd> vlan set (inner|outer) tpid (value) (port_id)
700
701.. note::
702
703   TPID value must be a 16-bit number (value <= 65536).
704
705rx_vlan add
706~~~~~~~~~~~
707
708Add a VLAN ID, or all identifiers, to the set of VLAN identifiers filtered by port ID::
709
710   testpmd> rx_vlan add (vlan_id|all) (port_id)
711
712.. note::
713
714   VLAN filter must be set on that port. VLAN ID < 4096.
715   Depending on the NIC used, number of vlan_ids may be limited to the maximum entries
716   in VFTA table. This is important if enabling all vlan_ids.
717
718rx_vlan rm
719~~~~~~~~~~
720
721Remove a VLAN ID, or all identifiers, from the set of VLAN identifiers filtered by port ID::
722
723   testpmd> rx_vlan rm (vlan_id|all) (port_id)
724
725rx_vlan add (for VF)
726~~~~~~~~~~~~~~~~~~~~
727
728Add a VLAN ID, to the set of VLAN identifiers filtered for VF(s) for port ID::
729
730   testpmd> rx_vlan add (vlan_id) port (port_id) vf (vf_mask)
731
732rx_vlan rm (for VF)
733~~~~~~~~~~~~~~~~~~~
734
735Remove a VLAN ID, from the set of VLAN identifiers filtered for VF(s) for port ID::
736
737   testpmd> rx_vlan rm (vlan_id) port (port_id) vf (vf_mask)
738
739tunnel_filter add
740~~~~~~~~~~~~~~~~~
741
742Add a tunnel filter on a port::
743
744   testpmd> tunnel_filter add (port_id) (outer_mac) (inner_mac) (ip_addr) \
745            (inner_vlan) (vxlan|nvgre|ipingre) (imac-ivlan|imac-ivlan-tenid|\
746            imac-tenid|imac|omac-imac-tenid|oip|iip) (tenant_id) (queue_id)
747
748The available information categories are:
749
750* ``vxlan``: Set tunnel type as VXLAN.
751
752* ``nvgre``: Set tunnel type as NVGRE.
753
754* ``ipingre``: Set tunnel type as IP-in-GRE.
755
756* ``imac-ivlan``: Set filter type as Inner MAC and VLAN.
757
758* ``imac-ivlan-tenid``: Set filter type as Inner MAC, VLAN and tenant ID.
759
760* ``imac-tenid``: Set filter type as Inner MAC and tenant ID.
761
762* ``imac``: Set filter type as Inner MAC.
763
764* ``omac-imac-tenid``: Set filter type as Outer MAC, Inner MAC and tenant ID.
765
766* ``oip``: Set filter type as Outer IP.
767
768* ``iip``: Set filter type as Inner IP.
769
770Example::
771
772   testpmd> tunnel_filter add 0 68:05:CA:28:09:82 00:00:00:00:00:00 \
773            192.168.2.2 0 ipingre oip 1 1
774
775   Set an IP-in-GRE tunnel on port 0, and the filter type is Outer IP.
776
777tunnel_filter remove
778~~~~~~~~~~~~~~~~~~~~
779
780Remove a tunnel filter on a port::
781
782   testpmd> tunnel_filter rm (port_id) (outer_mac) (inner_mac) (ip_addr) \
783            (inner_vlan) (vxlan|nvgre|ipingre) (imac-ivlan|imac-ivlan-tenid|\
784            imac-tenid|imac|omac-imac-tenid|oip|iip) (tenant_id) (queue_id)
785
786rx_vxlan_port add
787~~~~~~~~~~~~~~~~~
788
789Add an UDP port for VXLAN packet filter on a port::
790
791   testpmd> rx_vxlan_port add (udp_port) (port_id)
792
793rx_vxlan_port remove
794~~~~~~~~~~~~~~~~~~~~
795
796Remove an UDP port for VXLAN packet filter on a port::
797
798   testpmd> rx_vxlan_port rm (udp_port) (port_id)
799
800tx_vlan set
801~~~~~~~~~~~
802
803Set hardware insertion of VLAN IDs in packets sent on a port::
804
805   testpmd> tx_vlan set (port_id) vlan_id[, vlan_id_outer]
806
807For example, set a single VLAN ID (5) insertion on port 0::
808
809   tx_vlan set 0 5
810
811Or, set double VLAN ID (inner: 2, outer: 3) insertion on port 1::
812
813   tx_vlan set 1 2 3
814
815
816tx_vlan set pvid
817~~~~~~~~~~~~~~~~
818
819Set port based hardware insertion of VLAN ID in packets sent on a port::
820
821   testpmd> tx_vlan set pvid (port_id) (vlan_id) (on|off)
822
823tx_vlan reset
824~~~~~~~~~~~~~
825
826Disable hardware insertion of a VLAN header in packets sent on a port::
827
828   testpmd> tx_vlan reset (port_id)
829
830csum set
831~~~~~~~~
832
833Select hardware or software calculation of the checksum when
834transmitting a packet using the ``csum`` forwarding engine::
835
836   testpmd> csum set (ip|udp|tcp|sctp|outer-ip) (hw|sw) (port_id)
837
838Where:
839
840* ``ip|udp|tcp|sctp`` always relate to  the inner layer.
841
842* ``outer-ip`` relates to the outer IP layer (only for IPv4) in the case where the packet is recognized
843  as a tunnel packet by the forwarding engine (vxlan, gre and ipip are
844  supported). See also the ``csum parse-tunnel`` command.
845
846.. note::
847
848   Check the NIC Datasheet for hardware limits.
849
850RSS queue region
851~~~~~~~~~~~~~~~~
852
853Set RSS queue region span on a port::
854
855   testpmd> set port (port_id) queue-region region_id (value) \
856		queue_start_index (value) queue_num (value)
857
858Set flowtype mapping on a RSS queue region on a port::
859
860   testpmd> set port (port_id) queue-region region_id (value) flowtype (value)
861
862where:
863
864* For the flowtype(pctype) of packet,the specific index for each type has
865  been defined in file i40e_type.h as enum i40e_filter_pctype.
866
867Set user priority mapping on a RSS queue region on a port::
868
869   testpmd> set port (port_id) queue-region UP (value) region_id (value)
870
871Flush all queue region related configuration on a port::
872
873   testpmd> set port (port_id) queue-region flush (on|off)
874
875where:
876
877* "on"is just an enable function which server for other configuration,
878  it is for all configuration about queue region from up layer,
879  at first will only keep in DPDK softwarestored in driver,
880  only after "flush on", it commit all configuration to HW.
881  "off" is just clean all configuration about queue region just now,
882  and restore all to DPDK i40e driver default config when start up.
883
884Show all queue region related configuration info on a port::
885
886   testpmd> show port (port_id) queue-region
887
888.. note::
889
890  Queue region only support on PF by now, so these command is
891  only for configuration of queue region on PF port.
892
893csum parse-tunnel
894~~~~~~~~~~~~~~~~~
895
896Define how tunneled packets should be handled by the csum forward
897engine::
898
899   testpmd> csum parse-tunnel (on|off) (tx_port_id)
900
901If enabled, the csum forward engine will try to recognize supported
902tunnel headers (vxlan, gre, ipip).
903
904If disabled, treat tunnel packets as non-tunneled packets (a inner
905header is handled as a packet payload).
906
907.. note::
908
909   The port argument is the TX port like in the ``csum set`` command.
910
911Example:
912
913Consider a packet in packet like the following::
914
915   eth_out/ipv4_out/udp_out/vxlan/eth_in/ipv4_in/tcp_in
916
917* If parse-tunnel is enabled, the ``ip|udp|tcp|sctp`` parameters of ``csum set``
918  command relate to the inner headers (here ``ipv4_in`` and ``tcp_in``), and the
919  ``outer-ip parameter`` relates to the outer headers (here ``ipv4_out``).
920
921* If parse-tunnel is disabled, the ``ip|udp|tcp|sctp`` parameters of ``csum  set``
922   command relate to the outer headers, here ``ipv4_out`` and ``udp_out``.
923
924csum show
925~~~~~~~~~
926
927Display tx checksum offload configuration::
928
929   testpmd> csum show (port_id)
930
931tso set
932~~~~~~~
933
934Enable TCP Segmentation Offload (TSO) in the ``csum`` forwarding engine::
935
936   testpmd> tso set (segsize) (port_id)
937
938.. note::
939
940   Check the NIC datasheet for hardware limits.
941
942tso show
943~~~~~~~~
944
945Display the status of TCP Segmentation Offload::
946
947   testpmd> tso show (port_id)
948
949set port - gro
950~~~~~~~~~~~~~~
951
952Enable or disable GRO in ``csum`` forwarding engine::
953
954   testpmd> set port <port_id> gro on|off
955
956If enabled, the csum forwarding engine will perform GRO on the TCP/IPv4
957packets received from the given port.
958
959If disabled, packets received from the given port won't be performed
960GRO. By default, GRO is disabled for all ports.
961
962.. note::
963
964   When enable GRO for a port, TCP/IPv4 packets received from the port
965   will be performed GRO. After GRO, all merged packets have bad
966   checksums, since the GRO library doesn't re-calculate checksums for
967   the merged packets. Therefore, if users want the merged packets to
968   have correct checksums, please select HW IP checksum calculation and
969   HW TCP checksum calculation for the port which the merged packets are
970   transmitted to.
971
972show port - gro
973~~~~~~~~~~~~~~~
974
975Display GRO configuration for a given port::
976
977   testpmd> show port <port_id> gro
978
979set gro flush
980~~~~~~~~~~~~~
981
982Set the cycle to flush the GROed packets from reassembly tables::
983
984   testpmd> set gro flush <cycles>
985
986When enable GRO, the csum forwarding engine performs GRO on received
987packets, and the GROed packets are stored in reassembly tables. Users
988can use this command to determine when the GROed packets are flushed
989from the reassembly tables.
990
991The ``cycles`` is measured in GRO operation times. The csum forwarding
992engine flushes the GROed packets from the tables every ``cycles`` GRO
993operations.
994
995By default, the value of ``cycles`` is 1, which means flush GROed packets
996from the reassembly tables as soon as one GRO operation finishes. The value
997of ``cycles`` should be in the range of 1 to ``GRO_MAX_FLUSH_CYCLES``.
998
999Please note that the large value of ``cycles`` may cause the poor TCP/IP
1000stack performance. Because the GROed packets are delayed to arrive the
1001stack, thus causing more duplicated ACKs and TCP retransmissions.
1002
1003set port - gso
1004~~~~~~~~~~~~~~
1005
1006Toggle per-port GSO support in ``csum`` forwarding engine::
1007
1008   testpmd> set port <port_id> gso on|off
1009
1010If enabled, the csum forwarding engine will perform GSO on supported IPv4
1011packets, transmitted on the given port.
1012
1013If disabled, packets transmitted on the given port will not undergo GSO.
1014By default, GSO is disabled for all ports.
1015
1016.. note::
1017
1018   When GSO is enabled on a port, supported IPv4 packets transmitted on that
1019   port undergo GSO. Afterwards, the segmented packets are represented by
1020   multi-segment mbufs; however, the csum forwarding engine doesn't calculation
1021   of checksums for GSO'd segments in SW. As a result, if users want correct
1022   checksums in GSO segments, they should enable HW checksum calculation for
1023   GSO-enabled ports.
1024
1025   For example, HW checksum calculation for VxLAN GSO'd packets may be enabled
1026   by setting the following options in the csum forwarding engine:
1027
1028   testpmd> csum set outer_ip hw <port_id>
1029
1030   testpmd> csum set ip hw <port_id>
1031
1032   testpmd> csum set tcp hw <port_id>
1033
1034set gso segsz
1035~~~~~~~~~~~~~
1036
1037Set the maximum GSO segment size (measured in bytes), which includes the
1038packet header and the packet payload for GSO-enabled ports (global)::
1039
1040   testpmd> set gso segsz <length>
1041
1042show port - gso
1043~~~~~~~~~~~~~~~
1044
1045Display the status of Generic Segmentation Offload for a given port::
1046
1047   testpmd> show port <port_id> gso
1048
1049mac_addr add
1050~~~~~~~~~~~~
1051
1052Add an alternative MAC address to a port::
1053
1054   testpmd> mac_addr add (port_id) (XX:XX:XX:XX:XX:XX)
1055
1056mac_addr remove
1057~~~~~~~~~~~~~~~
1058
1059Remove a MAC address from a port::
1060
1061   testpmd> mac_addr remove (port_id) (XX:XX:XX:XX:XX:XX)
1062
1063mac_addr add (for VF)
1064~~~~~~~~~~~~~~~~~~~~~
1065
1066Add an alternative MAC address for a VF to a port::
1067
1068   testpmd> mac_add add port (port_id) vf (vf_id) (XX:XX:XX:XX:XX:XX)
1069
1070mac_addr set
1071~~~~~~~~~~~~
1072
1073Set the default MAC address for a port::
1074
1075   testpmd> mac_addr set (port_id) (XX:XX:XX:XX:XX:XX)
1076
1077mac_addr set (for VF)
1078~~~~~~~~~~~~~~~~~~~~~
1079
1080Set the MAC address for a VF from the PF::
1081
1082   testpmd> set vf mac addr (port_id) (vf_id) (XX:XX:XX:XX:XX:XX)
1083
1084set eth-peer
1085~~~~~~~~~~~~
1086
1087Set the forwarding peer address for certain port::
1088
1089   testpmd> set eth-peer (port_id) (perr_addr)
1090
1091This is equivalent to the ``--eth-peer`` command-line option.
1092
1093set port-uta
1094~~~~~~~~~~~~
1095
1096Set the unicast hash filter(s) on/off for a port::
1097
1098   testpmd> set port (port_id) uta (XX:XX:XX:XX:XX:XX|all) (on|off)
1099
1100set promisc
1101~~~~~~~~~~~
1102
1103Set the promiscuous mode on for a port or for all ports.
1104In promiscuous mode packets are not dropped if they aren't for the specified MAC address::
1105
1106   testpmd> set promisc (port_id|all) (on|off)
1107
1108set allmulti
1109~~~~~~~~~~~~
1110
1111Set the allmulti mode for a port or for all ports::
1112
1113   testpmd> set allmulti (port_id|all) (on|off)
1114
1115Same as the ifconfig (8) option. Controls how multicast packets are handled.
1116
1117set promisc (for VF)
1118~~~~~~~~~~~~~~~~~~~~
1119
1120Set the unicast promiscuous mode for a VF from PF.
1121It's supported by Intel i40e NICs now.
1122In promiscuous mode packets are not dropped if they aren't for the specified MAC address::
1123
1124   testpmd> set vf promisc (port_id) (vf_id) (on|off)
1125
1126set allmulticast (for VF)
1127~~~~~~~~~~~~~~~~~~~~~~~~~
1128
1129Set the multicast promiscuous mode for a VF from PF.
1130It's supported by Intel i40e NICs now.
1131In promiscuous mode packets are not dropped if they aren't for the specified MAC address::
1132
1133   testpmd> set vf allmulti (port_id) (vf_id) (on|off)
1134
1135set tx max bandwidth (for VF)
1136~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
1137
1138Set TX max absolute bandwidth (Mbps) for a VF from PF::
1139
1140   testpmd> set vf tx max-bandwidth (port_id) (vf_id) (max_bandwidth)
1141
1142set tc tx min bandwidth (for VF)
1143~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
1144
1145Set all TCs' TX min relative bandwidth (%) for a VF from PF::
1146
1147   testpmd> set vf tc tx min-bandwidth (port_id) (vf_id) (bw1, bw2, ...)
1148
1149set tc tx max bandwidth (for VF)
1150~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
1151
1152Set a TC's TX max absolute bandwidth (Mbps) for a VF from PF::
1153
1154   testpmd> set vf tc tx max-bandwidth (port_id) (vf_id) (tc_no) (max_bandwidth)
1155
1156set tc strict link priority mode
1157~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
1158
1159Set some TCs' strict link priority mode on a physical port::
1160
1161   testpmd> set tx strict-link-priority (port_id) (tc_bitmap)
1162
1163set tc tx min bandwidth
1164~~~~~~~~~~~~~~~~~~~~~~~
1165
1166Set all TCs' TX min relative bandwidth (%) globally for all PF and VFs::
1167
1168   testpmd> set tc tx min-bandwidth (port_id) (bw1, bw2, ...)
1169
1170set flow_ctrl rx
1171~~~~~~~~~~~~~~~~
1172
1173Set the link flow control parameter on a port::
1174
1175   testpmd> set flow_ctrl rx (on|off) tx (on|off) (high_water) (low_water) \
1176            (pause_time) (send_xon) mac_ctrl_frame_fwd (on|off) \
1177	    autoneg (on|off) (port_id)
1178
1179Where:
1180
1181* ``high_water`` (integer): High threshold value to trigger XOFF.
1182
1183* ``low_water`` (integer): Low threshold value to trigger XON.
1184
1185* ``pause_time`` (integer): Pause quota in the Pause frame.
1186
1187* ``send_xon`` (0/1): Send XON frame.
1188
1189* ``mac_ctrl_frame_fwd``: Enable receiving MAC control frames.
1190
1191* ``autoneg``: Change the auto-negotiation parameter.
1192
1193set pfc_ctrl rx
1194~~~~~~~~~~~~~~~
1195
1196Set the priority flow control parameter on a port::
1197
1198   testpmd> set pfc_ctrl rx (on|off) tx (on|off) (high_water) (low_water) \
1199            (pause_time) (priority) (port_id)
1200
1201Where:
1202
1203* ``high_water`` (integer): High threshold value.
1204
1205* ``low_water`` (integer): Low threshold value.
1206
1207* ``pause_time`` (integer): Pause quota in the Pause frame.
1208
1209* ``priority`` (0-7): VLAN User Priority.
1210
1211set stat_qmap
1212~~~~~~~~~~~~~
1213
1214Set statistics mapping (qmapping 0..15) for RX/TX queue on port::
1215
1216   testpmd> set stat_qmap (tx|rx) (port_id) (queue_id) (qmapping)
1217
1218For example, to set rx queue 2 on port 0 to mapping 5::
1219
1220   testpmd>set stat_qmap rx 0 2 5
1221
1222set xstats-hide-zero
1223~~~~~~~~~~~~~~~~~~~~
1224
1225Set the option to hide zero values for xstats display::
1226
1227	testpmd> set xstats-hide-zero on|off
1228
1229.. note::
1230
1231	By default, the zero values are displayed for xstats.
1232
1233set port - rx/tx (for VF)
1234~~~~~~~~~~~~~~~~~~~~~~~~~
1235
1236Set VF receive/transmit from a port::
1237
1238   testpmd> set port (port_id) vf (vf_id) (rx|tx) (on|off)
1239
1240set port - mac address filter (for VF)
1241~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
1242
1243Add/Remove unicast or multicast MAC addr filter for a VF::
1244
1245   testpmd> set port (port_id) vf (vf_id) (mac_addr) \
1246            (exact-mac|exact-mac-vlan|hashmac|hashmac-vlan) (on|off)
1247
1248set port - rx mode(for VF)
1249~~~~~~~~~~~~~~~~~~~~~~~~~~
1250
1251Set the VF receive mode of a port::
1252
1253   testpmd> set port (port_id) vf (vf_id) \
1254            rxmode (AUPE|ROPE|BAM|MPE) (on|off)
1255
1256The available receive modes are:
1257
1258* ``AUPE``: Accepts untagged VLAN.
1259
1260* ``ROPE``: Accepts unicast hash.
1261
1262* ``BAM``: Accepts broadcast packets.
1263
1264* ``MPE``: Accepts all multicast packets.
1265
1266set port - tx_rate (for Queue)
1267~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
1268
1269Set TX rate limitation for a queue on a port::
1270
1271   testpmd> set port (port_id) queue (queue_id) rate (rate_value)
1272
1273set port - tx_rate (for VF)
1274~~~~~~~~~~~~~~~~~~~~~~~~~~~
1275
1276Set TX rate limitation for queues in VF on a port::
1277
1278   testpmd> set port (port_id) vf (vf_id) rate (rate_value) queue_mask (queue_mask)
1279
1280set port - mirror rule
1281~~~~~~~~~~~~~~~~~~~~~~
1282
1283Set pool or vlan type mirror rule for a port::
1284
1285   testpmd> set port (port_id) mirror-rule (rule_id) \
1286            (pool-mirror-up|pool-mirror-down|vlan-mirror) \
1287            (poolmask|vlanid[,vlanid]*) dst-pool (pool_id) (on|off)
1288
1289Set link mirror rule for a port::
1290
1291   testpmd> set port (port_id) mirror-rule (rule_id) \
1292           (uplink-mirror|downlink-mirror) dst-pool (pool_id) (on|off)
1293
1294For example to enable mirror traffic with vlan 0,1 to pool 0::
1295
1296   set port 0 mirror-rule 0 vlan-mirror 0,1 dst-pool 0 on
1297
1298reset port - mirror rule
1299~~~~~~~~~~~~~~~~~~~~~~~~
1300
1301Reset a mirror rule for a port::
1302
1303   testpmd> reset port (port_id) mirror-rule (rule_id)
1304
1305set flush_rx
1306~~~~~~~~~~~~
1307
1308Set the flush on RX streams before forwarding.
1309The default is flush ``on``.
1310Mainly used with PCAP drivers to turn off the default behavior of flushing the first 512 packets on RX streams::
1311
1312   testpmd> set flush_rx off
1313
1314set bypass mode
1315~~~~~~~~~~~~~~~
1316
1317Set the bypass mode for the lowest port on bypass enabled NIC::
1318
1319   testpmd> set bypass mode (normal|bypass|isolate) (port_id)
1320
1321set bypass event
1322~~~~~~~~~~~~~~~~
1323
1324Set the event required to initiate specified bypass mode for the lowest port on a bypass enabled::
1325
1326   testpmd> set bypass event (timeout|os_on|os_off|power_on|power_off) \
1327            mode (normal|bypass|isolate) (port_id)
1328
1329Where:
1330
1331* ``timeout``: Enable bypass after watchdog timeout.
1332
1333* ``os_on``: Enable bypass when OS/board is powered on.
1334
1335* ``os_off``: Enable bypass when OS/board is powered off.
1336
1337* ``power_on``: Enable bypass when power supply is turned on.
1338
1339* ``power_off``: Enable bypass when power supply is turned off.
1340
1341
1342set bypass timeout
1343~~~~~~~~~~~~~~~~~~
1344
1345Set the bypass watchdog timeout to ``n`` seconds where 0 = instant::
1346
1347   testpmd> set bypass timeout (0|1.5|2|3|4|8|16|32)
1348
1349show bypass config
1350~~~~~~~~~~~~~~~~~~
1351
1352Show the bypass configuration for a bypass enabled NIC using the lowest port on the NIC::
1353
1354   testpmd> show bypass config (port_id)
1355
1356set link up
1357~~~~~~~~~~~
1358
1359Set link up for a port::
1360
1361   testpmd> set link-up port (port id)
1362
1363set link down
1364~~~~~~~~~~~~~
1365
1366Set link down for a port::
1367
1368   testpmd> set link-down port (port id)
1369
1370E-tag set
1371~~~~~~~~~
1372
1373Enable E-tag insertion for a VF on a port::
1374
1375   testpmd> E-tag set insertion on port-tag-id (value) port (port_id) vf (vf_id)
1376
1377Disable E-tag insertion for a VF on a port::
1378
1379   testpmd> E-tag set insertion off port (port_id) vf (vf_id)
1380
1381Enable/disable E-tag stripping on a port::
1382
1383   testpmd> E-tag set stripping (on|off) port (port_id)
1384
1385Enable/disable E-tag based forwarding on a port::
1386
1387   testpmd> E-tag set forwarding (on|off) port (port_id)
1388
1389Add an E-tag forwarding filter on a port::
1390
1391   testpmd> E-tag set filter add e-tag-id (value) dst-pool (pool_id) port (port_id)
1392
1393Delete an E-tag forwarding filter on a port::
1394   testpmd> E-tag set filter del e-tag-id (value) port (port_id)
1395
1396ddp add
1397~~~~~~~
1398
1399Load a dynamic device personalization (DDP) profile and store backup profile::
1400
1401   testpmd> ddp add (port_id) (profile_path[,backup_profile_path])
1402
1403ddp del
1404~~~~~~~
1405
1406Delete a dynamic device personalization profile and restore backup profile::
1407
1408   testpmd> ddp del (port_id) (backup_profile_path)
1409
1410ptype mapping
1411~~~~~~~~~~~~~
1412
1413List all items from the ptype mapping table::
1414
1415   testpmd> ptype mapping get (port_id) (valid_only)
1416
1417Where:
1418
1419* ``valid_only``: A flag indicates if only list valid items(=1) or all itemss(=0).
1420
1421Replace a specific or a group of software defined ptype with a new one::
1422
1423   testpmd> ptype mapping replace  (port_id) (target) (mask) (pkt_type)
1424
1425where:
1426
1427* ``target``: A specific software ptype or a mask to represent a group of software ptypes.
1428
1429* ``mask``: A flag indicate if "target" is a specific software ptype(=0) or a ptype mask(=1).
1430
1431* ``pkt_type``: The new software ptype to replace the old ones.
1432
1433Update hardware defined ptype to software defined packet type mapping table::
1434
1435   testpmd> ptype mapping update (port_id) (hw_ptype) (sw_ptype)
1436
1437where:
1438
1439* ``hw_ptype``: hardware ptype as the index of the ptype mapping table.
1440
1441* ``sw_ptype``: software ptype as the value of the ptype mapping table.
1442
1443Reset ptype mapping table::
1444
1445   testpmd> ptype mapping reset (port_id)
1446
1447Port Functions
1448--------------
1449
1450The following sections show functions for configuring ports.
1451
1452.. note::
1453
1454   Port configuration changes only become active when forwarding is started/restarted.
1455
1456port attach
1457~~~~~~~~~~~
1458
1459Attach a port specified by pci address or virtual device args::
1460
1461   testpmd> port attach (identifier)
1462
1463To attach a new pci device, the device should be recognized by kernel first.
1464Then it should be moved under DPDK management.
1465Finally the port can be attached to testpmd.
1466
1467For example, to move a pci device using ixgbe under DPDK management:
1468
1469.. code-block:: console
1470
1471   # Check the status of the available devices.
1472   ./usertools/dpdk-devbind.py --status
1473
1474   Network devices using DPDK-compatible driver
1475   ============================================
1476   <none>
1477
1478   Network devices using kernel driver
1479   ===================================
1480   0000:0a:00.0 '82599ES 10-Gigabit' if=eth2 drv=ixgbe unused=
1481
1482
1483   # Bind the device to igb_uio.
1484   sudo ./usertools/dpdk-devbind.py -b igb_uio 0000:0a:00.0
1485
1486
1487   # Recheck the status of the devices.
1488   ./usertools/dpdk-devbind.py --status
1489   Network devices using DPDK-compatible driver
1490   ============================================
1491   0000:0a:00.0 '82599ES 10-Gigabit' drv=igb_uio unused=
1492
1493To attach a port created by virtual device, above steps are not needed.
1494
1495For example, to attach a port whose pci address is 0000:0a:00.0.
1496
1497.. code-block:: console
1498
1499   testpmd> port attach 0000:0a:00.0
1500   Attaching a new port...
1501   EAL: PCI device 0000:0a:00.0 on NUMA socket -1
1502   EAL:   probe driver: 8086:10fb rte_ixgbe_pmd
1503   EAL:   PCI memory mapped at 0x7f83bfa00000
1504   EAL:   PCI memory mapped at 0x7f83bfa80000
1505   PMD: eth_ixgbe_dev_init(): MAC: 2, PHY: 18, SFP+: 5
1506   PMD: eth_ixgbe_dev_init(): port 0 vendorID=0x8086 deviceID=0x10fb
1507   Port 0 is attached. Now total ports is 1
1508   Done
1509
1510For example, to attach a port created by pcap PMD.
1511
1512.. code-block:: console
1513
1514   testpmd> port attach net_pcap0
1515   Attaching a new port...
1516   PMD: Initializing pmd_pcap for net_pcap0
1517   PMD: Creating pcap-backed ethdev on numa socket 0
1518   Port 0 is attached. Now total ports is 1
1519   Done
1520
1521In this case, identifier is ``net_pcap0``.
1522This identifier format is the same as ``--vdev`` format of DPDK applications.
1523
1524For example, to re-attach a bonded port which has been previously detached,
1525the mode and slave parameters must be given.
1526
1527.. code-block:: console
1528
1529   testpmd> port attach net_bond_0,mode=0,slave=1
1530   Attaching a new port...
1531   EAL: Initializing pmd_bond for net_bond_0
1532   EAL: Create bonded device net_bond_0 on port 0 in mode 0 on socket 0.
1533   Port 0 is attached. Now total ports is 1
1534   Done
1535
1536
1537port detach
1538~~~~~~~~~~~
1539
1540Detach a specific port::
1541
1542   testpmd> port detach (port_id)
1543
1544Before detaching a port, the port should be stopped and closed.
1545
1546For example, to detach a pci device port 0.
1547
1548.. code-block:: console
1549
1550   testpmd> port stop 0
1551   Stopping ports...
1552   Done
1553   testpmd> port close 0
1554   Closing ports...
1555   Done
1556
1557   testpmd> port detach 0
1558   Detaching a port...
1559   EAL: PCI device 0000:0a:00.0 on NUMA socket -1
1560   EAL:   remove driver: 8086:10fb rte_ixgbe_pmd
1561   EAL:   PCI memory unmapped at 0x7f83bfa00000
1562   EAL:   PCI memory unmapped at 0x7f83bfa80000
1563   Done
1564
1565
1566For example, to detach a virtual device port 0.
1567
1568.. code-block:: console
1569
1570   testpmd> port stop 0
1571   Stopping ports...
1572   Done
1573   testpmd> port close 0
1574   Closing ports...
1575   Done
1576
1577   testpmd> port detach 0
1578   Detaching a port...
1579   PMD: Closing pcap ethdev on numa socket 0
1580   Port 'net_pcap0' is detached. Now total ports is 0
1581   Done
1582
1583To remove a pci device completely from the system, first detach the port from testpmd.
1584Then the device should be moved under kernel management.
1585Finally the device can be removed using kernel pci hotplug functionality.
1586
1587For example, to move a pci device under kernel management:
1588
1589.. code-block:: console
1590
1591   sudo ./usertools/dpdk-devbind.py -b ixgbe 0000:0a:00.0
1592
1593   ./usertools/dpdk-devbind.py --status
1594
1595   Network devices using DPDK-compatible driver
1596   ============================================
1597   <none>
1598
1599   Network devices using kernel driver
1600   ===================================
1601   0000:0a:00.0 '82599ES 10-Gigabit' if=eth2 drv=ixgbe unused=igb_uio
1602
1603To remove a port created by a virtual device, above steps are not needed.
1604
1605port start
1606~~~~~~~~~~
1607
1608Start all ports or a specific port::
1609
1610   testpmd> port start (port_id|all)
1611
1612port stop
1613~~~~~~~~~
1614
1615Stop all ports or a specific port::
1616
1617   testpmd> port stop (port_id|all)
1618
1619port close
1620~~~~~~~~~~
1621
1622Close all ports or a specific port::
1623
1624   testpmd> port close (port_id|all)
1625
1626port start/stop queue
1627~~~~~~~~~~~~~~~~~~~~~
1628
1629Start/stop a rx/tx queue on a specific port::
1630
1631   testpmd> port (port_id) (rxq|txq) (queue_id) (start|stop)
1632
1633port setup queue
1634~~~~~~~~~~~~~~~~~~~~~
1635
1636Setup a rx/tx queue on a specific port::
1637
1638   testpmd> port (port_id) (rxq|txq) (queue_id) setup
1639
1640Only take effect when port is started.
1641
1642port config - speed
1643~~~~~~~~~~~~~~~~~~~
1644
1645Set the speed and duplex mode for all ports or a specific port::
1646
1647   testpmd> port config (port_id|all) speed (10|100|1000|10000|25000|40000|50000|100000|auto) \
1648            duplex (half|full|auto)
1649
1650port config - queues/descriptors
1651~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
1652
1653Set number of queues/descriptors for rxq, txq, rxd and txd::
1654
1655   testpmd> port config all (rxq|txq|rxd|txd) (value)
1656
1657This is equivalent to the ``--rxq``, ``--txq``, ``--rxd`` and ``--txd`` command-line options.
1658
1659port config - max-pkt-len
1660~~~~~~~~~~~~~~~~~~~~~~~~~
1661
1662Set the maximum packet length::
1663
1664   testpmd> port config all max-pkt-len (value)
1665
1666This is equivalent to the ``--max-pkt-len`` command-line option.
1667
1668port config - CRC Strip
1669~~~~~~~~~~~~~~~~~~~~~~~
1670
1671Set hardware CRC stripping on or off for all ports::
1672
1673   testpmd> port config all crc-strip (on|off)
1674
1675CRC stripping is on by default.
1676
1677The ``off`` option is equivalent to the ``--disable-crc-strip`` command-line option.
1678
1679port config - scatter
1680~~~~~~~~~~~~~~~~~~~~~~~
1681
1682Set RX scatter mode on or off for all ports::
1683
1684   testpmd> port config all scatter (on|off)
1685
1686RX scatter mode is off by default.
1687
1688The ``on`` option is equivalent to the ``--enable-scatter`` command-line option.
1689
1690port config - RX Checksum
1691~~~~~~~~~~~~~~~~~~~~~~~~~
1692
1693Set hardware RX checksum offload to on or off for all ports::
1694
1695   testpmd> port config all rx-cksum (on|off)
1696
1697Checksum offload is off by default.
1698
1699The ``on`` option is equivalent to the ``--enable-rx-cksum`` command-line option.
1700
1701port config - VLAN
1702~~~~~~~~~~~~~~~~~~
1703
1704Set hardware VLAN on or off for all ports::
1705
1706   testpmd> port config all hw-vlan (on|off)
1707
1708Hardware VLAN is off by default.
1709
1710The ``on`` option is equivalent to the ``--enable-hw-vlan`` command-line option.
1711
1712port config - VLAN filter
1713~~~~~~~~~~~~~~~~~~~~~~~~~
1714
1715Set hardware VLAN filter on or off for all ports::
1716
1717   testpmd> port config all hw-vlan-filter (on|off)
1718
1719Hardware VLAN filter is off by default.
1720
1721The ``on`` option is equivalent to the ``--enable-hw-vlan-filter`` command-line option.
1722
1723port config - VLAN strip
1724~~~~~~~~~~~~~~~~~~~~~~~~
1725
1726Set hardware VLAN strip on or off for all ports::
1727
1728   testpmd> port config all hw-vlan-strip (on|off)
1729
1730Hardware VLAN strip is off by default.
1731
1732The ``on`` option is equivalent to the ``--enable-hw-vlan-strip`` command-line option.
1733
1734port config - VLAN extend
1735~~~~~~~~~~~~~~~~~~~~~~~~~
1736
1737Set hardware VLAN extend on or off for all ports::
1738
1739   testpmd> port config all hw-vlan-extend (on|off)
1740
1741Hardware VLAN extend is off by default.
1742
1743The ``on`` option is equivalent to the ``--enable-hw-vlan-extend`` command-line option.
1744
1745port config - Drop Packets
1746~~~~~~~~~~~~~~~~~~~~~~~~~~
1747
1748Set packet drop for packets with no descriptors on or off for all ports::
1749
1750   testpmd> port config all drop-en (on|off)
1751
1752Packet dropping for packets with no descriptors is off by default.
1753
1754The ``on`` option is equivalent to the ``--enable-drop-en`` command-line option.
1755
1756port config - RSS
1757~~~~~~~~~~~~~~~~~
1758
1759Set the RSS (Receive Side Scaling) mode on or off::
1760
1761   testpmd> port config all rss (all|default|ip|tcp|udp|sctp|ether|port|vxlan|geneve|nvgre|none)
1762
1763RSS is on by default.
1764
1765The ``all`` option is equivalent to ip|tcp|udp|sctp|ether.
1766The ``default`` option enables all supported RSS types reported by device info.
1767The ``none`` option is equivalent to the ``--disable-rss`` command-line option.
1768
1769port config - RSS Reta
1770~~~~~~~~~~~~~~~~~~~~~~
1771
1772Set the RSS (Receive Side Scaling) redirection table::
1773
1774   testpmd> port config all rss reta (hash,queue)[,(hash,queue)]
1775
1776port config - DCB
1777~~~~~~~~~~~~~~~~~
1778
1779Set the DCB mode for an individual port::
1780
1781   testpmd> port config (port_id) dcb vt (on|off) (traffic_class) pfc (on|off)
1782
1783The traffic class should be 4 or 8.
1784
1785port config - Burst
1786~~~~~~~~~~~~~~~~~~~
1787
1788Set the number of packets per burst::
1789
1790   testpmd> port config all burst (value)
1791
1792This is equivalent to the ``--burst`` command-line option.
1793
1794port config - Threshold
1795~~~~~~~~~~~~~~~~~~~~~~~
1796
1797Set thresholds for TX/RX queues::
1798
1799   testpmd> port config all (threshold) (value)
1800
1801Where the threshold type can be:
1802
1803* ``txpt:`` Set the prefetch threshold register of the TX rings, 0 <= value <= 255.
1804
1805* ``txht:`` Set the host threshold register of the TX rings, 0 <= value <= 255.
1806
1807* ``txwt:`` Set the write-back threshold register of the TX rings, 0 <= value <= 255.
1808
1809* ``rxpt:`` Set the prefetch threshold register of the RX rings, 0 <= value <= 255.
1810
1811* ``rxht:`` Set the host threshold register of the RX rings, 0 <= value <= 255.
1812
1813* ``rxwt:`` Set the write-back threshold register of the RX rings, 0 <= value <= 255.
1814
1815* ``txfreet:`` Set the transmit free threshold of the TX rings, 0 <= value <= txd.
1816
1817* ``rxfreet:`` Set the transmit free threshold of the RX rings, 0 <= value <= rxd.
1818
1819* ``txrst:`` Set the transmit RS bit threshold of TX rings, 0 <= value <= txd.
1820
1821These threshold options are also available from the command-line.
1822
1823port config - E-tag
1824~~~~~~~~~~~~~~~~~~~
1825
1826Set the value of ether-type for E-tag::
1827
1828   testpmd> port config (port_id|all) l2-tunnel E-tag ether-type (value)
1829
1830Enable/disable the E-tag support::
1831
1832   testpmd> port config (port_id|all) l2-tunnel E-tag (enable|disable)
1833
1834port config pctype mapping
1835~~~~~~~~~~~~~~~~~~~~~~~~~~
1836
1837Reset pctype mapping table::
1838
1839   testpmd> port config (port_id) pctype mapping reset
1840
1841Update hardware defined pctype to software defined flow type mapping table::
1842
1843   testpmd> port config (port_id) pctype mapping update (pctype_id_0[,pctype_id_1]*) (flow_type_id)
1844
1845where:
1846
1847* ``pctype_id_x``: hardware pctype id as index of bit in bitmask value of the pctype mapping table.
1848
1849* ``flow_type_id``: software flow type id as the index of the pctype mapping table.
1850
1851port config input set
1852~~~~~~~~~~~~~~~~~~~~~
1853
1854Config RSS/FDIR/FDIR flexible payload input set for some pctype::
1855   testpmd> port config (port_id) pctype (pctype_id) \
1856            (hash_inset|fdir_inset|fdir_flx_inset) \
1857	    (get|set|clear) field (field_idx)
1858
1859Clear RSS/FDIR/FDIR flexible payload input set for some pctype::
1860   testpmd> port config (port_id) pctype (pctype_id) \
1861            (hash_inset|fdir_inset|fdir_flx_inset) clear all
1862
1863where:
1864
1865* ``pctype_id``: hardware packet classification types.
1866* ``field_idx``: hardware field index.
1867
1868Link Bonding Functions
1869----------------------
1870
1871The Link Bonding functions make it possible to dynamically create and
1872manage link bonding devices from within testpmd interactive prompt.
1873
1874create bonded device
1875~~~~~~~~~~~~~~~~~~~~
1876
1877Create a new bonding device::
1878
1879   testpmd> create bonded device (mode) (socket)
1880
1881For example, to create a bonded device in mode 1 on socket 0::
1882
1883   testpmd> create bonded 1 0
1884   created new bonded device (port X)
1885
1886add bonding slave
1887~~~~~~~~~~~~~~~~~
1888
1889Adds Ethernet device to a Link Bonding device::
1890
1891   testpmd> add bonding slave (slave id) (port id)
1892
1893For example, to add Ethernet device (port 6) to a Link Bonding device (port 10)::
1894
1895   testpmd> add bonding slave 6 10
1896
1897
1898remove bonding slave
1899~~~~~~~~~~~~~~~~~~~~
1900
1901Removes an Ethernet slave device from a Link Bonding device::
1902
1903   testpmd> remove bonding slave (slave id) (port id)
1904
1905For example, to remove Ethernet slave device (port 6) to a Link Bonding device (port 10)::
1906
1907   testpmd> remove bonding slave 6 10
1908
1909set bonding mode
1910~~~~~~~~~~~~~~~~
1911
1912Set the Link Bonding mode of a Link Bonding device::
1913
1914   testpmd> set bonding mode (value) (port id)
1915
1916For example, to set the bonding mode of a Link Bonding device (port 10) to broadcast (mode 3)::
1917
1918   testpmd> set bonding mode 3 10
1919
1920set bonding primary
1921~~~~~~~~~~~~~~~~~~~
1922
1923Set an Ethernet slave device as the primary device on a Link Bonding device::
1924
1925   testpmd> set bonding primary (slave id) (port id)
1926
1927For example, to set the Ethernet slave device (port 6) as the primary port of a Link Bonding device (port 10)::
1928
1929   testpmd> set bonding primary 6 10
1930
1931set bonding mac
1932~~~~~~~~~~~~~~~
1933
1934Set the MAC address of a Link Bonding device::
1935
1936   testpmd> set bonding mac (port id) (mac)
1937
1938For example, to set the MAC address of a Link Bonding device (port 10) to 00:00:00:00:00:01::
1939
1940   testpmd> set bonding mac 10 00:00:00:00:00:01
1941
1942set bonding xmit_balance_policy
1943~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
1944
1945Set the transmission policy for a Link Bonding device when it is in Balance XOR mode::
1946
1947   testpmd> set bonding xmit_balance_policy (port_id) (l2|l23|l34)
1948
1949For example, set a Link Bonding device (port 10) to use a balance policy of layer 3+4 (IP addresses & UDP ports)::
1950
1951   testpmd> set bonding xmit_balance_policy 10 l34
1952
1953
1954set bonding mon_period
1955~~~~~~~~~~~~~~~~~~~~~~
1956
1957Set the link status monitoring polling period in milliseconds for a bonding device.
1958
1959This adds support for PMD slave devices which do not support link status interrupts.
1960When the mon_period is set to a value greater than 0 then all PMD's which do not support
1961link status ISR will be queried every polling interval to check if their link status has changed::
1962
1963   testpmd> set bonding mon_period (port_id) (value)
1964
1965For example, to set the link status monitoring polling period of bonded device (port 5) to 150ms::
1966
1967   testpmd> set bonding mon_period 5 150
1968
1969
1970set bonding lacp dedicated_queue
1971~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
1972
1973Enable dedicated tx/rx queues on bonding devices slaves to handle LACP control plane traffic
1974when in mode 4 (link-aggregration-802.3ad)::
1975
1976   testpmd> set bonding lacp dedicated_queues (port_id) (enable|disable)
1977
1978
1979set bonding agg_mode
1980~~~~~~~~~~~~~~~~~~~~
1981
1982Enable one of the specific aggregators mode when in mode 4 (link-aggregration-802.3ad)::
1983
1984   testpmd> set bonding agg_mode (port_id) (bandwidth|count|stable)
1985
1986
1987show bonding config
1988~~~~~~~~~~~~~~~~~~~
1989
1990Show the current configuration of a Link Bonding device::
1991
1992   testpmd> show bonding config (port id)
1993
1994For example,
1995to show the configuration a Link Bonding device (port 9) with 3 slave devices (1, 3, 4)
1996in balance mode with a transmission policy of layer 2+3::
1997
1998   testpmd> show bonding config 9
1999        Bonding mode: 2
2000        Balance Xmit Policy: BALANCE_XMIT_POLICY_LAYER23
2001        Slaves (3): [1 3 4]
2002        Active Slaves (3): [1 3 4]
2003        Primary: [3]
2004
2005
2006Register Functions
2007------------------
2008
2009The Register Functions can be used to read from and write to registers on the network card referenced by a port number.
2010This is mainly useful for debugging purposes.
2011Reference should be made to the appropriate datasheet for the network card for details on the register addresses
2012and fields that can be accessed.
2013
2014read reg
2015~~~~~~~~
2016
2017Display the value of a port register::
2018
2019   testpmd> read reg (port_id) (address)
2020
2021For example, to examine the Flow Director control register (FDIRCTL, 0x0000EE000) on an Intel 82599 10 GbE Controller::
2022
2023   testpmd> read reg 0 0xEE00
2024   port 0 PCI register at offset 0xEE00: 0x4A060029 (1241907241)
2025
2026read regfield
2027~~~~~~~~~~~~~
2028
2029Display a port register bit field::
2030
2031   testpmd> read regfield (port_id) (address) (bit_x) (bit_y)
2032
2033For example, reading the lowest two bits from the register in the example above::
2034
2035   testpmd> read regfield 0 0xEE00 0 1
2036   port 0 PCI register at offset 0xEE00: bits[0, 1]=0x1 (1)
2037
2038read regbit
2039~~~~~~~~~~~
2040
2041Display a single port register bit::
2042
2043   testpmd> read regbit (port_id) (address) (bit_x)
2044
2045For example, reading the lowest bit from the register in the example above::
2046
2047   testpmd> read regbit 0 0xEE00 0
2048   port 0 PCI register at offset 0xEE00: bit 0=1
2049
2050write reg
2051~~~~~~~~~
2052
2053Set the value of a port register::
2054
2055   testpmd> write reg (port_id) (address) (value)
2056
2057For example, to clear a register::
2058
2059   testpmd> write reg 0 0xEE00 0x0
2060   port 0 PCI register at offset 0xEE00: 0x00000000 (0)
2061
2062write regfield
2063~~~~~~~~~~~~~~
2064
2065Set bit field of a port register::
2066
2067   testpmd> write regfield (port_id) (address) (bit_x) (bit_y) (value)
2068
2069For example, writing to the register cleared in the example above::
2070
2071   testpmd> write regfield 0 0xEE00 0 1 2
2072   port 0 PCI register at offset 0xEE00: 0x00000002 (2)
2073
2074write regbit
2075~~~~~~~~~~~~
2076
2077Set single bit value of a port register::
2078
2079   testpmd> write regbit (port_id) (address) (bit_x) (value)
2080
2081For example, to set the high bit in the register from the example above::
2082
2083   testpmd> write regbit 0 0xEE00 31 1
2084   port 0 PCI register at offset 0xEE00: 0x8000000A (2147483658)
2085
2086Traffic Metering and Policing
2087-----------------------------
2088
2089The following section shows functions for configuring traffic metering and
2090policing on the ethernet device through the use of generic ethdev API.
2091
2092show port traffic management capability
2093~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
2094
2095Show traffic metering and policing capability of the port::
2096
2097   testpmd> show port meter cap (port_id)
2098
2099add port meter profile (srTCM rfc2967)
2100~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
2101
2102Add meter profile (srTCM rfc2697) to the ethernet device::
2103
2104   testpmd> add port meter profile srtcm_rfc2697 (port_id) (profile_id) \
2105   (cir) (cbs) (ebs)
2106
2107where:
2108
2109* ``profile_id``: ID for the meter profile.
2110* ``cir``: Committed Information Rate (CIR) (bytes/second).
2111* ``cbs``: Committed Burst Size (CBS) (bytes).
2112* ``ebs``: Excess Burst Size (EBS) (bytes).
2113
2114add port meter profile (trTCM rfc2968)
2115~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
2116
2117Add meter profile (srTCM rfc2698) to the ethernet device::
2118
2119   testpmd> add port meter profile trtcm_rfc2698 (port_id) (profile_id) \
2120   (cir) (pir) (cbs) (pbs)
2121
2122where:
2123
2124* ``profile_id``: ID for the meter profile.
2125* ``cir``: Committed information rate (bytes/second).
2126* ``pir``: Peak information rate (bytes/second).
2127* ``cbs``: Committed burst size (bytes).
2128* ``pbs``: Peak burst size (bytes).
2129
2130add port meter profile (trTCM rfc4115)
2131~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
2132
2133Add meter profile (trTCM rfc4115) to the ethernet device::
2134
2135   testpmd> add port meter profile trtcm_rfc4115 (port_id) (profile_id) \
2136   (cir) (eir) (cbs) (ebs)
2137
2138where:
2139
2140* ``profile_id``: ID for the meter profile.
2141* ``cir``: Committed information rate (bytes/second).
2142* ``eir``: Excess information rate (bytes/second).
2143* ``cbs``: Committed burst size (bytes).
2144* ``ebs``: Excess burst size (bytes).
2145
2146delete port meter profile
2147~~~~~~~~~~~~~~~~~~~~~~~~~
2148
2149Delete meter profile from the ethernet device::
2150
2151   testpmd> del port meter profile (port_id) (profile_id)
2152
2153create port meter
2154~~~~~~~~~~~~~~~~~
2155
2156Create new meter object for the ethernet device::
2157
2158   testpmd> create port meter (port_id) (mtr_id) (profile_id) \
2159   (meter_enable) (g_action) (y_action) (r_action) (stats_mask) (shared) \
2160   (use_pre_meter_color) [(dscp_tbl_entry0) (dscp_tbl_entry1)...\
2161   (dscp_tbl_entry63)]
2162
2163where:
2164
2165* ``mtr_id``: meter object ID.
2166* ``profile_id``: ID for the meter profile.
2167* ``meter_enable``: When this parameter has a non-zero value, the meter object
2168  gets enabled at the time of creation, otherwise remains disabled.
2169* ``g_action``: Policer action for the packet with green color.
2170* ``y_action``: Policer action for the packet with yellow color.
2171* ``r_action``: Policer action for the packet with red color.
2172* ``stats_mask``: Mask of statistics counter types to be enabled for the
2173  meter object.
2174* ``shared``:  When this parameter has a non-zero value, the meter object is
2175  shared by multiple flows. Otherwise, meter object is used by single flow.
2176* ``use_pre_meter_color``: When this parameter has a non-zero value, the
2177  input color for the current meter object is determined by the latest meter
2178  object in the same flow. Otherwise, the current meter object uses the
2179  *dscp_table* to determine the input color.
2180* ``dscp_tbl_entryx``: DSCP table entry x providing meter providing input
2181  color, 0 <= x <= 63.
2182
2183enable port meter
2184~~~~~~~~~~~~~~~~~
2185
2186Enable meter for the ethernet device::
2187
2188   testpmd> enable port meter (port_id) (mtr_id)
2189
2190disable port meter
2191~~~~~~~~~~~~~~~~~~
2192
2193Disable meter for the ethernet device::
2194
2195   testpmd> disable port meter (port_id) (mtr_id)
2196
2197delete port meter
2198~~~~~~~~~~~~~~~~~
2199
2200Delete meter for the ethernet device::
2201
2202   testpmd> del port meter (port_id) (mtr_id)
2203
2204Set port meter profile
2205~~~~~~~~~~~~~~~~~~~~~~
2206
2207Set meter profile for the ethernet device::
2208
2209   testpmd> set port meter profile (port_id) (mtr_id) (profile_id)
2210
2211set port meter dscp table
2212~~~~~~~~~~~~~~~~~~~~~~~~~
2213
2214Set meter dscp table for the ethernet device::
2215
2216   testpmd> set port meter dscp table (port_id) (mtr_id) [(dscp_tbl_entry0) \
2217   (dscp_tbl_entry1)...(dscp_tbl_entry63)]
2218
2219set port meter policer action
2220~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
2221
2222Set meter policer action for the ethernet device::
2223
2224   testpmd> set port meter policer action (port_id) (mtr_id) (action_mask) \
2225   (action0) [(action1) (action1)]
2226
2227where:
2228
2229* ``action_mask``: Bit mask indicating which policer actions need to be
2230  updated. One or more policer actions can be updated in a single function
2231  invocation. To update the policer action associated with color C, bit
2232  (1 << C) needs to be set in *action_mask* and element at position C
2233  in the *actions* array needs to be valid.
2234* ``actionx``: Policer action for the color x,
2235  RTE_MTR_GREEN <= x < RTE_MTR_COLORS
2236
2237set port meter stats mask
2238~~~~~~~~~~~~~~~~~~~~~~~~~
2239
2240Set meter stats mask for the ethernet device::
2241
2242   testpmd> set port meter stats mask (port_id) (mtr_id) (stats_mask)
2243
2244where:
2245
2246* ``stats_mask``: Bit mask indicating statistics counter types to be enabled.
2247
2248show port meter stats
2249~~~~~~~~~~~~~~~~~~~~~
2250
2251Show meter stats of the ethernet device::
2252
2253   testpmd> show port meter stats (port_id) (mtr_id) (clear)
2254
2255where:
2256
2257* ``clear``: Flag that indicates whether the statistics counters should
2258  be cleared (i.e. set to zero) immediately after they have been read or not.
2259
2260Traffic Management
2261------------------
2262
2263The following section shows functions for configuring traffic management on
2264on the ethernet device through the use of generic TM API.
2265
2266show port traffic management capability
2267~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
2268
2269Show traffic management capability of the port::
2270
2271   testpmd> show port tm cap (port_id)
2272
2273show port traffic management capability (hierarchy level)
2274~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
2275
2276Show traffic management hierarchy level capability of the port::
2277
2278   testpmd> show port tm level cap (port_id) (level_id)
2279
2280show port traffic management capability (hierarchy node level)
2281~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
2282
2283Show the traffic management hierarchy node capability of the port::
2284
2285   testpmd> show port tm node cap (port_id) (node_id)
2286
2287show port traffic management hierarchy node type
2288~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
2289
2290Show the port traffic management hierarchy node type::
2291
2292   testpmd> show port tm node type (port_id) (node_id)
2293
2294show port traffic management hierarchy node stats
2295~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
2296
2297Show the port traffic management hierarchy node statistics::
2298
2299   testpmd> show port tm node stats (port_id) (node_id) (clear)
2300
2301where:
2302
2303* ``clear``: When this parameter has a non-zero value, the statistics counters
2304  are cleared (i.e. set to zero) immediately after they have been read,
2305  otherwise the statistics counters are left untouched.
2306
2307Add port traffic management private shaper profile
2308~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
2309
2310Add the port traffic management private shaper profile::
2311
2312   testpmd> add port tm node shaper profile (port_id) (shaper_profile_id) \
2313   (tb_rate) (tb_size) (packet_length_adjust)
2314
2315where:
2316
2317* ``shaper_profile id``: Shaper profile ID for the new profile.
2318* ``tb_rate``: Token bucket rate (bytes per second).
2319* ``tb_size``: Token bucket size (bytes).
2320* ``packet_length_adjust``: The value (bytes) to be added to the length of
2321  each packet for the purpose of shaping. This parameter value can be used to
2322  correct the packet length with the framing overhead bytes that are consumed
2323  on the wire.
2324
2325Delete port traffic management private shaper profile
2326~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
2327
2328Delete the port traffic management private shaper::
2329
2330   testpmd> del port tm node shaper profile (port_id) (shaper_profile_id)
2331
2332where:
2333
2334* ``shaper_profile id``: Shaper profile ID that needs to be deleted.
2335
2336Add port traffic management shared shaper
2337~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
2338
2339Create the port traffic management shared shaper::
2340
2341   testpmd> add port tm node shared shaper (port_id) (shared_shaper_id) \
2342   (shaper_profile_id)
2343
2344where:
2345
2346* ``shared_shaper_id``: Shared shaper ID to be created.
2347* ``shaper_profile id``: Shaper profile ID for shared shaper.
2348
2349Set port traffic management shared shaper
2350~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
2351
2352Update the port traffic management shared shaper::
2353
2354   testpmd> set port tm node shared shaper (port_id) (shared_shaper_id) \
2355   (shaper_profile_id)
2356
2357where:
2358
2359* ``shared_shaper_id``: Shared shaper ID to be update.
2360* ``shaper_profile id``: Shaper profile ID for shared shaper.
2361
2362Delete port traffic management shared shaper
2363~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
2364
2365Delete the port traffic management shared shaper::
2366
2367   testpmd> del port tm node shared shaper (port_id) (shared_shaper_id)
2368
2369where:
2370
2371* ``shared_shaper_id``: Shared shaper ID to be deleted.
2372
2373Set port traffic management hiearchy node private shaper
2374~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
2375
2376set the port traffic management hierarchy node private shaper::
2377
2378   testpmd> set port tm node shaper profile (port_id) (node_id) \
2379   (shaper_profile_id)
2380
2381where:
2382
2383* ``shaper_profile id``: Private shaper profile ID to be enabled on the
2384  hierarchy node.
2385
2386Add port traffic management WRED profile
2387~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
2388
2389Create a new WRED profile::
2390
2391   testpmd> add port tm node wred profile (port_id) (wred_profile_id) \
2392   (color_g) (min_th_g) (max_th_g) (maxp_inv_g) (wq_log2_g) \
2393   (color_y) (min_th_y) (max_th_y) (maxp_inv_y) (wq_log2_y) \
2394   (color_r) (min_th_r) (max_th_r) (maxp_inv_r) (wq_log2_r)
2395
2396where:
2397
2398* ``wred_profile id``: Identifier for the newly create WRED profile
2399* ``color_g``: Packet color (green)
2400* ``min_th_g``: Minimum queue threshold for packet with green color
2401* ``max_th_g``: Minimum queue threshold for packet with green color
2402* ``maxp_inv_g``: Inverse of packet marking probability maximum value (maxp)
2403* ``wq_log2_g``: Negated log2 of queue weight (wq)
2404* ``color_y``: Packet color (yellow)
2405* ``min_th_y``: Minimum queue threshold for packet with yellow color
2406* ``max_th_y``: Minimum queue threshold for packet with yellow color
2407* ``maxp_inv_y``: Inverse of packet marking probability maximum value (maxp)
2408* ``wq_log2_y``: Negated log2 of queue weight (wq)
2409* ``color_r``: Packet color (red)
2410* ``min_th_r``: Minimum queue threshold for packet with yellow color
2411* ``max_th_r``: Minimum queue threshold for packet with yellow color
2412* ``maxp_inv_r``: Inverse of packet marking probability maximum value (maxp)
2413* ``wq_log2_r``: Negated log2 of queue weight (wq)
2414
2415Delete port traffic management WRED profile
2416~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
2417
2418Delete the WRED profile::
2419
2420   testpmd> del port tm node wred profile (port_id) (wred_profile_id)
2421
2422Add port traffic management hierarchy nonleaf node
2423~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
2424
2425Add nonleaf node to port traffic management hiearchy::
2426
2427   testpmd> add port tm nonleaf node (port_id) (node_id) (parent_node_id) \
2428   (priority) (weight) (level_id) (shaper_profile_id) \
2429   (n_sp_priorities) (stats_mask) (n_shared_shapers) \
2430   [(shared_shaper_0) (shared_shaper_1) ...] \
2431
2432where:
2433
2434* ``parent_node_id``: Node ID of the parent.
2435* ``priority``: Node priority (highest node priority is zero). This is used by
2436  the SP algorithm running on the parent node for scheduling this node.
2437* ``weight``: Node weight (lowest weight is one). The node weight is relative
2438  to the weight sum of all siblings that have the same priority. It is used by
2439  the WFQ algorithm running on the parent node for scheduling this node.
2440* ``level_id``: Hiearchy level of the node.
2441* ``shaper_profile_id``: Shaper profile ID of the private shaper to be used by
2442  the node.
2443* ``n_sp_priorities``: Number of strict priorities.
2444* ``stats_mask``: Mask of statistics counter types to be enabled for this node.
2445* ``n_shared_shapers``: Number of shared shapers.
2446* ``shared_shaper_id``: Shared shaper id.
2447
2448Add port traffic management hierarchy leaf node
2449~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
2450
2451Add leaf node to port traffic management hiearchy::
2452
2453   testpmd> add port tm leaf node (port_id) (node_id) (parent_node_id) \
2454   (priority) (weight) (level_id) (shaper_profile_id) \
2455   (cman_mode) (wred_profile_id) (stats_mask) (n_shared_shapers) \
2456   [(shared_shaper_id) (shared_shaper_id) ...] \
2457
2458where:
2459
2460* ``parent_node_id``: Node ID of the parent.
2461* ``priority``: Node priority (highest node priority is zero). This is used by
2462  the SP algorithm running on the parent node for scheduling this node.
2463* ``weight``: Node weight (lowest weight is one). The node weight is relative
2464  to the weight sum of all siblings that have the same priority. It is used by
2465  the WFQ algorithm running on the parent node for scheduling this node.
2466* ``level_id``: Hiearchy level of the node.
2467* ``shaper_profile_id``: Shaper profile ID of the private shaper to be used by
2468  the node.
2469* ``cman_mode``: Congestion management mode to be enabled for this node.
2470* ``wred_profile_id``: WRED profile id to be enabled for this node.
2471* ``stats_mask``: Mask of statistics counter types to be enabled for this node.
2472* ``n_shared_shapers``: Number of shared shapers.
2473* ``shared_shaper_id``: Shared shaper id.
2474
2475Delete port traffic management hierarchy node
2476~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
2477
2478Delete node from port traffic management hiearchy::
2479
2480   testpmd> del port tm node (port_id) (node_id)
2481
2482Update port traffic management hierarchy parent node
2483~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
2484
2485Update port traffic management hierarchy parent node::
2486
2487   testpmd> set port tm node parent (port_id) (node_id) (parent_node_id) \
2488   (priority) (weight)
2489
2490This function can only be called after the hierarchy commit invocation. Its
2491success depends on the port support for this operation, as advertised through
2492the port capability set. This function is valid for all nodes of the traffic
2493management hierarchy except root node.
2494
2495Commit port traffic management hierarchy
2496~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
2497
2498Commit the traffic management hierarchy on the port::
2499
2500   testpmd> port tm hierarchy commit (port_id) (clean_on_fail)
2501
2502where:
2503
2504* ``clean_on_fail``: When set to non-zero, hierarchy is cleared on function
2505  call failure. On the other hand, hierarchy is preserved when this parameter
2506  is equal to zero.
2507
2508Set port traffic management default hierarchy (tm forwarding mode)
2509~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
2510
2511set the traffic management default hierarchy on the port::
2512
2513   testpmd> set port tm hierarchy default (port_id)
2514
2515Filter Functions
2516----------------
2517
2518This section details the available filter functions that are available.
2519
2520Note these functions interface the deprecated legacy filtering framework,
2521superseded by *rte_flow*. See `Flow rules management`_.
2522
2523ethertype_filter
2524~~~~~~~~~~~~~~~~~~~~
2525
2526Add or delete a L2 Ethertype filter, which identify packets by their L2 Ethertype mainly assign them to a receive queue::
2527
2528   ethertype_filter (port_id) (add|del) (mac_addr|mac_ignr) (mac_address) \
2529                    ethertype (ether_type) (drop|fwd) queue (queue_id)
2530
2531The available information parameters are:
2532
2533* ``port_id``: The port which the Ethertype filter assigned on.
2534
2535* ``mac_addr``: Compare destination mac address.
2536
2537* ``mac_ignr``: Ignore destination mac address match.
2538
2539* ``mac_address``: Destination mac address to match.
2540
2541* ``ether_type``: The EtherType value want to match,
2542  for example 0x0806 for ARP packet. 0x0800 (IPv4) and 0x86DD (IPv6) are invalid.
2543
2544* ``queue_id``: The receive queue associated with this EtherType filter.
2545  It is meaningless when deleting or dropping.
2546
2547Example, to add/remove an ethertype filter rule::
2548
2549   testpmd> ethertype_filter 0 add mac_ignr 00:11:22:33:44:55 \
2550                             ethertype 0x0806 fwd queue 3
2551
2552   testpmd> ethertype_filter 0 del mac_ignr 00:11:22:33:44:55 \
2553                             ethertype 0x0806 fwd queue 3
2554
25552tuple_filter
2556~~~~~~~~~~~~~~~~~
2557
2558Add or delete a 2-tuple filter,
2559which identifies packets by specific protocol and destination TCP/UDP port
2560and forwards packets into one of the receive queues::
2561
2562   2tuple_filter (port_id) (add|del) dst_port (dst_port_value) \
2563                 protocol (protocol_value) mask (mask_value) \
2564                 tcp_flags (tcp_flags_value) priority (prio_value) \
2565                 queue (queue_id)
2566
2567The available information parameters are:
2568
2569* ``port_id``: The port which the 2-tuple filter assigned on.
2570
2571* ``dst_port_value``: Destination port in L4.
2572
2573* ``protocol_value``: IP L4 protocol.
2574
2575* ``mask_value``: Participates in the match or not by bit for field above, 1b means participate.
2576
2577* ``tcp_flags_value``: TCP control bits. The non-zero value is invalid, when the pro_value is not set to 0x06 (TCP).
2578
2579* ``prio_value``: Priority of this filter.
2580
2581* ``queue_id``: The receive queue associated with this 2-tuple filter.
2582
2583Example, to add/remove an 2tuple filter rule::
2584
2585   testpmd> 2tuple_filter 0 add dst_port 32 protocol 0x06 mask 0x03 \
2586                          tcp_flags 0x02 priority 3 queue 3
2587
2588   testpmd> 2tuple_filter 0 del dst_port 32 protocol 0x06 mask 0x03 \
2589                          tcp_flags 0x02 priority 3 queue 3
2590
25915tuple_filter
2592~~~~~~~~~~~~~~~~~
2593
2594Add or delete a 5-tuple filter,
2595which consists of a 5-tuple (protocol, source and destination IP addresses, source and destination TCP/UDP/SCTP port)
2596and routes packets into one of the receive queues::
2597
2598   5tuple_filter (port_id) (add|del) dst_ip (dst_address) src_ip \
2599                 (src_address) dst_port (dst_port_value) \
2600                 src_port (src_port_value) protocol (protocol_value) \
2601                 mask (mask_value) tcp_flags (tcp_flags_value) \
2602                 priority (prio_value) queue (queue_id)
2603
2604The available information parameters are:
2605
2606* ``port_id``: The port which the 5-tuple filter assigned on.
2607
2608* ``dst_address``: Destination IP address.
2609
2610* ``src_address``: Source IP address.
2611
2612* ``dst_port_value``: TCP/UDP destination port.
2613
2614* ``src_port_value``: TCP/UDP source port.
2615
2616* ``protocol_value``: L4 protocol.
2617
2618* ``mask_value``: Participates in the match or not by bit for field above, 1b means participate
2619
2620* ``tcp_flags_value``: TCP control bits. The non-zero value is invalid, when the protocol_value is not set to 0x06 (TCP).
2621
2622* ``prio_value``: The priority of this filter.
2623
2624* ``queue_id``: The receive queue associated with this 5-tuple filter.
2625
2626Example, to add/remove an 5tuple filter rule::
2627
2628   testpmd> 5tuple_filter 0 add dst_ip 2.2.2.5 src_ip 2.2.2.4 \
2629            dst_port 64 src_port 32 protocol 0x06 mask 0x1F \
2630            flags 0x0 priority 3 queue 3
2631
2632   testpmd> 5tuple_filter 0 del dst_ip 2.2.2.5 src_ip 2.2.2.4 \
2633            dst_port 64 src_port 32 protocol 0x06 mask 0x1F \
2634            flags 0x0 priority 3 queue 3
2635
2636syn_filter
2637~~~~~~~~~~
2638
2639Using the  SYN filter, TCP packets whose *SYN* flag is set can be forwarded to a separate queue::
2640
2641   syn_filter (port_id) (add|del) priority (high|low) queue (queue_id)
2642
2643The available information parameters are:
2644
2645* ``port_id``: The port which the SYN filter assigned on.
2646
2647* ``high``: This SYN filter has higher priority than other filters.
2648
2649* ``low``: This SYN filter has lower priority than other filters.
2650
2651* ``queue_id``: The receive queue associated with this SYN filter
2652
2653Example::
2654
2655   testpmd> syn_filter 0 add priority high queue 3
2656
2657flex_filter
2658~~~~~~~~~~~
2659
2660With flex filter, packets can be recognized by any arbitrary pattern within the first 128 bytes of the packet
2661and routed into one of the receive queues::
2662
2663   flex_filter (port_id) (add|del) len (len_value) bytes (bytes_value) \
2664               mask (mask_value) priority (prio_value) queue (queue_id)
2665
2666The available information parameters are:
2667
2668* ``port_id``: The port which the Flex filter is assigned on.
2669
2670* ``len_value``: Filter length in bytes, no greater than 128.
2671
2672* ``bytes_value``: A string in hexadecimal, means the value the flex filter needs to match.
2673
2674* ``mask_value``: A string in hexadecimal, bit 1 means corresponding byte participates in the match.
2675
2676* ``prio_value``: The priority of this filter.
2677
2678* ``queue_id``: The receive queue associated with this Flex filter.
2679
2680Example::
2681
2682   testpmd> flex_filter 0 add len 16 bytes 0x00000000000000000000000008060000 \
2683                          mask 000C priority 3 queue 3
2684
2685   testpmd> flex_filter 0 del len 16 bytes 0x00000000000000000000000008060000 \
2686                          mask 000C priority 3 queue 3
2687
2688
2689.. _testpmd_flow_director:
2690
2691flow_director_filter
2692~~~~~~~~~~~~~~~~~~~~
2693
2694The Flow Director works in receive mode to identify specific flows or sets of flows and route them to specific queues.
2695
2696Four types of filtering are supported which are referred to as Perfect Match, Signature, Perfect-mac-vlan and
2697Perfect-tunnel filters, the match mode is set by the ``--pkt-filter-mode`` command-line parameter:
2698
2699* Perfect match filters.
2700  The hardware checks a match between the masked fields of the received packets and the programmed filters.
2701  The masked fields are for IP flow.
2702
2703* Signature filters.
2704  The hardware checks a match between a hash-based signature of the masked fields of the received packet.
2705
2706* Perfect-mac-vlan match filters.
2707  The hardware checks a match between the masked fields of the received packets and the programmed filters.
2708  The masked fields are for MAC VLAN flow.
2709
2710* Perfect-tunnel match filters.
2711  The hardware checks a match between the masked fields of the received packets and the programmed filters.
2712  The masked fields are for tunnel flow.
2713
2714* Perfect-raw-flow-type match filters.
2715  The hardware checks a match between the masked fields of the received packets and pre-loaded raw (template) packet.
2716  The masked fields are specified by input sets.
2717
2718The Flow Director filters can match the different fields for different type of packet: flow type, specific input set
2719per flow type and the flexible payload.
2720
2721The Flow Director can also mask out parts of all of these fields so that filters
2722are only applied to certain fields or parts of the fields.
2723
2724Note that for raw flow type mode the source and destination fields in the
2725raw packet buffer need to be presented in a reversed order with respect
2726to the expected received packets.
2727For example: IP source and destination addresses or TCP/UDP/SCTP
2728source and destination ports
2729
2730Different NICs may have different capabilities, command show port fdir (port_id) can be used to acquire the information.
2731
2732# Commands to add flow director filters of different flow types::
2733
2734   flow_director_filter (port_id) mode IP (add|del|update) \
2735                        flow (ipv4-other|ipv4-frag|ipv6-other|ipv6-frag) \
2736                        src (src_ip_address) dst (dst_ip_address) \
2737                        tos (tos_value) proto (proto_value) ttl (ttl_value) \
2738                        vlan (vlan_value) flexbytes (flexbytes_value) \
2739                        (drop|fwd) pf|vf(vf_id) queue (queue_id) \
2740                        fd_id (fd_id_value)
2741
2742   flow_director_filter (port_id) mode IP (add|del|update) \
2743                        flow (ipv4-tcp|ipv4-udp|ipv6-tcp|ipv6-udp) \
2744                        src (src_ip_address) (src_port) \
2745                        dst (dst_ip_address) (dst_port) \
2746                        tos (tos_value) ttl (ttl_value) \
2747                        vlan (vlan_value) flexbytes (flexbytes_value) \
2748                        (drop|fwd) queue pf|vf(vf_id) (queue_id) \
2749                        fd_id (fd_id_value)
2750
2751   flow_director_filter (port_id) mode IP (add|del|update) \
2752                        flow (ipv4-sctp|ipv6-sctp) \
2753                        src (src_ip_address) (src_port) \
2754                        dst (dst_ip_address) (dst_port) \
2755                        tos (tos_value) ttl (ttl_value) \
2756                        tag (verification_tag) vlan (vlan_value) \
2757                        flexbytes (flexbytes_value) (drop|fwd) \
2758                        pf|vf(vf_id) queue (queue_id) fd_id (fd_id_value)
2759
2760   flow_director_filter (port_id) mode IP (add|del|update) flow l2_payload \
2761                        ether (ethertype) flexbytes (flexbytes_value) \
2762                        (drop|fwd) pf|vf(vf_id) queue (queue_id)
2763                        fd_id (fd_id_value)
2764
2765   flow_director_filter (port_id) mode MAC-VLAN (add|del|update) \
2766                        mac (mac_address) vlan (vlan_value) \
2767                        flexbytes (flexbytes_value) (drop|fwd) \
2768                        queue (queue_id) fd_id (fd_id_value)
2769
2770   flow_director_filter (port_id) mode Tunnel (add|del|update) \
2771                        mac (mac_address) vlan (vlan_value) \
2772                        tunnel (NVGRE|VxLAN) tunnel-id (tunnel_id_value) \
2773                        flexbytes (flexbytes_value) (drop|fwd) \
2774                        queue (queue_id) fd_id (fd_id_value)
2775
2776   flow_director_filter (port_id) mode raw (add|del|update) flow (flow_id) \
2777                        (drop|fwd) queue (queue_id) fd_id (fd_id_value) \
2778                        packet (packet file name)
2779
2780For example, to add an ipv4-udp flow type filter::
2781
2782   testpmd> flow_director_filter 0 mode IP add flow ipv4-udp src 2.2.2.3 32 \
2783            dst 2.2.2.5 33 tos 2 ttl 40 vlan 0x1 flexbytes (0x88,0x48) \
2784            fwd pf queue 1 fd_id 1
2785
2786For example, add an ipv4-other flow type filter::
2787
2788   testpmd> flow_director_filter 0 mode IP add flow ipv4-other src 2.2.2.3 \
2789             dst 2.2.2.5 tos 2 proto 20 ttl 40 vlan 0x1 \
2790             flexbytes (0x88,0x48) fwd pf queue 1 fd_id 1
2791
2792flush_flow_director
2793~~~~~~~~~~~~~~~~~~~
2794
2795Flush all flow director filters on a device::
2796
2797   testpmd> flush_flow_director (port_id)
2798
2799Example, to flush all flow director filter on port 0::
2800
2801   testpmd> flush_flow_director 0
2802
2803flow_director_mask
2804~~~~~~~~~~~~~~~~~~
2805
2806Set flow director's input masks::
2807
2808   flow_director_mask (port_id) mode IP vlan (vlan_value) \
2809                      src_mask (ipv4_src) (ipv6_src) (src_port) \
2810                      dst_mask (ipv4_dst) (ipv6_dst) (dst_port)
2811
2812   flow_director_mask (port_id) mode MAC-VLAN vlan (vlan_value)
2813
2814   flow_director_mask (port_id) mode Tunnel vlan (vlan_value) \
2815                      mac (mac_value) tunnel-type (tunnel_type_value) \
2816                      tunnel-id (tunnel_id_value)
2817
2818Example, to set flow director mask on port 0::
2819
2820   testpmd> flow_director_mask 0 mode IP vlan 0xefff \
2821            src_mask 255.255.255.255 \
2822                FFFF:FFFF:FFFF:FFFF:FFFF:FFFF:FFFF:FFFF 0xFFFF \
2823            dst_mask 255.255.255.255 \
2824                FFFF:FFFF:FFFF:FFFF:FFFF:FFFF:FFFF:FFFF 0xFFFF
2825
2826flow_director_flex_mask
2827~~~~~~~~~~~~~~~~~~~~~~~
2828
2829set masks of flow director's flexible payload based on certain flow type::
2830
2831   testpmd> flow_director_flex_mask (port_id) \
2832            flow (none|ipv4-other|ipv4-frag|ipv4-tcp|ipv4-udp|ipv4-sctp| \
2833                  ipv6-other|ipv6-frag|ipv6-tcp|ipv6-udp|ipv6-sctp| \
2834                  l2_payload|all) (mask)
2835
2836Example, to set flow director's flex mask for all flow type on port 0::
2837
2838   testpmd> flow_director_flex_mask 0 flow all \
2839            (0xff,0xff,0,0,0,0,0,0,0,0,0,0,0,0,0,0)
2840
2841
2842flow_director_flex_payload
2843~~~~~~~~~~~~~~~~~~~~~~~~~~
2844
2845Configure flexible payload selection::
2846
2847   flow_director_flex_payload (port_id) (raw|l2|l3|l4) (config)
2848
2849For example, to select the first 16 bytes from the offset 4 (bytes) of packet's payload as flexible payload::
2850
2851   testpmd> flow_director_flex_payload 0 l4 \
2852            (4,5,6,7,8,9,10,11,12,13,14,15,16,17,18,19)
2853
2854get_sym_hash_ena_per_port
2855~~~~~~~~~~~~~~~~~~~~~~~~~
2856
2857Get symmetric hash enable configuration per port::
2858
2859   get_sym_hash_ena_per_port (port_id)
2860
2861For example, to get symmetric hash enable configuration of port 1::
2862
2863   testpmd> get_sym_hash_ena_per_port 1
2864
2865set_sym_hash_ena_per_port
2866~~~~~~~~~~~~~~~~~~~~~~~~~
2867
2868Set symmetric hash enable configuration per port to enable or disable::
2869
2870   set_sym_hash_ena_per_port (port_id) (enable|disable)
2871
2872For example, to set symmetric hash enable configuration of port 1 to enable::
2873
2874   testpmd> set_sym_hash_ena_per_port 1 enable
2875
2876get_hash_global_config
2877~~~~~~~~~~~~~~~~~~~~~~
2878
2879Get the global configurations of hash filters::
2880
2881   get_hash_global_config (port_id)
2882
2883For example, to get the global configurations of hash filters of port 1::
2884
2885   testpmd> get_hash_global_config 1
2886
2887set_hash_global_config
2888~~~~~~~~~~~~~~~~~~~~~~
2889
2890Set the global configurations of hash filters::
2891
2892   set_hash_global_config (port_id) (toeplitz|simple_xor|default) \
2893   (ipv4|ipv4-frag|ipv4-tcp|ipv4-udp|ipv4-sctp|ipv4-other|ipv6|ipv6-frag| \
2894   ipv6-tcp|ipv6-udp|ipv6-sctp|ipv6-other|l2_payload|<flow_id>) \
2895   (enable|disable)
2896
2897For example, to enable simple_xor for flow type of ipv6 on port 2::
2898
2899   testpmd> set_hash_global_config 2 simple_xor ipv6 enable
2900
2901set_hash_input_set
2902~~~~~~~~~~~~~~~~~~
2903
2904Set the input set for hash::
2905
2906   set_hash_input_set (port_id) (ipv4-frag|ipv4-tcp|ipv4-udp|ipv4-sctp| \
2907   ipv4-other|ipv6-frag|ipv6-tcp|ipv6-udp|ipv6-sctp|ipv6-other| \
2908   l2_payload|<flow_id>) (ovlan|ivlan|src-ipv4|dst-ipv4|src-ipv6|dst-ipv6| \
2909   ipv4-tos|ipv4-proto|ipv6-tc|ipv6-next-header|udp-src-port|udp-dst-port| \
2910   tcp-src-port|tcp-dst-port|sctp-src-port|sctp-dst-port|sctp-veri-tag| \
2911   udp-key|gre-key|fld-1st|fld-2nd|fld-3rd|fld-4th|fld-5th|fld-6th|fld-7th| \
2912   fld-8th|none) (select|add)
2913
2914For example, to add source IP to hash input set for flow type of ipv4-udp on port 0::
2915
2916   testpmd> set_hash_input_set 0 ipv4-udp src-ipv4 add
2917
2918set_fdir_input_set
2919~~~~~~~~~~~~~~~~~~
2920
2921The Flow Director filters can match the different fields for different type of packet, i.e. specific input set
2922on per flow type and the flexible payload. This command can be used to change input set for each flow type.
2923
2924Set the input set for flow director::
2925
2926   set_fdir_input_set (port_id) (ipv4-frag|ipv4-tcp|ipv4-udp|ipv4-sctp| \
2927   ipv4-other|ipv6|ipv6-frag|ipv6-tcp|ipv6-udp|ipv6-sctp|ipv6-other| \
2928   l2_payload|<flow_id>) (ivlan|ethertype|src-ipv4|dst-ipv4|src-ipv6|dst-ipv6| \
2929   ipv4-tos|ipv4-proto|ipv4-ttl|ipv6-tc|ipv6-next-header|ipv6-hop-limits| \
2930   tudp-src-port|udp-dst-port|cp-src-port|tcp-dst-port|sctp-src-port| \
2931   sctp-dst-port|sctp-veri-tag|none) (select|add)
2932
2933For example to add source IP to FD input set for flow type of ipv4-udp on port 0::
2934
2935   testpmd> set_fdir_input_set 0 ipv4-udp src-ipv4 add
2936
2937global_config
2938~~~~~~~~~~~~~
2939
2940Set different GRE key length for input set::
2941
2942   global_config (port_id) gre-key-len (number in bytes)
2943
2944For example to set GRE key length for input set to 4 bytes on port 0::
2945
2946   testpmd> global_config 0 gre-key-len 4
2947
2948
2949.. _testpmd_rte_flow:
2950
2951Flow rules management
2952---------------------
2953
2954Control of the generic flow API (*rte_flow*) is fully exposed through the
2955``flow`` command (validation, creation, destruction, queries and operation
2956modes).
2957
2958Considering *rte_flow* overlaps with all `Filter Functions`_, using both
2959features simultaneously may cause undefined side-effects and is therefore
2960not recommended.
2961
2962``flow`` syntax
2963~~~~~~~~~~~~~~~
2964
2965Because the ``flow`` command uses dynamic tokens to handle the large number
2966of possible flow rules combinations, its behavior differs slightly from
2967other commands, in particular:
2968
2969- Pressing *?* or the *<tab>* key displays contextual help for the current
2970  token, not that of the entire command.
2971
2972- Optional and repeated parameters are supported (provided they are listed
2973  in the contextual help).
2974
2975The first parameter stands for the operation mode. Possible operations and
2976their general syntax are described below. They are covered in detail in the
2977following sections.
2978
2979- Check whether a flow rule can be created::
2980
2981   flow validate {port_id}
2982       [group {group_id}] [priority {level}] [ingress] [egress]
2983       pattern {item} [/ {item} [...]] / end
2984       actions {action} [/ {action} [...]] / end
2985
2986- Create a flow rule::
2987
2988   flow create {port_id}
2989       [group {group_id}] [priority {level}] [ingress] [egress]
2990       pattern {item} [/ {item} [...]] / end
2991       actions {action} [/ {action} [...]] / end
2992
2993- Destroy specific flow rules::
2994
2995   flow destroy {port_id} rule {rule_id} [...]
2996
2997- Destroy all flow rules::
2998
2999   flow flush {port_id}
3000
3001- Query an existing flow rule::
3002
3003   flow query {port_id} {rule_id} {action}
3004
3005- List existing flow rules sorted by priority, filtered by group
3006  identifiers::
3007
3008   flow list {port_id} [group {group_id}] [...]
3009
3010- Restrict ingress traffic to the defined flow rules::
3011
3012   flow isolate {port_id} {boolean}
3013
3014Validating flow rules
3015~~~~~~~~~~~~~~~~~~~~~
3016
3017``flow validate`` reports whether a flow rule would be accepted by the
3018underlying device in its current state but stops short of creating it. It is
3019bound to ``rte_flow_validate()``::
3020
3021   flow validate {port_id}
3022      [group {group_id}] [priority {level}] [ingress] [egress]
3023      pattern {item} [/ {item} [...]] / end
3024      actions {action} [/ {action} [...]] / end
3025
3026If successful, it will show::
3027
3028   Flow rule validated
3029
3030Otherwise it will show an error message of the form::
3031
3032   Caught error type [...] ([...]): [...]
3033
3034This command uses the same parameters as ``flow create``, their format is
3035described in `Creating flow rules`_.
3036
3037Check whether redirecting any Ethernet packet received on port 0 to RX queue
3038index 6 is supported::
3039
3040   testpmd> flow validate 0 ingress pattern eth / end
3041      actions queue index 6 / end
3042   Flow rule validated
3043   testpmd>
3044
3045Port 0 does not support TCPv6 rules::
3046
3047   testpmd> flow validate 0 ingress pattern eth / ipv6 / tcp / end
3048      actions drop / end
3049   Caught error type 9 (specific pattern item): Invalid argument
3050   testpmd>
3051
3052Creating flow rules
3053~~~~~~~~~~~~~~~~~~~
3054
3055``flow create`` validates and creates the specified flow rule. It is bound
3056to ``rte_flow_create()``::
3057
3058   flow create {port_id}
3059      [group {group_id}] [priority {level}] [ingress] [egress]
3060      pattern {item} [/ {item} [...]] / end
3061      actions {action} [/ {action} [...]] / end
3062
3063If successful, it will return a flow rule ID usable with other commands::
3064
3065   Flow rule #[...] created
3066
3067Otherwise it will show an error message of the form::
3068
3069   Caught error type [...] ([...]): [...]
3070
3071Parameters describe in the following order:
3072
3073- Attributes (*group*, *priority*, *ingress*, *egress* tokens).
3074- A matching pattern, starting with the *pattern* token and terminated by an
3075  *end* pattern item.
3076- Actions, starting with the *actions* token and terminated by an *end*
3077  action.
3078
3079These translate directly to *rte_flow* objects provided as-is to the
3080underlying functions.
3081
3082The shortest valid definition only comprises mandatory tokens::
3083
3084   testpmd> flow create 0 pattern end actions end
3085
3086Note that PMDs may refuse rules that essentially do nothing such as this
3087one.
3088
3089**All unspecified object values are automatically initialized to 0.**
3090
3091Attributes
3092^^^^^^^^^^
3093
3094These tokens affect flow rule attributes (``struct rte_flow_attr``) and are
3095specified before the ``pattern`` token.
3096
3097- ``group {group id}``: priority group.
3098- ``priority {level}``: priority level within group.
3099- ``ingress``: rule applies to ingress traffic.
3100- ``egress``: rule applies to egress traffic.
3101
3102Each instance of an attribute specified several times overrides the previous
3103value as shown below (group 4 is used)::
3104
3105   testpmd> flow create 0 group 42 group 24 group 4 [...]
3106
3107Note that once enabled, ``ingress`` and ``egress`` cannot be disabled.
3108
3109While not specifying a direction is an error, some rules may allow both
3110simultaneously.
3111
3112Most rules affect RX therefore contain the ``ingress`` token::
3113
3114   testpmd> flow create 0 ingress pattern [...]
3115
3116Matching pattern
3117^^^^^^^^^^^^^^^^
3118
3119A matching pattern starts after the ``pattern`` token. It is made of pattern
3120items and is terminated by a mandatory ``end`` item.
3121
3122Items are named after their type (*RTE_FLOW_ITEM_TYPE_* from ``enum
3123rte_flow_item_type``).
3124
3125The ``/`` token is used as a separator between pattern items as shown
3126below::
3127
3128   testpmd> flow create 0 ingress pattern eth / ipv4 / udp / end [...]
3129
3130Note that protocol items like these must be stacked from lowest to highest
3131layer to make sense. For instance, the following rule is either invalid or
3132unlikely to match any packet::
3133
3134   testpmd> flow create 0 ingress pattern eth / udp / ipv4 / end [...]
3135
3136More information on these restrictions can be found in the *rte_flow*
3137documentation.
3138
3139Several items support additional specification structures, for example
3140``ipv4`` allows specifying source and destination addresses as follows::
3141
3142   testpmd> flow create 0 ingress pattern eth / ipv4 src is 10.1.1.1
3143      dst is 10.2.0.0 / end [...]
3144
3145This rule matches all IPv4 traffic with the specified properties.
3146
3147In this example, ``src`` and ``dst`` are field names of the underlying
3148``struct rte_flow_item_ipv4`` object. All item properties can be specified
3149in a similar fashion.
3150
3151The ``is`` token means that the subsequent value must be matched exactly,
3152and assigns ``spec`` and ``mask`` fields in ``struct rte_flow_item``
3153accordingly. Possible assignment tokens are:
3154
3155- ``is``: match value perfectly (with full bit-mask).
3156- ``spec``: match value according to configured bit-mask.
3157- ``last``: specify upper bound to establish a range.
3158- ``mask``: specify bit-mask with relevant bits set to one.
3159- ``prefix``: generate bit-mask from a prefix length.
3160
3161These yield identical results::
3162
3163   ipv4 src is 10.1.1.1
3164
3165::
3166
3167   ipv4 src spec 10.1.1.1 src mask 255.255.255.255
3168
3169::
3170
3171   ipv4 src spec 10.1.1.1 src prefix 32
3172
3173::
3174
3175   ipv4 src is 10.1.1.1 src last 10.1.1.1 # range with a single value
3176
3177::
3178
3179   ipv4 src is 10.1.1.1 src last 0 # 0 disables range
3180
3181Inclusive ranges can be defined with ``last``::
3182
3183   ipv4 src is 10.1.1.1 src last 10.2.3.4 # 10.1.1.1 to 10.2.3.4
3184
3185Note that ``mask`` affects both ``spec`` and ``last``::
3186
3187   ipv4 src is 10.1.1.1 src last 10.2.3.4 src mask 255.255.0.0
3188      # matches 10.1.0.0 to 10.2.255.255
3189
3190Properties can be modified multiple times::
3191
3192   ipv4 src is 10.1.1.1 src is 10.1.2.3 src is 10.2.3.4 # matches 10.2.3.4
3193
3194::
3195
3196   ipv4 src is 10.1.1.1 src prefix 24 src prefix 16 # matches 10.1.0.0/16
3197
3198Pattern items
3199^^^^^^^^^^^^^
3200
3201This section lists supported pattern items and their attributes, if any.
3202
3203- ``end``: end list of pattern items.
3204
3205- ``void``: no-op pattern item.
3206
3207- ``invert``: perform actions when pattern does not match.
3208
3209- ``any``: match any protocol for the current layer.
3210
3211  - ``num {unsigned}``: number of layers covered.
3212
3213- ``pf``: match packets addressed to the physical function.
3214
3215- ``vf``: match packets addressed to a virtual function ID.
3216
3217  - ``id {unsigned}``: destination VF ID.
3218
3219- ``port``: device-specific physical port index to use.
3220
3221  - ``index {unsigned}``: physical port index.
3222
3223- ``raw``: match an arbitrary byte string.
3224
3225  - ``relative {boolean}``: look for pattern after the previous item.
3226  - ``search {boolean}``: search pattern from offset (see also limit).
3227  - ``offset {integer}``: absolute or relative offset for pattern.
3228  - ``limit {unsigned}``: search area limit for start of pattern.
3229  - ``pattern {string}``: byte string to look for.
3230
3231- ``eth``: match Ethernet header.
3232
3233  - ``dst {MAC-48}``: destination MAC.
3234  - ``src {MAC-48}``: source MAC.
3235  - ``type {unsigned}``: EtherType.
3236
3237- ``vlan``: match 802.1Q/ad VLAN tag.
3238
3239  - ``tpid {unsigned}``: tag protocol identifier.
3240  - ``tci {unsigned}``: tag control information.
3241  - ``pcp {unsigned}``: priority code point.
3242  - ``dei {unsigned}``: drop eligible indicator.
3243  - ``vid {unsigned}``: VLAN identifier.
3244
3245- ``ipv4``: match IPv4 header.
3246
3247  - ``tos {unsigned}``: type of service.
3248  - ``ttl {unsigned}``: time to live.
3249  - ``proto {unsigned}``: next protocol ID.
3250  - ``src {ipv4 address}``: source address.
3251  - ``dst {ipv4 address}``: destination address.
3252
3253- ``ipv6``: match IPv6 header.
3254
3255  - ``tc {unsigned}``: traffic class.
3256  - ``flow {unsigned}``: flow label.
3257  - ``proto {unsigned}``: protocol (next header).
3258  - ``hop {unsigned}``: hop limit.
3259  - ``src {ipv6 address}``: source address.
3260  - ``dst {ipv6 address}``: destination address.
3261
3262- ``icmp``: match ICMP header.
3263
3264  - ``type {unsigned}``: ICMP packet type.
3265  - ``code {unsigned}``: ICMP packet code.
3266
3267- ``udp``: match UDP header.
3268
3269  - ``src {unsigned}``: UDP source port.
3270  - ``dst {unsigned}``: UDP destination port.
3271
3272- ``tcp``: match TCP header.
3273
3274  - ``src {unsigned}``: TCP source port.
3275  - ``dst {unsigned}``: TCP destination port.
3276
3277- ``sctp``: match SCTP header.
3278
3279  - ``src {unsigned}``: SCTP source port.
3280  - ``dst {unsigned}``: SCTP destination port.
3281  - ``tag {unsigned}``: validation tag.
3282  - ``cksum {unsigned}``: checksum.
3283
3284- ``vxlan``: match VXLAN header.
3285
3286  - ``vni {unsigned}``: VXLAN identifier.
3287
3288- ``e_tag``: match IEEE 802.1BR E-Tag header.
3289
3290  - ``grp_ecid_b {unsigned}``: GRP and E-CID base.
3291
3292- ``nvgre``: match NVGRE header.
3293
3294  - ``tni {unsigned}``: virtual subnet ID.
3295
3296- ``mpls``: match MPLS header.
3297
3298  - ``label {unsigned}``: MPLS label.
3299
3300- ``gre``: match GRE header.
3301
3302  - ``protocol {unsigned}``: protocol type.
3303
3304- ``fuzzy``: fuzzy pattern match, expect faster than default.
3305
3306  - ``thresh {unsigned}``: accuracy threshold.
3307
3308- ``gtp``, ``gtpc``, ``gtpu``: match GTPv1 header.
3309
3310  - ``teid {unsigned}``: tunnel endpoint identifier.
3311
3312- ``geneve``: match GENEVE header.
3313
3314  - ``vni {unsigned}``: virtual network identifier.
3315  - ``protocol {unsigned}``: protocol type.
3316
3317Actions list
3318^^^^^^^^^^^^
3319
3320A list of actions starts after the ``actions`` token in the same fashion as
3321`Matching pattern`_; actions are separated by ``/`` tokens and the list is
3322terminated by a mandatory ``end`` action.
3323
3324Actions are named after their type (*RTE_FLOW_ACTION_TYPE_* from ``enum
3325rte_flow_action_type``).
3326
3327Dropping all incoming UDPv4 packets can be expressed as follows::
3328
3329   testpmd> flow create 0 ingress pattern eth / ipv4 / udp / end
3330      actions drop / end
3331
3332Several actions have configurable properties which must be specified when
3333there is no valid default value. For example, ``queue`` requires a target
3334queue index.
3335
3336This rule redirects incoming UDPv4 traffic to queue index 6::
3337
3338   testpmd> flow create 0 ingress pattern eth / ipv4 / udp / end
3339      actions queue index 6 / end
3340
3341While this one could be rejected by PMDs (unspecified queue index)::
3342
3343   testpmd> flow create 0 ingress pattern eth / ipv4 / udp / end
3344      actions queue / end
3345
3346As defined by *rte_flow*, the list is not ordered, all actions of a given
3347rule are performed simultaneously. These are equivalent::
3348
3349   queue index 6 / void / mark id 42 / end
3350
3351::
3352
3353   void / mark id 42 / queue index 6 / end
3354
3355All actions in a list should have different types, otherwise only the last
3356action of a given type is taken into account::
3357
3358   queue index 4 / queue index 5 / queue index 6 / end # will use queue 6
3359
3360::
3361
3362   drop / drop / drop / end # drop is performed only once
3363
3364::
3365
3366   mark id 42 / queue index 3 / mark id 24 / end # mark will be 24
3367
3368Considering they are performed simultaneously, opposite and overlapping
3369actions can sometimes be combined when the end result is unambiguous::
3370
3371   drop / queue index 6 / end # drop has no effect
3372
3373::
3374
3375   drop / dup index 6 / end # same as above
3376
3377::
3378
3379   queue index 6 / rss queues 6 7 8 / end # queue has no effect
3380
3381::
3382
3383   drop / passthru / end # drop has no effect
3384
3385Note that PMDs may still refuse such combinations.
3386
3387Actions
3388^^^^^^^
3389
3390This section lists supported actions and their attributes, if any.
3391
3392- ``end``: end list of actions.
3393
3394- ``void``: no-op action.
3395
3396- ``passthru``: let subsequent rule process matched packets.
3397
3398- ``mark``: attach 32 bit value to packets.
3399
3400  - ``id {unsigned}``: 32 bit value to return with packets.
3401
3402- ``flag``: flag packets.
3403
3404- ``queue``: assign packets to a given queue index.
3405
3406  - ``index {unsigned}``: queue index to use.
3407
3408- ``drop``: drop packets (note: passthru has priority).
3409
3410- ``count``: enable counters for this rule.
3411
3412- ``dup``: duplicate packets to a given queue index.
3413
3414  - ``index {unsigned}``: queue index to duplicate packets to.
3415
3416- ``rss``: spread packets among several queues.
3417
3418  - ``types [{RSS hash type} [...]] end``: RSS hash types, allowed tokens
3419    are the same as `set_hash_input_set`_, an empty list means none (0).
3420
3421  - ``key {string}``: RSS hash key, overrides ``key_len``.
3422
3423  - ``key_len {unsigned}``: RSS hash key length in bytes, can be used in
3424    conjunction with ``key`` to pad or truncate it.
3425
3426  - ``queues [{unsigned} [...]] end``: queue indices to use.
3427
3428- ``pf``: redirect packets to physical device function.
3429
3430- ``vf``: redirect packets to virtual device function.
3431
3432  - ``original {boolean}``: use original VF ID if possible.
3433  - ``id {unsigned}``: VF ID to redirect packets to.
3434
3435Destroying flow rules
3436~~~~~~~~~~~~~~~~~~~~~
3437
3438``flow destroy`` destroys one or more rules from their rule ID (as returned
3439by ``flow create``), this command calls ``rte_flow_destroy()`` as many
3440times as necessary::
3441
3442   flow destroy {port_id} rule {rule_id} [...]
3443
3444If successful, it will show::
3445
3446   Flow rule #[...] destroyed
3447
3448It does not report anything for rule IDs that do not exist. The usual error
3449message is shown when a rule cannot be destroyed::
3450
3451   Caught error type [...] ([...]): [...]
3452
3453``flow flush`` destroys all rules on a device and does not take extra
3454arguments. It is bound to ``rte_flow_flush()``::
3455
3456   flow flush {port_id}
3457
3458Any errors are reported as above.
3459
3460Creating several rules and destroying them::
3461
3462   testpmd> flow create 0 ingress pattern eth / ipv6 / end
3463      actions queue index 2 / end
3464   Flow rule #0 created
3465   testpmd> flow create 0 ingress pattern eth / ipv4 / end
3466      actions queue index 3 / end
3467   Flow rule #1 created
3468   testpmd> flow destroy 0 rule 0 rule 1
3469   Flow rule #1 destroyed
3470   Flow rule #0 destroyed
3471   testpmd>
3472
3473The same result can be achieved using ``flow flush``::
3474
3475   testpmd> flow create 0 ingress pattern eth / ipv6 / end
3476      actions queue index 2 / end
3477   Flow rule #0 created
3478   testpmd> flow create 0 ingress pattern eth / ipv4 / end
3479      actions queue index 3 / end
3480   Flow rule #1 created
3481   testpmd> flow flush 0
3482   testpmd>
3483
3484Non-existent rule IDs are ignored::
3485
3486   testpmd> flow create 0 ingress pattern eth / ipv6 / end
3487      actions queue index 2 / end
3488   Flow rule #0 created
3489   testpmd> flow create 0 ingress pattern eth / ipv4 / end
3490      actions queue index 3 / end
3491   Flow rule #1 created
3492   testpmd> flow destroy 0 rule 42 rule 10 rule 2
3493   testpmd>
3494   testpmd> flow destroy 0 rule 0
3495   Flow rule #0 destroyed
3496   testpmd>
3497
3498Querying flow rules
3499~~~~~~~~~~~~~~~~~~~
3500
3501``flow query`` queries a specific action of a flow rule having that
3502ability. Such actions collect information that can be reported using this
3503command. It is bound to ``rte_flow_query()``::
3504
3505   flow query {port_id} {rule_id} {action}
3506
3507If successful, it will display either the retrieved data for known actions
3508or the following message::
3509
3510   Cannot display result for action type [...] ([...])
3511
3512Otherwise, it will complain either that the rule does not exist or that some
3513error occurred::
3514
3515   Flow rule #[...] not found
3516
3517::
3518
3519   Caught error type [...] ([...]): [...]
3520
3521Currently only the ``count`` action is supported. This action reports the
3522number of packets that hit the flow rule and the total number of bytes. Its
3523output has the following format::
3524
3525   count:
3526    hits_set: [...] # whether "hits" contains a valid value
3527    bytes_set: [...] # whether "bytes" contains a valid value
3528    hits: [...] # number of packets
3529    bytes: [...] # number of bytes
3530
3531Querying counters for TCPv6 packets redirected to queue 6::
3532
3533   testpmd> flow create 0 ingress pattern eth / ipv6 / tcp / end
3534      actions queue index 6 / count / end
3535   Flow rule #4 created
3536   testpmd> flow query 0 4 count
3537   count:
3538    hits_set: 1
3539    bytes_set: 0
3540    hits: 386446
3541    bytes: 0
3542   testpmd>
3543
3544Listing flow rules
3545~~~~~~~~~~~~~~~~~~
3546
3547``flow list`` lists existing flow rules sorted by priority and optionally
3548filtered by group identifiers::
3549
3550   flow list {port_id} [group {group_id}] [...]
3551
3552This command only fails with the following message if the device does not
3553exist::
3554
3555   Invalid port [...]
3556
3557Output consists of a header line followed by a short description of each
3558flow rule, one per line. There is no output at all when no flow rules are
3559configured on the device::
3560
3561   ID      Group   Prio    Attr    Rule
3562   [...]   [...]   [...]   [...]   [...]
3563
3564``Attr`` column flags:
3565
3566- ``i`` for ``ingress``.
3567- ``e`` for ``egress``.
3568
3569Creating several flow rules and listing them::
3570
3571   testpmd> flow create 0 ingress pattern eth / ipv4 / end
3572      actions queue index 6 / end
3573   Flow rule #0 created
3574   testpmd> flow create 0 ingress pattern eth / ipv6 / end
3575      actions queue index 2 / end
3576   Flow rule #1 created
3577   testpmd> flow create 0 priority 5 ingress pattern eth / ipv4 / udp / end
3578      actions rss queues 6 7 8 end / end
3579   Flow rule #2 created
3580   testpmd> flow list 0
3581   ID      Group   Prio    Attr    Rule
3582   0       0       0       i-      ETH IPV4 => QUEUE
3583   1       0       0       i-      ETH IPV6 => QUEUE
3584   2       0       5       i-      ETH IPV4 UDP => RSS
3585   testpmd>
3586
3587Rules are sorted by priority (i.e. group ID first, then priority level)::
3588
3589   testpmd> flow list 1
3590   ID      Group   Prio    Attr    Rule
3591   0       0       0       i-      ETH => COUNT
3592   6       0       500     i-      ETH IPV6 TCP => DROP COUNT
3593   5       0       1000    i-      ETH IPV6 ICMP => QUEUE
3594   1       24      0       i-      ETH IPV4 UDP => QUEUE
3595   4       24      10      i-      ETH IPV4 TCP => DROP
3596   3       24      20      i-      ETH IPV4 => DROP
3597   2       24      42      i-      ETH IPV4 UDP => QUEUE
3598   7       63      0       i-      ETH IPV6 UDP VXLAN => MARK QUEUE
3599   testpmd>
3600
3601Output can be limited to specific groups::
3602
3603   testpmd> flow list 1 group 0 group 63
3604   ID      Group   Prio    Attr    Rule
3605   0       0       0       i-      ETH => COUNT
3606   6       0       500     i-      ETH IPV6 TCP => DROP COUNT
3607   5       0       1000    i-      ETH IPV6 ICMP => QUEUE
3608   7       63      0       i-      ETH IPV6 UDP VXLAN => MARK QUEUE
3609   testpmd>
3610
3611Toggling isolated mode
3612~~~~~~~~~~~~~~~~~~~~~~
3613
3614``flow isolate`` can be used to tell the underlying PMD that ingress traffic
3615must only be injected from the defined flow rules; that no default traffic
3616is expected outside those rules and the driver is free to assign more
3617resources to handle them. It is bound to ``rte_flow_isolate()``::
3618
3619 flow isolate {port_id} {boolean}
3620
3621If successful, enabling or disabling isolated mode shows either::
3622
3623 Ingress traffic on port [...]
3624    is now restricted to the defined flow rules
3625
3626Or::
3627
3628 Ingress traffic on port [...]
3629    is not restricted anymore to the defined flow rules
3630
3631Otherwise, in case of error::
3632
3633   Caught error type [...] ([...]): [...]
3634
3635Mainly due to its side effects, PMDs supporting this mode may not have the
3636ability to toggle it more than once without reinitializing affected ports
3637first (e.g. by exiting testpmd).
3638
3639Enabling isolated mode::
3640
3641 testpmd> flow isolate 0 true
3642 Ingress traffic on port 0 is now restricted to the defined flow rules
3643 testpmd>
3644
3645Disabling isolated mode::
3646
3647 testpmd> flow isolate 0 false
3648 Ingress traffic on port 0 is not restricted anymore to the defined flow rules
3649 testpmd>
3650
3651Sample QinQ flow rules
3652~~~~~~~~~~~~~~~~~~~~~~
3653
3654Before creating QinQ rule(s) the following commands should be issued to enable QinQ::
3655
3656   testpmd> port stop 0
3657   testpmd> vlan set qinq on 0
3658
3659The above command sets the inner and outer TPID's to 0x8100.
3660
3661To change the TPID's the following commands should be used::
3662
3663   testpmd> vlan set outer tpid 0xa100 0
3664   testpmd> vlan set inner tpid 0x9100 0
3665   testpmd> port start 0
3666
3667Validate and create a QinQ rule on port 0 to steer traffic to a VF queue in a VM.
3668
3669::
3670
3671   testpmd> flow validate 0 ingress pattern eth / vlan tci is 123 /
3672       vlan tci is 456 / end actions vf id 1 / queue index 0 / end
3673   Flow rule #0 validated
3674
3675   testpmd> flow create 0 ingress pattern eth / vlan tci is 4 /
3676       vlan tci is 456 / end actions vf id 123 / queue index 0 / end
3677   Flow rule #0 created
3678
3679   testpmd> flow list 0
3680   ID      Group   Prio    Attr    Rule
3681   0       0       0       i-      ETH VLAN VLAN=>VF QUEUE
3682
3683Validate and create a QinQ rule on port 0 to steer traffic to a queue on the host.
3684
3685::
3686
3687   testpmd> flow validate 0 ingress pattern eth / vlan tci is 321 /
3688        vlan tci is 654 / end actions pf / queue index 0 / end
3689   Flow rule #1 validated
3690
3691   testpmd> flow create 0 ingress pattern eth / vlan tci is 321 /
3692        vlan tci is 654 / end actions pf / queue index 1 / end
3693   Flow rule #1 created
3694
3695   testpmd> flow list 0
3696   ID      Group   Prio    Attr    Rule
3697   0       0       0       i-      ETH VLAN VLAN=>VF QUEUE
3698   1       0       0       i-      ETH VLAN VLAN=>PF QUEUE
3699