xref: /dpdk/doc/guides/testpmd_app_ug/testpmd_funcs.rst (revision f4d623f96119ed720e87b16fb33985879160bfb7)
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
1633Only take effect when port is started.
1634
1635port config - speed
1636~~~~~~~~~~~~~~~~~~~
1637
1638Set the speed and duplex mode for all ports or a specific port::
1639
1640   testpmd> port config (port_id|all) speed (10|100|1000|10000|25000|40000|50000|100000|auto) \
1641            duplex (half|full|auto)
1642
1643port config - queues/descriptors
1644~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
1645
1646Set number of queues/descriptors for rxq, txq, rxd and txd::
1647
1648   testpmd> port config all (rxq|txq|rxd|txd) (value)
1649
1650This is equivalent to the ``--rxq``, ``--txq``, ``--rxd`` and ``--txd`` command-line options.
1651
1652port config - max-pkt-len
1653~~~~~~~~~~~~~~~~~~~~~~~~~
1654
1655Set the maximum packet length::
1656
1657   testpmd> port config all max-pkt-len (value)
1658
1659This is equivalent to the ``--max-pkt-len`` command-line option.
1660
1661port config - CRC Strip
1662~~~~~~~~~~~~~~~~~~~~~~~
1663
1664Set hardware CRC stripping on or off for all ports::
1665
1666   testpmd> port config all crc-strip (on|off)
1667
1668CRC stripping is on by default.
1669
1670The ``off`` option is equivalent to the ``--disable-crc-strip`` command-line option.
1671
1672port config - scatter
1673~~~~~~~~~~~~~~~~~~~~~~~
1674
1675Set RX scatter mode on or off for all ports::
1676
1677   testpmd> port config all scatter (on|off)
1678
1679RX scatter mode is off by default.
1680
1681The ``on`` option is equivalent to the ``--enable-scatter`` command-line option.
1682
1683port config - RX Checksum
1684~~~~~~~~~~~~~~~~~~~~~~~~~
1685
1686Set hardware RX checksum offload to on or off for all ports::
1687
1688   testpmd> port config all rx-cksum (on|off)
1689
1690Checksum offload is off by default.
1691
1692The ``on`` option is equivalent to the ``--enable-rx-cksum`` command-line option.
1693
1694port config - VLAN
1695~~~~~~~~~~~~~~~~~~
1696
1697Set hardware VLAN on or off for all ports::
1698
1699   testpmd> port config all hw-vlan (on|off)
1700
1701Hardware VLAN is off by default.
1702
1703The ``on`` option is equivalent to the ``--enable-hw-vlan`` command-line option.
1704
1705port config - VLAN filter
1706~~~~~~~~~~~~~~~~~~~~~~~~~
1707
1708Set hardware VLAN filter on or off for all ports::
1709
1710   testpmd> port config all hw-vlan-filter (on|off)
1711
1712Hardware VLAN filter is off by default.
1713
1714The ``on`` option is equivalent to the ``--enable-hw-vlan-filter`` command-line option.
1715
1716port config - VLAN strip
1717~~~~~~~~~~~~~~~~~~~~~~~~
1718
1719Set hardware VLAN strip on or off for all ports::
1720
1721   testpmd> port config all hw-vlan-strip (on|off)
1722
1723Hardware VLAN strip is off by default.
1724
1725The ``on`` option is equivalent to the ``--enable-hw-vlan-strip`` command-line option.
1726
1727port config - VLAN extend
1728~~~~~~~~~~~~~~~~~~~~~~~~~
1729
1730Set hardware VLAN extend on or off for all ports::
1731
1732   testpmd> port config all hw-vlan-extend (on|off)
1733
1734Hardware VLAN extend is off by default.
1735
1736The ``on`` option is equivalent to the ``--enable-hw-vlan-extend`` command-line option.
1737
1738port config - Drop Packets
1739~~~~~~~~~~~~~~~~~~~~~~~~~~
1740
1741Set packet drop for packets with no descriptors on or off for all ports::
1742
1743   testpmd> port config all drop-en (on|off)
1744
1745Packet dropping for packets with no descriptors is off by default.
1746
1747The ``on`` option is equivalent to the ``--enable-drop-en`` command-line option.
1748
1749port config - RSS
1750~~~~~~~~~~~~~~~~~
1751
1752Set the RSS (Receive Side Scaling) mode on or off::
1753
1754   testpmd> port config all rss (all|ip|tcp|udp|sctp|ether|port|vxlan|geneve|nvgre|none)
1755
1756RSS is on by default.
1757
1758The ``none`` option is equivalent to the ``--disable-rss`` command-line option.
1759
1760port config - RSS Reta
1761~~~~~~~~~~~~~~~~~~~~~~
1762
1763Set the RSS (Receive Side Scaling) redirection table::
1764
1765   testpmd> port config all rss reta (hash,queue)[,(hash,queue)]
1766
1767port config - DCB
1768~~~~~~~~~~~~~~~~~
1769
1770Set the DCB mode for an individual port::
1771
1772   testpmd> port config (port_id) dcb vt (on|off) (traffic_class) pfc (on|off)
1773
1774The traffic class should be 4 or 8.
1775
1776port config - Burst
1777~~~~~~~~~~~~~~~~~~~
1778
1779Set the number of packets per burst::
1780
1781   testpmd> port config all burst (value)
1782
1783This is equivalent to the ``--burst`` command-line option.
1784
1785port config - Threshold
1786~~~~~~~~~~~~~~~~~~~~~~~
1787
1788Set thresholds for TX/RX queues::
1789
1790   testpmd> port config all (threshold) (value)
1791
1792Where the threshold type can be:
1793
1794* ``txpt:`` Set the prefetch threshold register of the TX rings, 0 <= value <= 255.
1795
1796* ``txht:`` Set the host threshold register of the TX rings, 0 <= value <= 255.
1797
1798* ``txwt:`` Set the write-back threshold register of the TX rings, 0 <= value <= 255.
1799
1800* ``rxpt:`` Set the prefetch threshold register of the RX rings, 0 <= value <= 255.
1801
1802* ``rxht:`` Set the host threshold register of the RX rings, 0 <= value <= 255.
1803
1804* ``rxwt:`` Set the write-back threshold register of the RX rings, 0 <= value <= 255.
1805
1806* ``txfreet:`` Set the transmit free threshold of the TX rings, 0 <= value <= txd.
1807
1808* ``rxfreet:`` Set the transmit free threshold of the RX rings, 0 <= value <= rxd.
1809
1810* ``txrst:`` Set the transmit RS bit threshold of TX rings, 0 <= value <= txd.
1811
1812These threshold options are also available from the command-line.
1813
1814port config - E-tag
1815~~~~~~~~~~~~~~~~~~~
1816
1817Set the value of ether-type for E-tag::
1818
1819   testpmd> port config (port_id|all) l2-tunnel E-tag ether-type (value)
1820
1821Enable/disable the E-tag support::
1822
1823   testpmd> port config (port_id|all) l2-tunnel E-tag (enable|disable)
1824
1825port config pctype mapping
1826~~~~~~~~~~~~~~~~~~~~~~~~~~
1827
1828Reset pctype mapping table::
1829
1830   testpmd> port config (port_id) pctype mapping reset
1831
1832Update hardware defined pctype to software defined flow type mapping table::
1833
1834   testpmd> port config (port_id) pctype mapping update (pctype_id_0[,pctype_id_1]*) (flow_type_id)
1835
1836where:
1837
1838* ``pctype_id_x``: hardware pctype id as index of bit in bitmask value of the pctype mapping table.
1839
1840* ``flow_type_id``: software flow type id as the index of the pctype mapping table.
1841
1842port config input set
1843~~~~~~~~~~~~~~~~~~~~~
1844
1845Config RSS/FDIR/FDIR flexible payload input set for some pctype::
1846   testpmd> port config (port_id) pctype (pctype_id) \
1847            (hash_inset|fdir_inset|fdir_flx_inset) \
1848	    (get|set|clear) field (field_idx)
1849
1850Clear RSS/FDIR/FDIR flexible payload input set for some pctype::
1851   testpmd> port config (port_id) pctype (pctype_id) \
1852            (hash_inset|fdir_inset|fdir_flx_inset) clear all
1853
1854where:
1855
1856* ``pctype_id``: hardware packet classification types.
1857* ``field_idx``: hardware field index.
1858
1859Link Bonding Functions
1860----------------------
1861
1862The Link Bonding functions make it possible to dynamically create and
1863manage link bonding devices from within testpmd interactive prompt.
1864
1865create bonded device
1866~~~~~~~~~~~~~~~~~~~~
1867
1868Create a new bonding device::
1869
1870   testpmd> create bonded device (mode) (socket)
1871
1872For example, to create a bonded device in mode 1 on socket 0::
1873
1874   testpmd> create bonded 1 0
1875   created new bonded device (port X)
1876
1877add bonding slave
1878~~~~~~~~~~~~~~~~~
1879
1880Adds Ethernet device to a Link Bonding device::
1881
1882   testpmd> add bonding slave (slave id) (port id)
1883
1884For example, to add Ethernet device (port 6) to a Link Bonding device (port 10)::
1885
1886   testpmd> add bonding slave 6 10
1887
1888
1889remove bonding slave
1890~~~~~~~~~~~~~~~~~~~~
1891
1892Removes an Ethernet slave device from a Link Bonding device::
1893
1894   testpmd> remove bonding slave (slave id) (port id)
1895
1896For example, to remove Ethernet slave device (port 6) to a Link Bonding device (port 10)::
1897
1898   testpmd> remove bonding slave 6 10
1899
1900set bonding mode
1901~~~~~~~~~~~~~~~~
1902
1903Set the Link Bonding mode of a Link Bonding device::
1904
1905   testpmd> set bonding mode (value) (port id)
1906
1907For example, to set the bonding mode of a Link Bonding device (port 10) to broadcast (mode 3)::
1908
1909   testpmd> set bonding mode 3 10
1910
1911set bonding primary
1912~~~~~~~~~~~~~~~~~~~
1913
1914Set an Ethernet slave device as the primary device on a Link Bonding device::
1915
1916   testpmd> set bonding primary (slave id) (port id)
1917
1918For example, to set the Ethernet slave device (port 6) as the primary port of a Link Bonding device (port 10)::
1919
1920   testpmd> set bonding primary 6 10
1921
1922set bonding mac
1923~~~~~~~~~~~~~~~
1924
1925Set the MAC address of a Link Bonding device::
1926
1927   testpmd> set bonding mac (port id) (mac)
1928
1929For example, to set the MAC address of a Link Bonding device (port 10) to 00:00:00:00:00:01::
1930
1931   testpmd> set bonding mac 10 00:00:00:00:00:01
1932
1933set bonding xmit_balance_policy
1934~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
1935
1936Set the transmission policy for a Link Bonding device when it is in Balance XOR mode::
1937
1938   testpmd> set bonding xmit_balance_policy (port_id) (l2|l23|l34)
1939
1940For example, set a Link Bonding device (port 10) to use a balance policy of layer 3+4 (IP addresses & UDP ports)::
1941
1942   testpmd> set bonding xmit_balance_policy 10 l34
1943
1944
1945set bonding mon_period
1946~~~~~~~~~~~~~~~~~~~~~~
1947
1948Set the link status monitoring polling period in milliseconds for a bonding device.
1949
1950This adds support for PMD slave devices which do not support link status interrupts.
1951When the mon_period is set to a value greater than 0 then all PMD's which do not support
1952link status ISR will be queried every polling interval to check if their link status has changed::
1953
1954   testpmd> set bonding mon_period (port_id) (value)
1955
1956For example, to set the link status monitoring polling period of bonded device (port 5) to 150ms::
1957
1958   testpmd> set bonding mon_period 5 150
1959
1960
1961set bonding lacp dedicated_queue
1962~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
1963
1964Enable dedicated tx/rx queues on bonding devices slaves to handle LACP control plane traffic
1965when in mode 4 (link-aggregration-802.3ad)::
1966
1967   testpmd> set bonding lacp dedicated_queues (port_id) (enable|disable)
1968
1969
1970set bonding agg_mode
1971~~~~~~~~~~~~~~~~~~~~
1972
1973Enable one of the specific aggregators mode when in mode 4 (link-aggregration-802.3ad)::
1974
1975   testpmd> set bonding agg_mode (port_id) (bandwidth|count|stable)
1976
1977
1978show bonding config
1979~~~~~~~~~~~~~~~~~~~
1980
1981Show the current configuration of a Link Bonding device::
1982
1983   testpmd> show bonding config (port id)
1984
1985For example,
1986to show the configuration a Link Bonding device (port 9) with 3 slave devices (1, 3, 4)
1987in balance mode with a transmission policy of layer 2+3::
1988
1989   testpmd> show bonding config 9
1990        Bonding mode: 2
1991        Balance Xmit Policy: BALANCE_XMIT_POLICY_LAYER23
1992        Slaves (3): [1 3 4]
1993        Active Slaves (3): [1 3 4]
1994        Primary: [3]
1995
1996
1997Register Functions
1998------------------
1999
2000The Register Functions can be used to read from and write to registers on the network card referenced by a port number.
2001This is mainly useful for debugging purposes.
2002Reference should be made to the appropriate datasheet for the network card for details on the register addresses
2003and fields that can be accessed.
2004
2005read reg
2006~~~~~~~~
2007
2008Display the value of a port register::
2009
2010   testpmd> read reg (port_id) (address)
2011
2012For example, to examine the Flow Director control register (FDIRCTL, 0x0000EE000) on an Intel 82599 10 GbE Controller::
2013
2014   testpmd> read reg 0 0xEE00
2015   port 0 PCI register at offset 0xEE00: 0x4A060029 (1241907241)
2016
2017read regfield
2018~~~~~~~~~~~~~
2019
2020Display a port register bit field::
2021
2022   testpmd> read regfield (port_id) (address) (bit_x) (bit_y)
2023
2024For example, reading the lowest two bits from the register in the example above::
2025
2026   testpmd> read regfield 0 0xEE00 0 1
2027   port 0 PCI register at offset 0xEE00: bits[0, 1]=0x1 (1)
2028
2029read regbit
2030~~~~~~~~~~~
2031
2032Display a single port register bit::
2033
2034   testpmd> read regbit (port_id) (address) (bit_x)
2035
2036For example, reading the lowest bit from the register in the example above::
2037
2038   testpmd> read regbit 0 0xEE00 0
2039   port 0 PCI register at offset 0xEE00: bit 0=1
2040
2041write reg
2042~~~~~~~~~
2043
2044Set the value of a port register::
2045
2046   testpmd> write reg (port_id) (address) (value)
2047
2048For example, to clear a register::
2049
2050   testpmd> write reg 0 0xEE00 0x0
2051   port 0 PCI register at offset 0xEE00: 0x00000000 (0)
2052
2053write regfield
2054~~~~~~~~~~~~~~
2055
2056Set bit field of a port register::
2057
2058   testpmd> write regfield (port_id) (address) (bit_x) (bit_y) (value)
2059
2060For example, writing to the register cleared in the example above::
2061
2062   testpmd> write regfield 0 0xEE00 0 1 2
2063   port 0 PCI register at offset 0xEE00: 0x00000002 (2)
2064
2065write regbit
2066~~~~~~~~~~~~
2067
2068Set single bit value of a port register::
2069
2070   testpmd> write regbit (port_id) (address) (bit_x) (value)
2071
2072For example, to set the high bit in the register from the example above::
2073
2074   testpmd> write regbit 0 0xEE00 31 1
2075   port 0 PCI register at offset 0xEE00: 0x8000000A (2147483658)
2076
2077Traffic Metering and Policing
2078-----------------------------
2079
2080The following section shows functions for configuring traffic metering and
2081policing on the ethernet device through the use of generic ethdev API.
2082
2083show port traffic management capability
2084~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
2085
2086Show traffic metering and policing capability of the port::
2087
2088   testpmd> show port meter cap (port_id)
2089
2090add port meter profile (srTCM rfc2967)
2091~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
2092
2093Add meter profile (srTCM rfc2697) to the ethernet device::
2094
2095   testpmd> add port meter profile srtcm_rfc2697 (port_id) (profile_id) \
2096   (cir) (cbs) (ebs)
2097
2098where:
2099
2100* ``profile_id``: ID for the meter profile.
2101* ``cir``: Committed Information Rate (CIR) (bytes/second).
2102* ``cbs``: Committed Burst Size (CBS) (bytes).
2103* ``ebs``: Excess Burst Size (EBS) (bytes).
2104
2105add port meter profile (trTCM rfc2968)
2106~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
2107
2108Add meter profile (srTCM rfc2698) to the ethernet device::
2109
2110   testpmd> add port meter profile trtcm_rfc2698 (port_id) (profile_id) \
2111   (cir) (pir) (cbs) (pbs)
2112
2113where:
2114
2115* ``profile_id``: ID for the meter profile.
2116* ``cir``: Committed information rate (bytes/second).
2117* ``pir``: Peak information rate (bytes/second).
2118* ``cbs``: Committed burst size (bytes).
2119* ``pbs``: Peak burst size (bytes).
2120
2121add port meter profile (trTCM rfc4115)
2122~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
2123
2124Add meter profile (trTCM rfc4115) to the ethernet device::
2125
2126   testpmd> add port meter profile trtcm_rfc4115 (port_id) (profile_id) \
2127   (cir) (eir) (cbs) (ebs)
2128
2129where:
2130
2131* ``profile_id``: ID for the meter profile.
2132* ``cir``: Committed information rate (bytes/second).
2133* ``eir``: Excess information rate (bytes/second).
2134* ``cbs``: Committed burst size (bytes).
2135* ``ebs``: Excess burst size (bytes).
2136
2137delete port meter profile
2138~~~~~~~~~~~~~~~~~~~~~~~~~
2139
2140Delete meter profile from the ethernet device::
2141
2142   testpmd> del port meter profile (port_id) (profile_id)
2143
2144create port meter
2145~~~~~~~~~~~~~~~~~
2146
2147Create new meter object for the ethernet device::
2148
2149   testpmd> create port meter (port_id) (mtr_id) (profile_id) \
2150   (meter_enable) (g_action) (y_action) (r_action) (stats_mask) (shared) \
2151   (use_pre_meter_color) [(dscp_tbl_entry0) (dscp_tbl_entry1)...\
2152   (dscp_tbl_entry63)]
2153
2154where:
2155
2156* ``mtr_id``: meter object ID.
2157* ``profile_id``: ID for the meter profile.
2158* ``meter_enable``: When this parameter has a non-zero value, the meter object
2159  gets enabled at the time of creation, otherwise remains disabled.
2160* ``g_action``: Policer action for the packet with green color.
2161* ``y_action``: Policer action for the packet with yellow color.
2162* ``r_action``: Policer action for the packet with red color.
2163* ``stats_mask``: Mask of statistics counter types to be enabled for the
2164  meter object.
2165* ``shared``:  When this parameter has a non-zero value, the meter object is
2166  shared by multiple flows. Otherwise, meter object is used by single flow.
2167* ``use_pre_meter_color``: When this parameter has a non-zero value, the
2168  input color for the current meter object is determined by the latest meter
2169  object in the same flow. Otherwise, the current meter object uses the
2170  *dscp_table* to determine the input color.
2171* ``dscp_tbl_entryx``: DSCP table entry x providing meter providing input
2172  color, 0 <= x <= 63.
2173
2174enable port meter
2175~~~~~~~~~~~~~~~~~
2176
2177Enable meter for the ethernet device::
2178
2179   testpmd> enable port meter (port_id) (mtr_id)
2180
2181disable port meter
2182~~~~~~~~~~~~~~~~~~
2183
2184Disable meter for the ethernet device::
2185
2186   testpmd> disable port meter (port_id) (mtr_id)
2187
2188delete port meter
2189~~~~~~~~~~~~~~~~~
2190
2191Delete meter for the ethernet device::
2192
2193   testpmd> del port meter (port_id) (mtr_id)
2194
2195Set port meter profile
2196~~~~~~~~~~~~~~~~~~~~~~
2197
2198Set meter profile for the ethernet device::
2199
2200   testpmd> set port meter profile (port_id) (mtr_id) (profile_id)
2201
2202set port meter dscp table
2203~~~~~~~~~~~~~~~~~~~~~~~~~
2204
2205Set meter dscp table for the ethernet device::
2206
2207   testpmd> set port meter dscp table (port_id) (mtr_id) [(dscp_tbl_entry0) \
2208   (dscp_tbl_entry1)...(dscp_tbl_entry63)]
2209
2210set port meter policer action
2211~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
2212
2213Set meter policer action for the ethernet device::
2214
2215   testpmd> set port meter policer action (port_id) (mtr_id) (action_mask) \
2216   (action0) [(action1) (action1)]
2217
2218where:
2219
2220* ``action_mask``: Bit mask indicating which policer actions need to be
2221  updated. One or more policer actions can be updated in a single function
2222  invocation. To update the policer action associated with color C, bit
2223  (1 << C) needs to be set in *action_mask* and element at position C
2224  in the *actions* array needs to be valid.
2225* ``actionx``: Policer action for the color x,
2226  RTE_MTR_GREEN <= x < RTE_MTR_COLORS
2227
2228set port meter stats mask
2229~~~~~~~~~~~~~~~~~~~~~~~~~
2230
2231Set meter stats mask for the ethernet device::
2232
2233   testpmd> set port meter stats mask (port_id) (mtr_id) (stats_mask)
2234
2235where:
2236
2237* ``stats_mask``: Bit mask indicating statistics counter types to be enabled.
2238
2239show port meter stats
2240~~~~~~~~~~~~~~~~~~~~~
2241
2242Show meter stats of the ethernet device::
2243
2244   testpmd> show port meter stats (port_id) (mtr_id) (clear)
2245
2246where:
2247
2248* ``clear``: Flag that indicates whether the statistics counters should
2249  be cleared (i.e. set to zero) immediately after they have been read or not.
2250
2251Traffic Management
2252------------------
2253
2254The following section shows functions for configuring traffic management on
2255on the ethernet device through the use of generic TM API.
2256
2257show port traffic management capability
2258~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
2259
2260Show traffic management capability of the port::
2261
2262   testpmd> show port tm cap (port_id)
2263
2264show port traffic management capability (hierarchy level)
2265~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
2266
2267Show traffic management hierarchy level capability of the port::
2268
2269   testpmd> show port tm level cap (port_id) (level_id)
2270
2271show port traffic management capability (hierarchy node level)
2272~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
2273
2274Show the traffic management hierarchy node capability of the port::
2275
2276   testpmd> show port tm node cap (port_id) (node_id)
2277
2278show port traffic management hierarchy node type
2279~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
2280
2281Show the port traffic management hierarchy node type::
2282
2283   testpmd> show port tm node type (port_id) (node_id)
2284
2285show port traffic management hierarchy node stats
2286~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
2287
2288Show the port traffic management hierarchy node statistics::
2289
2290   testpmd> show port tm node stats (port_id) (node_id) (clear)
2291
2292where:
2293
2294* ``clear``: When this parameter has a non-zero value, the statistics counters
2295  are cleared (i.e. set to zero) immediately after they have been read,
2296  otherwise the statistics counters are left untouched.
2297
2298Add port traffic management private shaper profile
2299~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
2300
2301Add the port traffic management private shaper profile::
2302
2303   testpmd> add port tm node shaper profile (port_id) (shaper_profile_id) \
2304   (tb_rate) (tb_size) (packet_length_adjust)
2305
2306where:
2307
2308* ``shaper_profile id``: Shaper profile ID for the new profile.
2309* ``tb_rate``: Token bucket rate (bytes per second).
2310* ``tb_size``: Token bucket size (bytes).
2311* ``packet_length_adjust``: The value (bytes) to be added to the length of
2312  each packet for the purpose of shaping. This parameter value can be used to
2313  correct the packet length with the framing overhead bytes that are consumed
2314  on the wire.
2315
2316Delete port traffic management private shaper profile
2317~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
2318
2319Delete the port traffic management private shaper::
2320
2321   testpmd> del port tm node shaper profile (port_id) (shaper_profile_id)
2322
2323where:
2324
2325* ``shaper_profile id``: Shaper profile ID that needs to be deleted.
2326
2327Add port traffic management shared shaper
2328~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
2329
2330Create the port traffic management shared shaper::
2331
2332   testpmd> add port tm node shared shaper (port_id) (shared_shaper_id) \
2333   (shaper_profile_id)
2334
2335where:
2336
2337* ``shared_shaper_id``: Shared shaper ID to be created.
2338* ``shaper_profile id``: Shaper profile ID for shared shaper.
2339
2340Set port traffic management shared shaper
2341~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
2342
2343Update the port traffic management shared shaper::
2344
2345   testpmd> set port tm node shared shaper (port_id) (shared_shaper_id) \
2346   (shaper_profile_id)
2347
2348where:
2349
2350* ``shared_shaper_id``: Shared shaper ID to be update.
2351* ``shaper_profile id``: Shaper profile ID for shared shaper.
2352
2353Delete port traffic management shared shaper
2354~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
2355
2356Delete the port traffic management shared shaper::
2357
2358   testpmd> del port tm node shared shaper (port_id) (shared_shaper_id)
2359
2360where:
2361
2362* ``shared_shaper_id``: Shared shaper ID to be deleted.
2363
2364Set port traffic management hiearchy node private shaper
2365~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
2366
2367set the port traffic management hierarchy node private shaper::
2368
2369   testpmd> set port tm node shaper profile (port_id) (node_id) \
2370   (shaper_profile_id)
2371
2372where:
2373
2374* ``shaper_profile id``: Private shaper profile ID to be enabled on the
2375  hierarchy node.
2376
2377Add port traffic management WRED profile
2378~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
2379
2380Create a new WRED profile::
2381
2382   testpmd> add port tm node wred profile (port_id) (wred_profile_id) \
2383   (color_g) (min_th_g) (max_th_g) (maxp_inv_g) (wq_log2_g) \
2384   (color_y) (min_th_y) (max_th_y) (maxp_inv_y) (wq_log2_y) \
2385   (color_r) (min_th_r) (max_th_r) (maxp_inv_r) (wq_log2_r)
2386
2387where:
2388
2389* ``wred_profile id``: Identifier for the newly create WRED profile
2390* ``color_g``: Packet color (green)
2391* ``min_th_g``: Minimum queue threshold for packet with green color
2392* ``max_th_g``: Minimum queue threshold for packet with green color
2393* ``maxp_inv_g``: Inverse of packet marking probability maximum value (maxp)
2394* ``wq_log2_g``: Negated log2 of queue weight (wq)
2395* ``color_y``: Packet color (yellow)
2396* ``min_th_y``: Minimum queue threshold for packet with yellow color
2397* ``max_th_y``: Minimum queue threshold for packet with yellow color
2398* ``maxp_inv_y``: Inverse of packet marking probability maximum value (maxp)
2399* ``wq_log2_y``: Negated log2 of queue weight (wq)
2400* ``color_r``: Packet color (red)
2401* ``min_th_r``: Minimum queue threshold for packet with yellow color
2402* ``max_th_r``: Minimum queue threshold for packet with yellow color
2403* ``maxp_inv_r``: Inverse of packet marking probability maximum value (maxp)
2404* ``wq_log2_r``: Negated log2 of queue weight (wq)
2405
2406Delete port traffic management WRED profile
2407~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
2408
2409Delete the WRED profile::
2410
2411   testpmd> del port tm node wred profile (port_id) (wred_profile_id)
2412
2413Add port traffic management hierarchy nonleaf node
2414~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
2415
2416Add nonleaf node to port traffic management hiearchy::
2417
2418   testpmd> add port tm nonleaf node (port_id) (node_id) (parent_node_id) \
2419   (priority) (weight) (level_id) (shaper_profile_id) \
2420   (n_sp_priorities) (stats_mask) (n_shared_shapers) \
2421   [(shared_shaper_0) (shared_shaper_1) ...] \
2422
2423where:
2424
2425* ``parent_node_id``: Node ID of the parent.
2426* ``priority``: Node priority (highest node priority is zero). This is used by
2427  the SP algorithm running on the parent node for scheduling this node.
2428* ``weight``: Node weight (lowest weight is one). The node weight is relative
2429  to the weight sum of all siblings that have the same priority. It is used by
2430  the WFQ algorithm running on the parent node for scheduling this node.
2431* ``level_id``: Hiearchy level of the node.
2432* ``shaper_profile_id``: Shaper profile ID of the private shaper to be used by
2433  the node.
2434* ``n_sp_priorities``: Number of strict priorities.
2435* ``stats_mask``: Mask of statistics counter types to be enabled for this node.
2436* ``n_shared_shapers``: Number of shared shapers.
2437* ``shared_shaper_id``: Shared shaper id.
2438
2439Add port traffic management hierarchy leaf node
2440~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
2441
2442Add leaf node to port traffic management hiearchy::
2443
2444   testpmd> add port tm leaf node (port_id) (node_id) (parent_node_id) \
2445   (priority) (weight) (level_id) (shaper_profile_id) \
2446   (cman_mode) (wred_profile_id) (stats_mask) (n_shared_shapers) \
2447   [(shared_shaper_id) (shared_shaper_id) ...] \
2448
2449where:
2450
2451* ``parent_node_id``: Node ID of the parent.
2452* ``priority``: Node priority (highest node priority is zero). This is used by
2453  the SP algorithm running on the parent node for scheduling this node.
2454* ``weight``: Node weight (lowest weight is one). The node weight is relative
2455  to the weight sum of all siblings that have the same priority. It is used by
2456  the WFQ algorithm running on the parent node for scheduling this node.
2457* ``level_id``: Hiearchy level of the node.
2458* ``shaper_profile_id``: Shaper profile ID of the private shaper to be used by
2459  the node.
2460* ``cman_mode``: Congestion management mode to be enabled for this node.
2461* ``wred_profile_id``: WRED profile id to be enabled for this node.
2462* ``stats_mask``: Mask of statistics counter types to be enabled for this node.
2463* ``n_shared_shapers``: Number of shared shapers.
2464* ``shared_shaper_id``: Shared shaper id.
2465
2466Delete port traffic management hierarchy node
2467~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
2468
2469Delete node from port traffic management hiearchy::
2470
2471   testpmd> del port tm node (port_id) (node_id)
2472
2473Update port traffic management hierarchy parent node
2474~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
2475
2476Update port traffic management hierarchy parent node::
2477
2478   testpmd> set port tm node parent (port_id) (node_id) (parent_node_id) \
2479   (priority) (weight)
2480
2481This function can only be called after the hierarchy commit invocation. Its
2482success depends on the port support for this operation, as advertised through
2483the port capability set. This function is valid for all nodes of the traffic
2484management hierarchy except root node.
2485
2486Commit port traffic management hierarchy
2487~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
2488
2489Commit the traffic management hierarchy on the port::
2490
2491   testpmd> port tm hierarchy commit (port_id) (clean_on_fail)
2492
2493where:
2494
2495* ``clean_on_fail``: When set to non-zero, hierarchy is cleared on function
2496  call failure. On the other hand, hierarchy is preserved when this parameter
2497  is equal to zero.
2498
2499Set port traffic management default hierarchy (tm forwarding mode)
2500~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
2501
2502set the traffic management default hierarchy on the port::
2503
2504   testpmd> set port tm hierarchy default (port_id)
2505
2506Filter Functions
2507----------------
2508
2509This section details the available filter functions that are available.
2510
2511Note these functions interface the deprecated legacy filtering framework,
2512superseded by *rte_flow*. See `Flow rules management`_.
2513
2514ethertype_filter
2515~~~~~~~~~~~~~~~~~~~~
2516
2517Add or delete a L2 Ethertype filter, which identify packets by their L2 Ethertype mainly assign them to a receive queue::
2518
2519   ethertype_filter (port_id) (add|del) (mac_addr|mac_ignr) (mac_address) \
2520                    ethertype (ether_type) (drop|fwd) queue (queue_id)
2521
2522The available information parameters are:
2523
2524* ``port_id``: The port which the Ethertype filter assigned on.
2525
2526* ``mac_addr``: Compare destination mac address.
2527
2528* ``mac_ignr``: Ignore destination mac address match.
2529
2530* ``mac_address``: Destination mac address to match.
2531
2532* ``ether_type``: The EtherType value want to match,
2533  for example 0x0806 for ARP packet. 0x0800 (IPv4) and 0x86DD (IPv6) are invalid.
2534
2535* ``queue_id``: The receive queue associated with this EtherType filter.
2536  It is meaningless when deleting or dropping.
2537
2538Example, to add/remove an ethertype filter rule::
2539
2540   testpmd> ethertype_filter 0 add mac_ignr 00:11:22:33:44:55 \
2541                             ethertype 0x0806 fwd queue 3
2542
2543   testpmd> ethertype_filter 0 del mac_ignr 00:11:22:33:44:55 \
2544                             ethertype 0x0806 fwd queue 3
2545
25462tuple_filter
2547~~~~~~~~~~~~~~~~~
2548
2549Add or delete a 2-tuple filter,
2550which identifies packets by specific protocol and destination TCP/UDP port
2551and forwards packets into one of the receive queues::
2552
2553   2tuple_filter (port_id) (add|del) dst_port (dst_port_value) \
2554                 protocol (protocol_value) mask (mask_value) \
2555                 tcp_flags (tcp_flags_value) priority (prio_value) \
2556                 queue (queue_id)
2557
2558The available information parameters are:
2559
2560* ``port_id``: The port which the 2-tuple filter assigned on.
2561
2562* ``dst_port_value``: Destination port in L4.
2563
2564* ``protocol_value``: IP L4 protocol.
2565
2566* ``mask_value``: Participates in the match or not by bit for field above, 1b means participate.
2567
2568* ``tcp_flags_value``: TCP control bits. The non-zero value is invalid, when the pro_value is not set to 0x06 (TCP).
2569
2570* ``prio_value``: Priority of this filter.
2571
2572* ``queue_id``: The receive queue associated with this 2-tuple filter.
2573
2574Example, to add/remove an 2tuple filter rule::
2575
2576   testpmd> 2tuple_filter 0 add dst_port 32 protocol 0x06 mask 0x03 \
2577                          tcp_flags 0x02 priority 3 queue 3
2578
2579   testpmd> 2tuple_filter 0 del dst_port 32 protocol 0x06 mask 0x03 \
2580                          tcp_flags 0x02 priority 3 queue 3
2581
25825tuple_filter
2583~~~~~~~~~~~~~~~~~
2584
2585Add or delete a 5-tuple filter,
2586which consists of a 5-tuple (protocol, source and destination IP addresses, source and destination TCP/UDP/SCTP port)
2587and routes packets into one of the receive queues::
2588
2589   5tuple_filter (port_id) (add|del) dst_ip (dst_address) src_ip \
2590                 (src_address) dst_port (dst_port_value) \
2591                 src_port (src_port_value) protocol (protocol_value) \
2592                 mask (mask_value) tcp_flags (tcp_flags_value) \
2593                 priority (prio_value) queue (queue_id)
2594
2595The available information parameters are:
2596
2597* ``port_id``: The port which the 5-tuple filter assigned on.
2598
2599* ``dst_address``: Destination IP address.
2600
2601* ``src_address``: Source IP address.
2602
2603* ``dst_port_value``: TCP/UDP destination port.
2604
2605* ``src_port_value``: TCP/UDP source port.
2606
2607* ``protocol_value``: L4 protocol.
2608
2609* ``mask_value``: Participates in the match or not by bit for field above, 1b means participate
2610
2611* ``tcp_flags_value``: TCP control bits. The non-zero value is invalid, when the protocol_value is not set to 0x06 (TCP).
2612
2613* ``prio_value``: The priority of this filter.
2614
2615* ``queue_id``: The receive queue associated with this 5-tuple filter.
2616
2617Example, to add/remove an 5tuple filter rule::
2618
2619   testpmd> 5tuple_filter 0 add dst_ip 2.2.2.5 src_ip 2.2.2.4 \
2620            dst_port 64 src_port 32 protocol 0x06 mask 0x1F \
2621            flags 0x0 priority 3 queue 3
2622
2623   testpmd> 5tuple_filter 0 del dst_ip 2.2.2.5 src_ip 2.2.2.4 \
2624            dst_port 64 src_port 32 protocol 0x06 mask 0x1F \
2625            flags 0x0 priority 3 queue 3
2626
2627syn_filter
2628~~~~~~~~~~
2629
2630Using the  SYN filter, TCP packets whose *SYN* flag is set can be forwarded to a separate queue::
2631
2632   syn_filter (port_id) (add|del) priority (high|low) queue (queue_id)
2633
2634The available information parameters are:
2635
2636* ``port_id``: The port which the SYN filter assigned on.
2637
2638* ``high``: This SYN filter has higher priority than other filters.
2639
2640* ``low``: This SYN filter has lower priority than other filters.
2641
2642* ``queue_id``: The receive queue associated with this SYN filter
2643
2644Example::
2645
2646   testpmd> syn_filter 0 add priority high queue 3
2647
2648flex_filter
2649~~~~~~~~~~~
2650
2651With flex filter, packets can be recognized by any arbitrary pattern within the first 128 bytes of the packet
2652and routed into one of the receive queues::
2653
2654   flex_filter (port_id) (add|del) len (len_value) bytes (bytes_value) \
2655               mask (mask_value) priority (prio_value) queue (queue_id)
2656
2657The available information parameters are:
2658
2659* ``port_id``: The port which the Flex filter is assigned on.
2660
2661* ``len_value``: Filter length in bytes, no greater than 128.
2662
2663* ``bytes_value``: A string in hexadecimal, means the value the flex filter needs to match.
2664
2665* ``mask_value``: A string in hexadecimal, bit 1 means corresponding byte participates in the match.
2666
2667* ``prio_value``: The priority of this filter.
2668
2669* ``queue_id``: The receive queue associated with this Flex filter.
2670
2671Example::
2672
2673   testpmd> flex_filter 0 add len 16 bytes 0x00000000000000000000000008060000 \
2674                          mask 000C priority 3 queue 3
2675
2676   testpmd> flex_filter 0 del len 16 bytes 0x00000000000000000000000008060000 \
2677                          mask 000C priority 3 queue 3
2678
2679
2680.. _testpmd_flow_director:
2681
2682flow_director_filter
2683~~~~~~~~~~~~~~~~~~~~
2684
2685The Flow Director works in receive mode to identify specific flows or sets of flows and route them to specific queues.
2686
2687Four types of filtering are supported which are referred to as Perfect Match, Signature, Perfect-mac-vlan and
2688Perfect-tunnel filters, the match mode is set by the ``--pkt-filter-mode`` command-line parameter:
2689
2690* Perfect match filters.
2691  The hardware checks a match between the masked fields of the received packets and the programmed filters.
2692  The masked fields are for IP flow.
2693
2694* Signature filters.
2695  The hardware checks a match between a hash-based signature of the masked fields of the received packet.
2696
2697* Perfect-mac-vlan match filters.
2698  The hardware checks a match between the masked fields of the received packets and the programmed filters.
2699  The masked fields are for MAC VLAN flow.
2700
2701* Perfect-tunnel match filters.
2702  The hardware checks a match between the masked fields of the received packets and the programmed filters.
2703  The masked fields are for tunnel flow.
2704
2705* Perfect-raw-flow-type match filters.
2706  The hardware checks a match between the masked fields of the received packets and pre-loaded raw (template) packet.
2707  The masked fields are specified by input sets.
2708
2709The Flow Director filters can match the different fields for different type of packet: flow type, specific input set
2710per flow type and the flexible payload.
2711
2712The Flow Director can also mask out parts of all of these fields so that filters
2713are only applied to certain fields or parts of the fields.
2714
2715Note that for raw flow type mode the source and destination fields in the
2716raw packet buffer need to be presented in a reversed order with respect
2717to the expected received packets.
2718For example: IP source and destination addresses or TCP/UDP/SCTP
2719source and destination ports
2720
2721Different NICs may have different capabilities, command show port fdir (port_id) can be used to acquire the information.
2722
2723# Commands to add flow director filters of different flow types::
2724
2725   flow_director_filter (port_id) mode IP (add|del|update) \
2726                        flow (ipv4-other|ipv4-frag|ipv6-other|ipv6-frag) \
2727                        src (src_ip_address) dst (dst_ip_address) \
2728                        tos (tos_value) proto (proto_value) ttl (ttl_value) \
2729                        vlan (vlan_value) flexbytes (flexbytes_value) \
2730                        (drop|fwd) pf|vf(vf_id) queue (queue_id) \
2731                        fd_id (fd_id_value)
2732
2733   flow_director_filter (port_id) mode IP (add|del|update) \
2734                        flow (ipv4-tcp|ipv4-udp|ipv6-tcp|ipv6-udp) \
2735                        src (src_ip_address) (src_port) \
2736                        dst (dst_ip_address) (dst_port) \
2737                        tos (tos_value) ttl (ttl_value) \
2738                        vlan (vlan_value) flexbytes (flexbytes_value) \
2739                        (drop|fwd) queue pf|vf(vf_id) (queue_id) \
2740                        fd_id (fd_id_value)
2741
2742   flow_director_filter (port_id) mode IP (add|del|update) \
2743                        flow (ipv4-sctp|ipv6-sctp) \
2744                        src (src_ip_address) (src_port) \
2745                        dst (dst_ip_address) (dst_port) \
2746                        tos (tos_value) ttl (ttl_value) \
2747                        tag (verification_tag) vlan (vlan_value) \
2748                        flexbytes (flexbytes_value) (drop|fwd) \
2749                        pf|vf(vf_id) queue (queue_id) fd_id (fd_id_value)
2750
2751   flow_director_filter (port_id) mode IP (add|del|update) flow l2_payload \
2752                        ether (ethertype) flexbytes (flexbytes_value) \
2753                        (drop|fwd) pf|vf(vf_id) queue (queue_id)
2754                        fd_id (fd_id_value)
2755
2756   flow_director_filter (port_id) mode MAC-VLAN (add|del|update) \
2757                        mac (mac_address) vlan (vlan_value) \
2758                        flexbytes (flexbytes_value) (drop|fwd) \
2759                        queue (queue_id) fd_id (fd_id_value)
2760
2761   flow_director_filter (port_id) mode Tunnel (add|del|update) \
2762                        mac (mac_address) vlan (vlan_value) \
2763                        tunnel (NVGRE|VxLAN) tunnel-id (tunnel_id_value) \
2764                        flexbytes (flexbytes_value) (drop|fwd) \
2765                        queue (queue_id) fd_id (fd_id_value)
2766
2767   flow_director_filter (port_id) mode raw (add|del|update) flow (flow_id) \
2768                        (drop|fwd) queue (queue_id) fd_id (fd_id_value) \
2769                        packet (packet file name)
2770
2771For example, to add an ipv4-udp flow type filter::
2772
2773   testpmd> flow_director_filter 0 mode IP add flow ipv4-udp src 2.2.2.3 32 \
2774            dst 2.2.2.5 33 tos 2 ttl 40 vlan 0x1 flexbytes (0x88,0x48) \
2775            fwd pf queue 1 fd_id 1
2776
2777For example, add an ipv4-other flow type filter::
2778
2779   testpmd> flow_director_filter 0 mode IP add flow ipv4-other src 2.2.2.3 \
2780             dst 2.2.2.5 tos 2 proto 20 ttl 40 vlan 0x1 \
2781             flexbytes (0x88,0x48) fwd pf queue 1 fd_id 1
2782
2783flush_flow_director
2784~~~~~~~~~~~~~~~~~~~
2785
2786Flush all flow director filters on a device::
2787
2788   testpmd> flush_flow_director (port_id)
2789
2790Example, to flush all flow director filter on port 0::
2791
2792   testpmd> flush_flow_director 0
2793
2794flow_director_mask
2795~~~~~~~~~~~~~~~~~~
2796
2797Set flow director's input masks::
2798
2799   flow_director_mask (port_id) mode IP vlan (vlan_value) \
2800                      src_mask (ipv4_src) (ipv6_src) (src_port) \
2801                      dst_mask (ipv4_dst) (ipv6_dst) (dst_port)
2802
2803   flow_director_mask (port_id) mode MAC-VLAN vlan (vlan_value)
2804
2805   flow_director_mask (port_id) mode Tunnel vlan (vlan_value) \
2806                      mac (mac_value) tunnel-type (tunnel_type_value) \
2807                      tunnel-id (tunnel_id_value)
2808
2809Example, to set flow director mask on port 0::
2810
2811   testpmd> flow_director_mask 0 mode IP vlan 0xefff \
2812            src_mask 255.255.255.255 \
2813                FFFF:FFFF:FFFF:FFFF:FFFF:FFFF:FFFF:FFFF 0xFFFF \
2814            dst_mask 255.255.255.255 \
2815                FFFF:FFFF:FFFF:FFFF:FFFF:FFFF:FFFF:FFFF 0xFFFF
2816
2817flow_director_flex_mask
2818~~~~~~~~~~~~~~~~~~~~~~~
2819
2820set masks of flow director's flexible payload based on certain flow type::
2821
2822   testpmd> flow_director_flex_mask (port_id) \
2823            flow (none|ipv4-other|ipv4-frag|ipv4-tcp|ipv4-udp|ipv4-sctp| \
2824                  ipv6-other|ipv6-frag|ipv6-tcp|ipv6-udp|ipv6-sctp| \
2825                  l2_payload|all) (mask)
2826
2827Example, to set flow director's flex mask for all flow type on port 0::
2828
2829   testpmd> flow_director_flex_mask 0 flow all \
2830            (0xff,0xff,0,0,0,0,0,0,0,0,0,0,0,0,0,0)
2831
2832
2833flow_director_flex_payload
2834~~~~~~~~~~~~~~~~~~~~~~~~~~
2835
2836Configure flexible payload selection::
2837
2838   flow_director_flex_payload (port_id) (raw|l2|l3|l4) (config)
2839
2840For example, to select the first 16 bytes from the offset 4 (bytes) of packet's payload as flexible payload::
2841
2842   testpmd> flow_director_flex_payload 0 l4 \
2843            (4,5,6,7,8,9,10,11,12,13,14,15,16,17,18,19)
2844
2845get_sym_hash_ena_per_port
2846~~~~~~~~~~~~~~~~~~~~~~~~~
2847
2848Get symmetric hash enable configuration per port::
2849
2850   get_sym_hash_ena_per_port (port_id)
2851
2852For example, to get symmetric hash enable configuration of port 1::
2853
2854   testpmd> get_sym_hash_ena_per_port 1
2855
2856set_sym_hash_ena_per_port
2857~~~~~~~~~~~~~~~~~~~~~~~~~
2858
2859Set symmetric hash enable configuration per port to enable or disable::
2860
2861   set_sym_hash_ena_per_port (port_id) (enable|disable)
2862
2863For example, to set symmetric hash enable configuration of port 1 to enable::
2864
2865   testpmd> set_sym_hash_ena_per_port 1 enable
2866
2867get_hash_global_config
2868~~~~~~~~~~~~~~~~~~~~~~
2869
2870Get the global configurations of hash filters::
2871
2872   get_hash_global_config (port_id)
2873
2874For example, to get the global configurations of hash filters of port 1::
2875
2876   testpmd> get_hash_global_config 1
2877
2878set_hash_global_config
2879~~~~~~~~~~~~~~~~~~~~~~
2880
2881Set the global configurations of hash filters::
2882
2883   set_hash_global_config (port_id) (toeplitz|simple_xor|default) \
2884   (ipv4|ipv4-frag|ipv4-tcp|ipv4-udp|ipv4-sctp|ipv4-other|ipv6|ipv6-frag| \
2885   ipv6-tcp|ipv6-udp|ipv6-sctp|ipv6-other|l2_payload|<flow_id>) \
2886   (enable|disable)
2887
2888For example, to enable simple_xor for flow type of ipv6 on port 2::
2889
2890   testpmd> set_hash_global_config 2 simple_xor ipv6 enable
2891
2892set_hash_input_set
2893~~~~~~~~~~~~~~~~~~
2894
2895Set the input set for hash::
2896
2897   set_hash_input_set (port_id) (ipv4-frag|ipv4-tcp|ipv4-udp|ipv4-sctp| \
2898   ipv4-other|ipv6-frag|ipv6-tcp|ipv6-udp|ipv6-sctp|ipv6-other| \
2899   l2_payload|<flow_id>) (ovlan|ivlan|src-ipv4|dst-ipv4|src-ipv6|dst-ipv6| \
2900   ipv4-tos|ipv4-proto|ipv6-tc|ipv6-next-header|udp-src-port|udp-dst-port| \
2901   tcp-src-port|tcp-dst-port|sctp-src-port|sctp-dst-port|sctp-veri-tag| \
2902   udp-key|gre-key|fld-1st|fld-2nd|fld-3rd|fld-4th|fld-5th|fld-6th|fld-7th| \
2903   fld-8th|none) (select|add)
2904
2905For example, to add source IP to hash input set for flow type of ipv4-udp on port 0::
2906
2907   testpmd> set_hash_input_set 0 ipv4-udp src-ipv4 add
2908
2909set_fdir_input_set
2910~~~~~~~~~~~~~~~~~~
2911
2912The Flow Director filters can match the different fields for different type of packet, i.e. specific input set
2913on per flow type and the flexible payload. This command can be used to change input set for each flow type.
2914
2915Set the input set for flow director::
2916
2917   set_fdir_input_set (port_id) (ipv4-frag|ipv4-tcp|ipv4-udp|ipv4-sctp| \
2918   ipv4-other|ipv6|ipv6-frag|ipv6-tcp|ipv6-udp|ipv6-sctp|ipv6-other| \
2919   l2_payload|<flow_id>) (ivlan|ethertype|src-ipv4|dst-ipv4|src-ipv6|dst-ipv6| \
2920   ipv4-tos|ipv4-proto|ipv4-ttl|ipv6-tc|ipv6-next-header|ipv6-hop-limits| \
2921   tudp-src-port|udp-dst-port|cp-src-port|tcp-dst-port|sctp-src-port| \
2922   sctp-dst-port|sctp-veri-tag|none) (select|add)
2923
2924For example to add source IP to FD input set for flow type of ipv4-udp on port 0::
2925
2926   testpmd> set_fdir_input_set 0 ipv4-udp src-ipv4 add
2927
2928global_config
2929~~~~~~~~~~~~~
2930
2931Set different GRE key length for input set::
2932
2933   global_config (port_id) gre-key-len (number in bytes)
2934
2935For example to set GRE key length for input set to 4 bytes on port 0::
2936
2937   testpmd> global_config 0 gre-key-len 4
2938
2939
2940.. _testpmd_rte_flow:
2941
2942Flow rules management
2943---------------------
2944
2945Control of the generic flow API (*rte_flow*) is fully exposed through the
2946``flow`` command (validation, creation, destruction, queries and operation
2947modes).
2948
2949Considering *rte_flow* overlaps with all `Filter Functions`_, using both
2950features simultaneously may cause undefined side-effects and is therefore
2951not recommended.
2952
2953``flow`` syntax
2954~~~~~~~~~~~~~~~
2955
2956Because the ``flow`` command uses dynamic tokens to handle the large number
2957of possible flow rules combinations, its behavior differs slightly from
2958other commands, in particular:
2959
2960- Pressing *?* or the *<tab>* key displays contextual help for the current
2961  token, not that of the entire command.
2962
2963- Optional and repeated parameters are supported (provided they are listed
2964  in the contextual help).
2965
2966The first parameter stands for the operation mode. Possible operations and
2967their general syntax are described below. They are covered in detail in the
2968following sections.
2969
2970- Check whether a flow rule can be created::
2971
2972   flow validate {port_id}
2973       [group {group_id}] [priority {level}] [ingress] [egress]
2974       pattern {item} [/ {item} [...]] / end
2975       actions {action} [/ {action} [...]] / end
2976
2977- Create a flow rule::
2978
2979   flow create {port_id}
2980       [group {group_id}] [priority {level}] [ingress] [egress]
2981       pattern {item} [/ {item} [...]] / end
2982       actions {action} [/ {action} [...]] / end
2983
2984- Destroy specific flow rules::
2985
2986   flow destroy {port_id} rule {rule_id} [...]
2987
2988- Destroy all flow rules::
2989
2990   flow flush {port_id}
2991
2992- Query an existing flow rule::
2993
2994   flow query {port_id} {rule_id} {action}
2995
2996- List existing flow rules sorted by priority, filtered by group
2997  identifiers::
2998
2999   flow list {port_id} [group {group_id}] [...]
3000
3001- Restrict ingress traffic to the defined flow rules::
3002
3003   flow isolate {port_id} {boolean}
3004
3005Validating flow rules
3006~~~~~~~~~~~~~~~~~~~~~
3007
3008``flow validate`` reports whether a flow rule would be accepted by the
3009underlying device in its current state but stops short of creating it. It is
3010bound to ``rte_flow_validate()``::
3011
3012   flow validate {port_id}
3013      [group {group_id}] [priority {level}] [ingress] [egress]
3014      pattern {item} [/ {item} [...]] / end
3015      actions {action} [/ {action} [...]] / end
3016
3017If successful, it will show::
3018
3019   Flow rule validated
3020
3021Otherwise it will show an error message of the form::
3022
3023   Caught error type [...] ([...]): [...]
3024
3025This command uses the same parameters as ``flow create``, their format is
3026described in `Creating flow rules`_.
3027
3028Check whether redirecting any Ethernet packet received on port 0 to RX queue
3029index 6 is supported::
3030
3031   testpmd> flow validate 0 ingress pattern eth / end
3032      actions queue index 6 / end
3033   Flow rule validated
3034   testpmd>
3035
3036Port 0 does not support TCPv6 rules::
3037
3038   testpmd> flow validate 0 ingress pattern eth / ipv6 / tcp / end
3039      actions drop / end
3040   Caught error type 9 (specific pattern item): Invalid argument
3041   testpmd>
3042
3043Creating flow rules
3044~~~~~~~~~~~~~~~~~~~
3045
3046``flow create`` validates and creates the specified flow rule. It is bound
3047to ``rte_flow_create()``::
3048
3049   flow create {port_id}
3050      [group {group_id}] [priority {level}] [ingress] [egress]
3051      pattern {item} [/ {item} [...]] / end
3052      actions {action} [/ {action} [...]] / end
3053
3054If successful, it will return a flow rule ID usable with other commands::
3055
3056   Flow rule #[...] created
3057
3058Otherwise it will show an error message of the form::
3059
3060   Caught error type [...] ([...]): [...]
3061
3062Parameters describe in the following order:
3063
3064- Attributes (*group*, *priority*, *ingress*, *egress* tokens).
3065- A matching pattern, starting with the *pattern* token and terminated by an
3066  *end* pattern item.
3067- Actions, starting with the *actions* token and terminated by an *end*
3068  action.
3069
3070These translate directly to *rte_flow* objects provided as-is to the
3071underlying functions.
3072
3073The shortest valid definition only comprises mandatory tokens::
3074
3075   testpmd> flow create 0 pattern end actions end
3076
3077Note that PMDs may refuse rules that essentially do nothing such as this
3078one.
3079
3080**All unspecified object values are automatically initialized to 0.**
3081
3082Attributes
3083^^^^^^^^^^
3084
3085These tokens affect flow rule attributes (``struct rte_flow_attr``) and are
3086specified before the ``pattern`` token.
3087
3088- ``group {group id}``: priority group.
3089- ``priority {level}``: priority level within group.
3090- ``ingress``: rule applies to ingress traffic.
3091- ``egress``: rule applies to egress traffic.
3092
3093Each instance of an attribute specified several times overrides the previous
3094value as shown below (group 4 is used)::
3095
3096   testpmd> flow create 0 group 42 group 24 group 4 [...]
3097
3098Note that once enabled, ``ingress`` and ``egress`` cannot be disabled.
3099
3100While not specifying a direction is an error, some rules may allow both
3101simultaneously.
3102
3103Most rules affect RX therefore contain the ``ingress`` token::
3104
3105   testpmd> flow create 0 ingress pattern [...]
3106
3107Matching pattern
3108^^^^^^^^^^^^^^^^
3109
3110A matching pattern starts after the ``pattern`` token. It is made of pattern
3111items and is terminated by a mandatory ``end`` item.
3112
3113Items are named after their type (*RTE_FLOW_ITEM_TYPE_* from ``enum
3114rte_flow_item_type``).
3115
3116The ``/`` token is used as a separator between pattern items as shown
3117below::
3118
3119   testpmd> flow create 0 ingress pattern eth / ipv4 / udp / end [...]
3120
3121Note that protocol items like these must be stacked from lowest to highest
3122layer to make sense. For instance, the following rule is either invalid or
3123unlikely to match any packet::
3124
3125   testpmd> flow create 0 ingress pattern eth / udp / ipv4 / end [...]
3126
3127More information on these restrictions can be found in the *rte_flow*
3128documentation.
3129
3130Several items support additional specification structures, for example
3131``ipv4`` allows specifying source and destination addresses as follows::
3132
3133   testpmd> flow create 0 ingress pattern eth / ipv4 src is 10.1.1.1
3134      dst is 10.2.0.0 / end [...]
3135
3136This rule matches all IPv4 traffic with the specified properties.
3137
3138In this example, ``src`` and ``dst`` are field names of the underlying
3139``struct rte_flow_item_ipv4`` object. All item properties can be specified
3140in a similar fashion.
3141
3142The ``is`` token means that the subsequent value must be matched exactly,
3143and assigns ``spec`` and ``mask`` fields in ``struct rte_flow_item``
3144accordingly. Possible assignment tokens are:
3145
3146- ``is``: match value perfectly (with full bit-mask).
3147- ``spec``: match value according to configured bit-mask.
3148- ``last``: specify upper bound to establish a range.
3149- ``mask``: specify bit-mask with relevant bits set to one.
3150- ``prefix``: generate bit-mask from a prefix length.
3151
3152These yield identical results::
3153
3154   ipv4 src is 10.1.1.1
3155
3156::
3157
3158   ipv4 src spec 10.1.1.1 src mask 255.255.255.255
3159
3160::
3161
3162   ipv4 src spec 10.1.1.1 src prefix 32
3163
3164::
3165
3166   ipv4 src is 10.1.1.1 src last 10.1.1.1 # range with a single value
3167
3168::
3169
3170   ipv4 src is 10.1.1.1 src last 0 # 0 disables range
3171
3172Inclusive ranges can be defined with ``last``::
3173
3174   ipv4 src is 10.1.1.1 src last 10.2.3.4 # 10.1.1.1 to 10.2.3.4
3175
3176Note that ``mask`` affects both ``spec`` and ``last``::
3177
3178   ipv4 src is 10.1.1.1 src last 10.2.3.4 src mask 255.255.0.0
3179      # matches 10.1.0.0 to 10.2.255.255
3180
3181Properties can be modified multiple times::
3182
3183   ipv4 src is 10.1.1.1 src is 10.1.2.3 src is 10.2.3.4 # matches 10.2.3.4
3184
3185::
3186
3187   ipv4 src is 10.1.1.1 src prefix 24 src prefix 16 # matches 10.1.0.0/16
3188
3189Pattern items
3190^^^^^^^^^^^^^
3191
3192This section lists supported pattern items and their attributes, if any.
3193
3194- ``end``: end list of pattern items.
3195
3196- ``void``: no-op pattern item.
3197
3198- ``invert``: perform actions when pattern does not match.
3199
3200- ``any``: match any protocol for the current layer.
3201
3202  - ``num {unsigned}``: number of layers covered.
3203
3204- ``pf``: match packets addressed to the physical function.
3205
3206- ``vf``: match packets addressed to a virtual function ID.
3207
3208  - ``id {unsigned}``: destination VF ID.
3209
3210- ``port``: device-specific physical port index to use.
3211
3212  - ``index {unsigned}``: physical port index.
3213
3214- ``raw``: match an arbitrary byte string.
3215
3216  - ``relative {boolean}``: look for pattern after the previous item.
3217  - ``search {boolean}``: search pattern from offset (see also limit).
3218  - ``offset {integer}``: absolute or relative offset for pattern.
3219  - ``limit {unsigned}``: search area limit for start of pattern.
3220  - ``pattern {string}``: byte string to look for.
3221
3222- ``eth``: match Ethernet header.
3223
3224  - ``dst {MAC-48}``: destination MAC.
3225  - ``src {MAC-48}``: source MAC.
3226  - ``type {unsigned}``: EtherType.
3227
3228- ``vlan``: match 802.1Q/ad VLAN tag.
3229
3230  - ``tpid {unsigned}``: tag protocol identifier.
3231  - ``tci {unsigned}``: tag control information.
3232  - ``pcp {unsigned}``: priority code point.
3233  - ``dei {unsigned}``: drop eligible indicator.
3234  - ``vid {unsigned}``: VLAN identifier.
3235
3236- ``ipv4``: match IPv4 header.
3237
3238  - ``tos {unsigned}``: type of service.
3239  - ``ttl {unsigned}``: time to live.
3240  - ``proto {unsigned}``: next protocol ID.
3241  - ``src {ipv4 address}``: source address.
3242  - ``dst {ipv4 address}``: destination address.
3243
3244- ``ipv6``: match IPv6 header.
3245
3246  - ``tc {unsigned}``: traffic class.
3247  - ``flow {unsigned}``: flow label.
3248  - ``proto {unsigned}``: protocol (next header).
3249  - ``hop {unsigned}``: hop limit.
3250  - ``src {ipv6 address}``: source address.
3251  - ``dst {ipv6 address}``: destination address.
3252
3253- ``icmp``: match ICMP header.
3254
3255  - ``type {unsigned}``: ICMP packet type.
3256  - ``code {unsigned}``: ICMP packet code.
3257
3258- ``udp``: match UDP header.
3259
3260  - ``src {unsigned}``: UDP source port.
3261  - ``dst {unsigned}``: UDP destination port.
3262
3263- ``tcp``: match TCP header.
3264
3265  - ``src {unsigned}``: TCP source port.
3266  - ``dst {unsigned}``: TCP destination port.
3267
3268- ``sctp``: match SCTP header.
3269
3270  - ``src {unsigned}``: SCTP source port.
3271  - ``dst {unsigned}``: SCTP destination port.
3272  - ``tag {unsigned}``: validation tag.
3273  - ``cksum {unsigned}``: checksum.
3274
3275- ``vxlan``: match VXLAN header.
3276
3277  - ``vni {unsigned}``: VXLAN identifier.
3278
3279- ``e_tag``: match IEEE 802.1BR E-Tag header.
3280
3281  - ``grp_ecid_b {unsigned}``: GRP and E-CID base.
3282
3283- ``nvgre``: match NVGRE header.
3284
3285  - ``tni {unsigned}``: virtual subnet ID.
3286
3287- ``mpls``: match MPLS header.
3288
3289  - ``label {unsigned}``: MPLS label.
3290
3291- ``gre``: match GRE header.
3292
3293  - ``protocol {unsigned}``: protocol type.
3294
3295- ``fuzzy``: fuzzy pattern match, expect faster than default.
3296
3297  - ``thresh {unsigned}``: accuracy threshold.
3298
3299- ``gtp``, ``gtpc``, ``gtpu``: match GTPv1 header.
3300
3301  - ``teid {unsigned}``: tunnel endpoint identifier.
3302
3303- ``geneve``: match GENEVE header.
3304
3305  - ``vni {unsigned}``: virtual network identifier.
3306  - ``protocol {unsigned}``: protocol type.
3307
3308Actions list
3309^^^^^^^^^^^^
3310
3311A list of actions starts after the ``actions`` token in the same fashion as
3312`Matching pattern`_; actions are separated by ``/`` tokens and the list is
3313terminated by a mandatory ``end`` action.
3314
3315Actions are named after their type (*RTE_FLOW_ACTION_TYPE_* from ``enum
3316rte_flow_action_type``).
3317
3318Dropping all incoming UDPv4 packets can be expressed as follows::
3319
3320   testpmd> flow create 0 ingress pattern eth / ipv4 / udp / end
3321      actions drop / end
3322
3323Several actions have configurable properties which must be specified when
3324there is no valid default value. For example, ``queue`` requires a target
3325queue index.
3326
3327This rule redirects incoming UDPv4 traffic to queue index 6::
3328
3329   testpmd> flow create 0 ingress pattern eth / ipv4 / udp / end
3330      actions queue index 6 / end
3331
3332While this one could be rejected by PMDs (unspecified queue index)::
3333
3334   testpmd> flow create 0 ingress pattern eth / ipv4 / udp / end
3335      actions queue / end
3336
3337As defined by *rte_flow*, the list is not ordered, all actions of a given
3338rule are performed simultaneously. These are equivalent::
3339
3340   queue index 6 / void / mark id 42 / end
3341
3342::
3343
3344   void / mark id 42 / queue index 6 / end
3345
3346All actions in a list should have different types, otherwise only the last
3347action of a given type is taken into account::
3348
3349   queue index 4 / queue index 5 / queue index 6 / end # will use queue 6
3350
3351::
3352
3353   drop / drop / drop / end # drop is performed only once
3354
3355::
3356
3357   mark id 42 / queue index 3 / mark id 24 / end # mark will be 24
3358
3359Considering they are performed simultaneously, opposite and overlapping
3360actions can sometimes be combined when the end result is unambiguous::
3361
3362   drop / queue index 6 / end # drop has no effect
3363
3364::
3365
3366   drop / dup index 6 / end # same as above
3367
3368::
3369
3370   queue index 6 / rss queues 6 7 8 / end # queue has no effect
3371
3372::
3373
3374   drop / passthru / end # drop has no effect
3375
3376Note that PMDs may still refuse such combinations.
3377
3378Actions
3379^^^^^^^
3380
3381This section lists supported actions and their attributes, if any.
3382
3383- ``end``: end list of actions.
3384
3385- ``void``: no-op action.
3386
3387- ``passthru``: let subsequent rule process matched packets.
3388
3389- ``mark``: attach 32 bit value to packets.
3390
3391  - ``id {unsigned}``: 32 bit value to return with packets.
3392
3393- ``flag``: flag packets.
3394
3395- ``queue``: assign packets to a given queue index.
3396
3397  - ``index {unsigned}``: queue index to use.
3398
3399- ``drop``: drop packets (note: passthru has priority).
3400
3401- ``count``: enable counters for this rule.
3402
3403- ``dup``: duplicate packets to a given queue index.
3404
3405  - ``index {unsigned}``: queue index to duplicate packets to.
3406
3407- ``rss``: spread packets among several queues.
3408
3409  - ``types [{RSS hash type} [...]] end``: RSS hash types, allowed tokens
3410    are the same as `set_hash_input_set`_, an empty list means none (0).
3411
3412  - ``key {string}``: RSS hash key, overrides ``key_len``.
3413
3414  - ``key_len {unsigned}``: RSS hash key length in bytes, can be used in
3415    conjunction with ``key`` to pad or truncate it.
3416
3417  - ``queues [{unsigned} [...]] end``: queue indices to use.
3418
3419- ``pf``: redirect packets to physical device function.
3420
3421- ``vf``: redirect packets to virtual device function.
3422
3423  - ``original {boolean}``: use original VF ID if possible.
3424  - ``id {unsigned}``: VF ID to redirect packets to.
3425
3426Destroying flow rules
3427~~~~~~~~~~~~~~~~~~~~~
3428
3429``flow destroy`` destroys one or more rules from their rule ID (as returned
3430by ``flow create``), this command calls ``rte_flow_destroy()`` as many
3431times as necessary::
3432
3433   flow destroy {port_id} rule {rule_id} [...]
3434
3435If successful, it will show::
3436
3437   Flow rule #[...] destroyed
3438
3439It does not report anything for rule IDs that do not exist. The usual error
3440message is shown when a rule cannot be destroyed::
3441
3442   Caught error type [...] ([...]): [...]
3443
3444``flow flush`` destroys all rules on a device and does not take extra
3445arguments. It is bound to ``rte_flow_flush()``::
3446
3447   flow flush {port_id}
3448
3449Any errors are reported as above.
3450
3451Creating several rules and destroying them::
3452
3453   testpmd> flow create 0 ingress pattern eth / ipv6 / end
3454      actions queue index 2 / end
3455   Flow rule #0 created
3456   testpmd> flow create 0 ingress pattern eth / ipv4 / end
3457      actions queue index 3 / end
3458   Flow rule #1 created
3459   testpmd> flow destroy 0 rule 0 rule 1
3460   Flow rule #1 destroyed
3461   Flow rule #0 destroyed
3462   testpmd>
3463
3464The same result can be achieved using ``flow flush``::
3465
3466   testpmd> flow create 0 ingress pattern eth / ipv6 / end
3467      actions queue index 2 / end
3468   Flow rule #0 created
3469   testpmd> flow create 0 ingress pattern eth / ipv4 / end
3470      actions queue index 3 / end
3471   Flow rule #1 created
3472   testpmd> flow flush 0
3473   testpmd>
3474
3475Non-existent rule IDs are ignored::
3476
3477   testpmd> flow create 0 ingress pattern eth / ipv6 / end
3478      actions queue index 2 / end
3479   Flow rule #0 created
3480   testpmd> flow create 0 ingress pattern eth / ipv4 / end
3481      actions queue index 3 / end
3482   Flow rule #1 created
3483   testpmd> flow destroy 0 rule 42 rule 10 rule 2
3484   testpmd>
3485   testpmd> flow destroy 0 rule 0
3486   Flow rule #0 destroyed
3487   testpmd>
3488
3489Querying flow rules
3490~~~~~~~~~~~~~~~~~~~
3491
3492``flow query`` queries a specific action of a flow rule having that
3493ability. Such actions collect information that can be reported using this
3494command. It is bound to ``rte_flow_query()``::
3495
3496   flow query {port_id} {rule_id} {action}
3497
3498If successful, it will display either the retrieved data for known actions
3499or the following message::
3500
3501   Cannot display result for action type [...] ([...])
3502
3503Otherwise, it will complain either that the rule does not exist or that some
3504error occurred::
3505
3506   Flow rule #[...] not found
3507
3508::
3509
3510   Caught error type [...] ([...]): [...]
3511
3512Currently only the ``count`` action is supported. This action reports the
3513number of packets that hit the flow rule and the total number of bytes. Its
3514output has the following format::
3515
3516   count:
3517    hits_set: [...] # whether "hits" contains a valid value
3518    bytes_set: [...] # whether "bytes" contains a valid value
3519    hits: [...] # number of packets
3520    bytes: [...] # number of bytes
3521
3522Querying counters for TCPv6 packets redirected to queue 6::
3523
3524   testpmd> flow create 0 ingress pattern eth / ipv6 / tcp / end
3525      actions queue index 6 / count / end
3526   Flow rule #4 created
3527   testpmd> flow query 0 4 count
3528   count:
3529    hits_set: 1
3530    bytes_set: 0
3531    hits: 386446
3532    bytes: 0
3533   testpmd>
3534
3535Listing flow rules
3536~~~~~~~~~~~~~~~~~~
3537
3538``flow list`` lists existing flow rules sorted by priority and optionally
3539filtered by group identifiers::
3540
3541   flow list {port_id} [group {group_id}] [...]
3542
3543This command only fails with the following message if the device does not
3544exist::
3545
3546   Invalid port [...]
3547
3548Output consists of a header line followed by a short description of each
3549flow rule, one per line. There is no output at all when no flow rules are
3550configured on the device::
3551
3552   ID      Group   Prio    Attr    Rule
3553   [...]   [...]   [...]   [...]   [...]
3554
3555``Attr`` column flags:
3556
3557- ``i`` for ``ingress``.
3558- ``e`` for ``egress``.
3559
3560Creating several flow rules and listing them::
3561
3562   testpmd> flow create 0 ingress pattern eth / ipv4 / end
3563      actions queue index 6 / end
3564   Flow rule #0 created
3565   testpmd> flow create 0 ingress pattern eth / ipv6 / end
3566      actions queue index 2 / end
3567   Flow rule #1 created
3568   testpmd> flow create 0 priority 5 ingress pattern eth / ipv4 / udp / end
3569      actions rss queues 6 7 8 end / end
3570   Flow rule #2 created
3571   testpmd> flow list 0
3572   ID      Group   Prio    Attr    Rule
3573   0       0       0       i-      ETH IPV4 => QUEUE
3574   1       0       0       i-      ETH IPV6 => QUEUE
3575   2       0       5       i-      ETH IPV4 UDP => RSS
3576   testpmd>
3577
3578Rules are sorted by priority (i.e. group ID first, then priority level)::
3579
3580   testpmd> flow list 1
3581   ID      Group   Prio    Attr    Rule
3582   0       0       0       i-      ETH => COUNT
3583   6       0       500     i-      ETH IPV6 TCP => DROP COUNT
3584   5       0       1000    i-      ETH IPV6 ICMP => QUEUE
3585   1       24      0       i-      ETH IPV4 UDP => QUEUE
3586   4       24      10      i-      ETH IPV4 TCP => DROP
3587   3       24      20      i-      ETH IPV4 => DROP
3588   2       24      42      i-      ETH IPV4 UDP => QUEUE
3589   7       63      0       i-      ETH IPV6 UDP VXLAN => MARK QUEUE
3590   testpmd>
3591
3592Output can be limited to specific groups::
3593
3594   testpmd> flow list 1 group 0 group 63
3595   ID      Group   Prio    Attr    Rule
3596   0       0       0       i-      ETH => COUNT
3597   6       0       500     i-      ETH IPV6 TCP => DROP COUNT
3598   5       0       1000    i-      ETH IPV6 ICMP => QUEUE
3599   7       63      0       i-      ETH IPV6 UDP VXLAN => MARK QUEUE
3600   testpmd>
3601
3602Toggling isolated mode
3603~~~~~~~~~~~~~~~~~~~~~~
3604
3605``flow isolate`` can be used to tell the underlying PMD that ingress traffic
3606must only be injected from the defined flow rules; that no default traffic
3607is expected outside those rules and the driver is free to assign more
3608resources to handle them. It is bound to ``rte_flow_isolate()``::
3609
3610 flow isolate {port_id} {boolean}
3611
3612If successful, enabling or disabling isolated mode shows either::
3613
3614 Ingress traffic on port [...]
3615    is now restricted to the defined flow rules
3616
3617Or::
3618
3619 Ingress traffic on port [...]
3620    is not restricted anymore to the defined flow rules
3621
3622Otherwise, in case of error::
3623
3624   Caught error type [...] ([...]): [...]
3625
3626Mainly due to its side effects, PMDs supporting this mode may not have the
3627ability to toggle it more than once without reinitializing affected ports
3628first (e.g. by exiting testpmd).
3629
3630Enabling isolated mode::
3631
3632 testpmd> flow isolate 0 true
3633 Ingress traffic on port 0 is now restricted to the defined flow rules
3634 testpmd>
3635
3636Disabling isolated mode::
3637
3638 testpmd> flow isolate 0 false
3639 Ingress traffic on port 0 is not restricted anymore to the defined flow rules
3640 testpmd>
3641
3642Sample QinQ flow rules
3643~~~~~~~~~~~~~~~~~~~~~~
3644
3645Before creating QinQ rule(s) the following commands should be issued to enable QinQ::
3646
3647   testpmd> port stop 0
3648   testpmd> vlan set qinq on 0
3649
3650The above command sets the inner and outer TPID's to 0x8100.
3651
3652To change the TPID's the following commands should be used::
3653
3654   testpmd> vlan set outer tpid 0xa100 0
3655   testpmd> vlan set inner tpid 0x9100 0
3656   testpmd> port start 0
3657
3658Validate and create a QinQ rule on port 0 to steer traffic to a VF queue in a VM.
3659
3660::
3661
3662   testpmd> flow validate 0 ingress pattern eth / vlan tci is 123 /
3663       vlan tci is 456 / end actions vf id 1 / queue index 0 / end
3664   Flow rule #0 validated
3665
3666   testpmd> flow create 0 ingress pattern eth / vlan tci is 4 /
3667       vlan tci is 456 / end actions vf id 123 / queue index 0 / end
3668   Flow rule #0 created
3669
3670   testpmd> flow list 0
3671   ID      Group   Prio    Attr    Rule
3672   0       0       0       i-      ETH VLAN VLAN=>VF QUEUE
3673
3674Validate and create a QinQ rule on port 0 to steer traffic to a queue on the host.
3675
3676::
3677
3678   testpmd> flow validate 0 ingress pattern eth / vlan tci is 321 /
3679        vlan tci is 654 / end actions pf / queue index 0 / end
3680   Flow rule #1 validated
3681
3682   testpmd> flow create 0 ingress pattern eth / vlan tci is 321 /
3683        vlan tci is 654 / end actions pf / queue index 1 / end
3684   Flow rule #1 created
3685
3686   testpmd> flow list 0
3687   ID      Group   Prio    Attr    Rule
3688   0       0       0       i-      ETH VLAN VLAN=>VF QUEUE
3689   1       0       0       i-      ETH VLAN VLAN=>PF QUEUE
3690