xref: /dpdk/doc/guides/testpmd_app_ug/testpmd_funcs.rst (revision 2717246ecd7d27125a346a2c5c55b53a9c251a93)
1..  SPDX-License-Identifier: BSD-3-Clause
2    Copyright(c) 2010-2016 Intel Corporation.
3
4.. _testpmd_runtime:
5
6Testpmd Runtime Functions
7=========================
8
9Where the testpmd application is started in interactive mode, (``-i|--interactive``),
10it displays a prompt that can be used to start and stop forwarding,
11configure the application, display statistics (including the extended NIC
12statistics aka xstats) , set the Flow Director and other tasks::
13
14   testpmd>
15
16The testpmd prompt has some, limited, readline support.
17Common bash command-line functions such as ``Ctrl+a`` and ``Ctrl+e`` to go to the start and end of the prompt line are supported
18as well as access to the command history via the up-arrow.
19
20There is also support for tab completion.
21If you type a partial command and hit ``<TAB>`` you get a list of the available completions:
22
23.. code-block:: console
24
25   testpmd> show port <TAB>
26
27       info [Mul-choice STRING]: show|clear port info|stats|xstats|fdir|stat_qmap|dcb_tc|cap X
28       info [Mul-choice STRING]: show|clear port info|stats|xstats|fdir|stat_qmap|dcb_tc|cap all
29       stats [Mul-choice STRING]: show|clear port info|stats|xstats|fdir|stat_qmap|dcb_tc|cap X
30       stats [Mul-choice STRING]: show|clear port info|stats|xstats|fdir|stat_qmap|dcb_tc|cap all
31       ...
32
33
34.. note::
35
36   Some examples in this document are too long to fit on one line are are shown wrapped at `"\\"` for display purposes::
37
38      testpmd> set flow_ctrl rx (on|off) tx (on|off) (high_water) (low_water) \
39               (pause_time) (send_xon) (port_id)
40
41In the real ``testpmd>`` prompt these commands should be on a single line.
42
43Help Functions
44--------------
45
46The testpmd has on-line help for the functions that are available at runtime.
47These are divided into sections and can be accessed using help, help section or help all:
48
49.. code-block:: console
50
51   testpmd> help
52
53       help control    : Start and stop forwarding.
54       help display    : Displaying port, stats and config information.
55       help config     : Configuration information.
56       help ports      : Configuring ports.
57       help registers  : Reading and setting port registers.
58       help filters    : Filters configuration help.
59       help all        : All of the above sections.
60
61
62Command File Functions
63----------------------
64
65To facilitate loading large number of commands or to avoid cutting and pasting where not
66practical or possible testpmd supports alternative methods for executing commands.
67
68* If started with the ``--cmdline-file=FILENAME`` command line argument testpmd
69  will execute all CLI commands contained within the file immediately before
70  starting packet forwarding or entering interactive mode.
71
72.. code-block:: console
73
74   ./testpmd -n4 -r2 ... -- -i --cmdline-file=/home/ubuntu/flow-create-commands.txt
75   Interactive-mode selected
76   CLI commands to be read from /home/ubuntu/flow-create-commands.txt
77   Configuring Port 0 (socket 0)
78   Port 0: 7C:FE:90:CB:74:CE
79   Configuring Port 1 (socket 0)
80   Port 1: 7C:FE:90:CB:74:CA
81   Checking link statuses...
82   Port 0 Link Up - speed 10000 Mbps - full-duplex
83   Port 1 Link Up - speed 10000 Mbps - full-duplex
84   Done
85   Flow rule #0 created
86   Flow rule #1 created
87   ...
88   ...
89   Flow rule #498 created
90   Flow rule #499 created
91   Read all CLI commands from /home/ubuntu/flow-create-commands.txt
92   testpmd>
93
94
95* At run-time additional commands can be loaded in bulk by invoking the ``load FILENAME``
96  command.
97
98.. code-block:: console
99
100   testpmd> load /home/ubuntu/flow-create-commands.txt
101   Flow rule #0 created
102   Flow rule #1 created
103   ...
104   ...
105   Flow rule #498 created
106   Flow rule #499 created
107   Read all CLI commands from /home/ubuntu/flow-create-commands.txt
108   testpmd>
109
110
111In all cases output from any included command will be displayed as standard output.
112Execution will continue until the end of the file is reached regardless of
113whether any errors occur.  The end user must examine the output to determine if
114any failures occurred.
115
116
117Control Functions
118-----------------
119
120start
121~~~~~
122
123Start packet forwarding with current configuration::
124
125   testpmd> start
126
127start tx_first
128~~~~~~~~~~~~~~
129
130Start packet forwarding with current configuration after sending specified number of bursts of packets::
131
132   testpmd> start tx_first (""|burst_num)
133
134The default burst number is 1 when ``burst_num`` not presented.
135
136stop
137~~~~
138
139Stop packet forwarding, and display accumulated statistics::
140
141   testpmd> stop
142
143quit
144~~~~
145
146Quit to prompt::
147
148   testpmd> quit
149
150
151Display Functions
152-----------------
153
154The functions in the following sections are used to display information about the
155testpmd configuration or the NIC status.
156
157show port
158~~~~~~~~~
159
160Display information for a given port or all ports::
161
162   testpmd> show port (info|stats|xstats|fdir|stat_qmap|dcb_tc|cap) (port_id|all)
163
164The available information categories are:
165
166* ``info``: General port information such as MAC address.
167
168* ``stats``: RX/TX statistics.
169
170* ``xstats``: RX/TX extended NIC statistics.
171
172* ``fdir``: Flow Director information and statistics.
173
174* ``stat_qmap``: Queue statistics mapping.
175
176* ``dcb_tc``: DCB information such as TC mapping.
177
178* ``cap``: Supported offload capabilities.
179
180For example:
181
182.. code-block:: console
183
184   testpmd> show port info 0
185
186   ********************* Infos for port 0 *********************
187
188   MAC address: XX:XX:XX:XX:XX:XX
189   Connect to socket: 0
190   memory allocation on the socket: 0
191   Link status: up
192   Link speed: 40000 Mbps
193   Link duplex: full-duplex
194   Promiscuous mode: enabled
195   Allmulticast mode: disabled
196   Maximum number of MAC addresses: 64
197   Maximum number of MAC addresses of hash filtering: 0
198   VLAN offload:
199       strip on
200       filter on
201       qinq(extend) off
202   Redirection table size: 512
203   Supported flow types:
204     ipv4-frag
205     ipv4-tcp
206     ipv4-udp
207     ipv4-sctp
208     ipv4-other
209     ipv6-frag
210     ipv6-tcp
211     ipv6-udp
212     ipv6-sctp
213     ipv6-other
214     l2_payload
215     port
216     vxlan
217     geneve
218     nvgre
219
220show port rss reta
221~~~~~~~~~~~~~~~~~~
222
223Display the rss redirection table entry indicated by masks on port X::
224
225   testpmd> show port (port_id) rss reta (size) (mask0, mask1...)
226
227size is used to indicate the hardware supported reta size
228
229show port rss-hash
230~~~~~~~~~~~~~~~~~~
231
232Display the RSS hash functions and RSS hash key of a port::
233
234   testpmd> show port (port_id) rss-hash ipv4|ipv4-frag|ipv4-tcp|ipv4-udp|ipv4-sctp|ipv4-other|ipv6|ipv6-frag|ipv6-tcp|ipv6-udp|ipv6-sctp|ipv6-other|l2-payload|ipv6-ex|ipv6-tcp-ex|ipv6-udp-ex [key]
235
236clear port
237~~~~~~~~~~
238
239Clear the port statistics for a given port or for all ports::
240
241   testpmd> clear port (info|stats|xstats|fdir|stat_qmap) (port_id|all)
242
243For example::
244
245   testpmd> clear port stats all
246
247show (rxq|txq)
248~~~~~~~~~~~~~~
249
250Display information for a given port's RX/TX queue::
251
252   testpmd> show (rxq|txq) info (port_id) (queue_id)
253
254show config
255~~~~~~~~~~~
256
257Displays the configuration of the application.
258The configuration comes from the command-line, the runtime or the application defaults::
259
260   testpmd> show config (rxtx|cores|fwd|txpkts)
261
262The available information categories are:
263
264* ``rxtx``: RX/TX configuration items.
265
266* ``cores``: List of forwarding cores.
267
268* ``fwd``: Packet forwarding configuration.
269
270* ``txpkts``: Packets to TX configuration.
271
272For example:
273
274.. code-block:: console
275
276   testpmd> show config rxtx
277
278   io packet forwarding - CRC stripping disabled - packets/burst=16
279   nb forwarding cores=2 - nb forwarding ports=1
280   RX queues=1 - RX desc=128 - RX free threshold=0
281   RX threshold registers: pthresh=8 hthresh=8 wthresh=4
282   TX queues=1 - TX desc=512 - TX free threshold=0
283   TX threshold registers: pthresh=36 hthresh=0 wthresh=0
284   TX RS bit threshold=0 - TXQ flags=0x0
285
286set fwd
287~~~~~~~
288
289Set the packet forwarding mode::
290
291   testpmd> set fwd (io|mac|macswap|flowgen| \
292                     rxonly|txonly|csum|icmpecho) (""|retry)
293
294``retry`` can be specified for forwarding engines except ``rx_only``.
295
296The available information categories are:
297
298* ``io``: Forwards packets "as-is" in I/O mode.
299  This is the fastest possible forwarding operation as it does not access packets data.
300  This is the default mode.
301
302* ``mac``: Changes the source and the destination Ethernet addresses of packets before forwarding them.
303  Default application behaviour is to set source Ethernet address to that of the transmitting interface, and destination
304  address to a dummy value (set during init). The user may specify a target destination Ethernet address via the 'eth-peer' or
305  'eth-peer-configfile' command-line options. It is not currently possible to specify a specific source Ethernet address.
306
307* ``macswap``: MAC swap forwarding mode.
308  Swaps the source and the destination Ethernet addresses of packets before forwarding them.
309
310* ``flowgen``: Multi-flow generation mode.
311  Originates a number of flows (with varying destination IP addresses), and terminate receive traffic.
312
313* ``rxonly``: Receives packets but doesn't transmit them.
314
315* ``txonly``: Generates and transmits packets without receiving any.
316
317* ``csum``: Changes the checksum field with hardware or software methods depending on the offload flags on the packet.
318
319* ``icmpecho``: Receives a burst of packets, lookup for IMCP echo requests and, if any, send back ICMP echo replies.
320
321* ``ieee1588``: Demonstrate L2 IEEE1588 V2 PTP timestamping for RX and TX. Requires ``CONFIG_RTE_LIBRTE_IEEE1588=y``.
322
323* ``tm``: Traffic Management forwarding mode
324  Demonstrates the use of ethdev traffic management APIs and softnic PMD for
325  QoS traffic management. In this mode, 5-level hierarchical QoS scheduler is
326  available as an default option that can be enabled through CLI. The user can
327  also modify the default hierarchy or specify the new hierarchy through CLI for
328  implementing QoS scheduler.  Requires ``CONFIG_RTE_LIBRTE_PMD_SOFTNIC=y`` ``CONFIG_RTE_LIBRTE_SCHED=y``.
329
330Example::
331
332   testpmd> set fwd rxonly
333
334   Set rxonly packet forwarding mode
335
336
337read rxd
338~~~~~~~~
339
340Display an RX descriptor for a port RX queue::
341
342   testpmd> read rxd (port_id) (queue_id) (rxd_id)
343
344For example::
345
346   testpmd> read rxd 0 0 4
347        0x0000000B - 0x001D0180 / 0x0000000B - 0x001D0180
348
349read txd
350~~~~~~~~
351
352Display a TX descriptor for a port TX queue::
353
354   testpmd> read txd (port_id) (queue_id) (txd_id)
355
356For example::
357
358   testpmd> read txd 0 0 4
359        0x00000001 - 0x24C3C440 / 0x000F0000 - 0x2330003C
360
361ddp get list
362~~~~~~~~~~~~
363
364Get loaded dynamic device personalization (DDP) package info list::
365
366   testpmd> ddp get list (port_id)
367
368ddp get info
369~~~~~~~~~~~~
370
371Display information about dynamic device personalization (DDP) profile::
372
373   testpmd> ddp get info (profile_path)
374
375show vf stats
376~~~~~~~~~~~~~
377
378Display VF statistics::
379
380   testpmd> show vf stats (port_id) (vf_id)
381
382clear vf stats
383~~~~~~~~~~~~~~
384
385Reset VF statistics::
386
387   testpmd> clear vf stats (port_id) (vf_id)
388
389show port pctype mapping
390~~~~~~~~~~~~~~~~~~~~~~~~
391
392List all items from the pctype mapping table::
393
394   testpmd> show port (port_id) pctype mapping
395
396show rx offloading capabilities
397~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
398
399List all per queue and per port Rx offloading capabilities of a port::
400
401   testpmd> show port (port_id) rx_offload capabilities
402
403show rx offloading configuration
404~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
405
406List port level and all queue level Rx offloading configuration::
407
408   testpmd> show port (port_id) rx_offload configuration
409
410show tx offloading capabilities
411~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
412
413List all per queue and per port Tx offloading capabilities of a port::
414
415   testpmd> show port (port_id) tx_offload capabilities
416
417show tx offloading configuration
418~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
419
420List port level and all queue level Tx offloading configuration::
421
422   testpmd> show port (port_id) tx_offload configuration
423
424
425Configuration Functions
426-----------------------
427
428The testpmd application can be configured from the runtime as well as from the command-line.
429
430This section details the available configuration functions that are available.
431
432.. note::
433
434   Configuration changes only become active when forwarding is started/restarted.
435
436set default
437~~~~~~~~~~~
438
439Reset forwarding to the default configuration::
440
441   testpmd> set default
442
443set verbose
444~~~~~~~~~~~
445
446Set the debug verbosity level::
447
448   testpmd> set verbose (level)
449
450Currently the only available levels are 0 (silent except for error) and 1 (fully verbose).
451
452set log
453~~~~~~~
454
455Set the log level for a log type::
456
457	testpmd> set log global|(type) (level)
458
459Where:
460
461* ``type`` is the log name.
462
463* ``level`` is the log level.
464
465For example, to change the global log level::
466	testpmd> set log global (level)
467
468Regexes can also be used for type. To change log level of user1, user2 and user3::
469	testpmd> set log user[1-3] (level)
470
471set nbport
472~~~~~~~~~~
473
474Set the number of ports used by the application:
475
476set nbport (num)
477
478This is equivalent to the ``--nb-ports`` command-line option.
479
480set nbcore
481~~~~~~~~~~
482
483Set the number of cores used by the application::
484
485   testpmd> set nbcore (num)
486
487This is equivalent to the ``--nb-cores`` command-line option.
488
489.. note::
490
491   The number of cores used must not be greater than number of ports used multiplied by the number of queues per port.
492
493set coremask
494~~~~~~~~~~~~
495
496Set the forwarding cores hexadecimal mask::
497
498   testpmd> set coremask (mask)
499
500This is equivalent to the ``--coremask`` command-line option.
501
502.. note::
503
504   The master lcore is reserved for command line parsing only and cannot be masked on for packet forwarding.
505
506set portmask
507~~~~~~~~~~~~
508
509Set the forwarding ports hexadecimal mask::
510
511   testpmd> set portmask (mask)
512
513This is equivalent to the ``--portmask`` command-line option.
514
515set burst
516~~~~~~~~~
517
518Set number of packets per burst::
519
520   testpmd> set burst (num)
521
522This is equivalent to the ``--burst command-line`` option.
523
524When retry is enabled, the transmit delay time and number of retries can also be set::
525
526   testpmd> set burst tx delay (microseconds) retry (num)
527
528set txpkts
529~~~~~~~~~~
530
531Set the length of each segment of the TX-ONLY packets or length of packet for FLOWGEN mode::
532
533   testpmd> set txpkts (x[,y]*)
534
535Where x[,y]* represents a CSV list of values, without white space.
536
537set txsplit
538~~~~~~~~~~~
539
540Set the split policy for the TX packets, applicable for TX-ONLY and CSUM forwarding modes::
541
542   testpmd> set txsplit (off|on|rand)
543
544Where:
545
546* ``off`` disable packet copy & split for CSUM mode.
547
548* ``on`` split outgoing packet into multiple segments. Size of each segment
549  and number of segments per packet is determined by ``set txpkts`` command
550  (see above).
551
552* ``rand`` same as 'on', but number of segments per each packet is a random value between 1 and total number of segments.
553
554set corelist
555~~~~~~~~~~~~
556
557Set the list of forwarding cores::
558
559   testpmd> set corelist (x[,y]*)
560
561For example, to change the forwarding cores:
562
563.. code-block:: console
564
565   testpmd> set corelist 3,1
566   testpmd> show config fwd
567
568   io packet forwarding - ports=2 - cores=2 - streams=2 - NUMA support disabled
569   Logical Core 3 (socket 0) forwards packets on 1 streams:
570   RX P=0/Q=0 (socket 0) -> TX P=1/Q=0 (socket 0) peer=02:00:00:00:00:01
571   Logical Core 1 (socket 0) forwards packets on 1 streams:
572   RX P=1/Q=0 (socket 0) -> TX P=0/Q=0 (socket 0) peer=02:00:00:00:00:00
573
574.. note::
575
576   The cores are used in the same order as specified on the command line.
577
578set portlist
579~~~~~~~~~~~~
580
581Set the list of forwarding ports::
582
583   testpmd> set portlist (x[,y]*)
584
585For example, to change the port forwarding:
586
587.. code-block:: console
588
589   testpmd> set portlist 0,2,1,3
590   testpmd> show config fwd
591
592   io packet forwarding - ports=4 - cores=1 - streams=4
593   Logical Core 3 (socket 0) forwards packets on 4 streams:
594   RX P=0/Q=0 (socket 0) -> TX P=2/Q=0 (socket 0) peer=02:00:00:00:00:01
595   RX P=2/Q=0 (socket 0) -> TX P=0/Q=0 (socket 0) peer=02:00:00:00:00:00
596   RX P=1/Q=0 (socket 0) -> TX P=3/Q=0 (socket 0) peer=02:00:00:00:00:03
597   RX P=3/Q=0 (socket 0) -> TX P=1/Q=0 (socket 0) peer=02:00:00:00:00:02
598
599set tx loopback
600~~~~~~~~~~~~~~~
601
602Enable/disable tx loopback::
603
604   testpmd> set tx loopback (port_id) (on|off)
605
606set drop enable
607~~~~~~~~~~~~~~~
608
609set drop enable bit for all queues::
610
611   testpmd> set all queues drop (port_id) (on|off)
612
613set split drop enable (for VF)
614~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
615
616set split drop enable bit for VF from PF::
617
618   testpmd> set vf split drop (port_id) (vf_id) (on|off)
619
620set mac antispoof (for VF)
621~~~~~~~~~~~~~~~~~~~~~~~~~~
622
623Set mac antispoof for a VF from the PF::
624
625   testpmd> set vf mac antispoof  (port_id) (vf_id) (on|off)
626
627set macsec offload
628~~~~~~~~~~~~~~~~~~
629
630Enable/disable MACsec offload::
631
632   testpmd> set macsec offload (port_id) on encrypt (on|off) replay-protect (on|off)
633   testpmd> set macsec offload (port_id) off
634
635set macsec sc
636~~~~~~~~~~~~~
637
638Configure MACsec secure connection (SC)::
639
640   testpmd> set macsec sc (tx|rx) (port_id) (mac) (pi)
641
642.. note::
643
644   The pi argument is ignored for tx.
645   Check the NIC Datasheet for hardware limits.
646
647set macsec sa
648~~~~~~~~~~~~~
649
650Configure MACsec secure association (SA)::
651
652   testpmd> set macsec sa (tx|rx) (port_id) (idx) (an) (pn) (key)
653
654.. note::
655
656   The IDX value must be 0 or 1.
657   Check the NIC Datasheet for hardware limits.
658
659set broadcast mode (for VF)
660~~~~~~~~~~~~~~~~~~~~~~~~~~~
661
662Set broadcast mode for a VF from the PF::
663
664   testpmd> set vf broadcast (port_id) (vf_id) (on|off)
665
666vlan set strip
667~~~~~~~~~~~~~~
668
669Set the VLAN strip on a port::
670
671   testpmd> vlan set strip (on|off) (port_id)
672
673vlan set stripq
674~~~~~~~~~~~~~~~
675
676Set the VLAN strip for a queue on a port::
677
678   testpmd> vlan set stripq (on|off) (port_id,queue_id)
679
680vlan set stripq (for VF)
681~~~~~~~~~~~~~~~~~~~~~~~~
682
683Set VLAN strip for all queues in a pool for a VF from the PF::
684
685   testpmd> set vf vlan stripq (port_id) (vf_id) (on|off)
686
687vlan set insert (for VF)
688~~~~~~~~~~~~~~~~~~~~~~~~
689
690Set VLAN insert for a VF from the PF::
691
692   testpmd> set vf vlan insert (port_id) (vf_id) (vlan_id)
693
694vlan set tag (for VF)
695~~~~~~~~~~~~~~~~~~~~~
696
697Set VLAN tag for a VF from the PF::
698
699   testpmd> set vf vlan tag (port_id) (vf_id) (on|off)
700
701vlan set antispoof (for VF)
702~~~~~~~~~~~~~~~~~~~~~~~~~~~
703
704Set VLAN antispoof for a VF from the PF::
705
706   testpmd> set vf vlan antispoof (port_id) (vf_id) (on|off)
707
708vlan set filter
709~~~~~~~~~~~~~~~
710
711Set the VLAN filter on a port::
712
713   testpmd> vlan set filter (on|off) (port_id)
714
715vlan set qinq
716~~~~~~~~~~~~~
717
718Set the VLAN QinQ (extended queue in queue) on for a port::
719
720   testpmd> vlan set qinq (on|off) (port_id)
721
722vlan set tpid
723~~~~~~~~~~~~~
724
725Set the inner or outer VLAN TPID for packet filtering on a port::
726
727   testpmd> vlan set (inner|outer) tpid (value) (port_id)
728
729.. note::
730
731   TPID value must be a 16-bit number (value <= 65536).
732
733rx_vlan add
734~~~~~~~~~~~
735
736Add a VLAN ID, or all identifiers, to the set of VLAN identifiers filtered by port ID::
737
738   testpmd> rx_vlan add (vlan_id|all) (port_id)
739
740.. note::
741
742   VLAN filter must be set on that port. VLAN ID < 4096.
743   Depending on the NIC used, number of vlan_ids may be limited to the maximum entries
744   in VFTA table. This is important if enabling all vlan_ids.
745
746rx_vlan rm
747~~~~~~~~~~
748
749Remove a VLAN ID, or all identifiers, from the set of VLAN identifiers filtered by port ID::
750
751   testpmd> rx_vlan rm (vlan_id|all) (port_id)
752
753rx_vlan add (for VF)
754~~~~~~~~~~~~~~~~~~~~
755
756Add a VLAN ID, to the set of VLAN identifiers filtered for VF(s) for port ID::
757
758   testpmd> rx_vlan add (vlan_id) port (port_id) vf (vf_mask)
759
760rx_vlan rm (for VF)
761~~~~~~~~~~~~~~~~~~~
762
763Remove a VLAN ID, from the set of VLAN identifiers filtered for VF(s) for port ID::
764
765   testpmd> rx_vlan rm (vlan_id) port (port_id) vf (vf_mask)
766
767tunnel_filter add
768~~~~~~~~~~~~~~~~~
769
770Add a tunnel filter on a port::
771
772   testpmd> tunnel_filter add (port_id) (outer_mac) (inner_mac) (ip_addr) \
773            (inner_vlan) (vxlan|nvgre|ipingre) (imac-ivlan|imac-ivlan-tenid|\
774            imac-tenid|imac|omac-imac-tenid|oip|iip) (tenant_id) (queue_id)
775
776The available information categories are:
777
778* ``vxlan``: Set tunnel type as VXLAN.
779
780* ``nvgre``: Set tunnel type as NVGRE.
781
782* ``ipingre``: Set tunnel type as IP-in-GRE.
783
784* ``imac-ivlan``: Set filter type as Inner MAC and VLAN.
785
786* ``imac-ivlan-tenid``: Set filter type as Inner MAC, VLAN and tenant ID.
787
788* ``imac-tenid``: Set filter type as Inner MAC and tenant ID.
789
790* ``imac``: Set filter type as Inner MAC.
791
792* ``omac-imac-tenid``: Set filter type as Outer MAC, Inner MAC and tenant ID.
793
794* ``oip``: Set filter type as Outer IP.
795
796* ``iip``: Set filter type as Inner IP.
797
798Example::
799
800   testpmd> tunnel_filter add 0 68:05:CA:28:09:82 00:00:00:00:00:00 \
801            192.168.2.2 0 ipingre oip 1 1
802
803   Set an IP-in-GRE tunnel on port 0, and the filter type is Outer IP.
804
805tunnel_filter remove
806~~~~~~~~~~~~~~~~~~~~
807
808Remove a tunnel filter on a port::
809
810   testpmd> tunnel_filter rm (port_id) (outer_mac) (inner_mac) (ip_addr) \
811            (inner_vlan) (vxlan|nvgre|ipingre) (imac-ivlan|imac-ivlan-tenid|\
812            imac-tenid|imac|omac-imac-tenid|oip|iip) (tenant_id) (queue_id)
813
814rx_vxlan_port add
815~~~~~~~~~~~~~~~~~
816
817Add an UDP port for VXLAN packet filter on a port::
818
819   testpmd> rx_vxlan_port add (udp_port) (port_id)
820
821rx_vxlan_port remove
822~~~~~~~~~~~~~~~~~~~~
823
824Remove an UDP port for VXLAN packet filter on a port::
825
826   testpmd> rx_vxlan_port rm (udp_port) (port_id)
827
828tx_vlan set
829~~~~~~~~~~~
830
831Set hardware insertion of VLAN IDs in packets sent on a port::
832
833   testpmd> tx_vlan set (port_id) vlan_id[, vlan_id_outer]
834
835For example, set a single VLAN ID (5) insertion on port 0::
836
837   tx_vlan set 0 5
838
839Or, set double VLAN ID (inner: 2, outer: 3) insertion on port 1::
840
841   tx_vlan set 1 2 3
842
843
844tx_vlan set pvid
845~~~~~~~~~~~~~~~~
846
847Set port based hardware insertion of VLAN ID in packets sent on a port::
848
849   testpmd> tx_vlan set pvid (port_id) (vlan_id) (on|off)
850
851tx_vlan reset
852~~~~~~~~~~~~~
853
854Disable hardware insertion of a VLAN header in packets sent on a port::
855
856   testpmd> tx_vlan reset (port_id)
857
858csum set
859~~~~~~~~
860
861Select hardware or software calculation of the checksum when
862transmitting a packet using the ``csum`` forwarding engine::
863
864   testpmd> csum set (ip|udp|tcp|sctp|outer-ip) (hw|sw) (port_id)
865
866Where:
867
868* ``ip|udp|tcp|sctp`` always relate to  the inner layer.
869
870* ``outer-ip`` relates to the outer IP layer (only for IPv4) in the case where the packet is recognized
871  as a tunnel packet by the forwarding engine (vxlan, gre and ipip are
872  supported). See also the ``csum parse-tunnel`` command.
873
874.. note::
875
876   Check the NIC Datasheet for hardware limits.
877
878RSS queue region
879~~~~~~~~~~~~~~~~
880
881Set RSS queue region span on a port::
882
883   testpmd> set port (port_id) queue-region region_id (value) \
884		queue_start_index (value) queue_num (value)
885
886Set flowtype mapping on a RSS queue region on a port::
887
888   testpmd> set port (port_id) queue-region region_id (value) flowtype (value)
889
890where:
891
892* For the flowtype(pctype) of packet,the specific index for each type has
893  been defined in file i40e_type.h as enum i40e_filter_pctype.
894
895Set user priority mapping on a RSS queue region on a port::
896
897   testpmd> set port (port_id) queue-region UP (value) region_id (value)
898
899Flush all queue region related configuration on a port::
900
901   testpmd> set port (port_id) queue-region flush (on|off)
902
903where:
904
905* "on"is just an enable function which server for other configuration,
906  it is for all configuration about queue region from up layer,
907  at first will only keep in DPDK softwarestored in driver,
908  only after "flush on", it commit all configuration to HW.
909  "off" is just clean all configuration about queue region just now,
910  and restore all to DPDK i40e driver default config when start up.
911
912Show all queue region related configuration info on a port::
913
914   testpmd> show port (port_id) queue-region
915
916.. note::
917
918  Queue region only support on PF by now, so these command is
919  only for configuration of queue region on PF port.
920
921csum parse-tunnel
922~~~~~~~~~~~~~~~~~
923
924Define how tunneled packets should be handled by the csum forward
925engine::
926
927   testpmd> csum parse-tunnel (on|off) (tx_port_id)
928
929If enabled, the csum forward engine will try to recognize supported
930tunnel headers (vxlan, gre, ipip).
931
932If disabled, treat tunnel packets as non-tunneled packets (a inner
933header is handled as a packet payload).
934
935.. note::
936
937   The port argument is the TX port like in the ``csum set`` command.
938
939Example:
940
941Consider a packet in packet like the following::
942
943   eth_out/ipv4_out/udp_out/vxlan/eth_in/ipv4_in/tcp_in
944
945* If parse-tunnel is enabled, the ``ip|udp|tcp|sctp`` parameters of ``csum set``
946  command relate to the inner headers (here ``ipv4_in`` and ``tcp_in``), and the
947  ``outer-ip parameter`` relates to the outer headers (here ``ipv4_out``).
948
949* If parse-tunnel is disabled, the ``ip|udp|tcp|sctp`` parameters of ``csum  set``
950   command relate to the outer headers, here ``ipv4_out`` and ``udp_out``.
951
952csum show
953~~~~~~~~~
954
955Display tx checksum offload configuration::
956
957   testpmd> csum show (port_id)
958
959tso set
960~~~~~~~
961
962Enable TCP Segmentation Offload (TSO) in the ``csum`` forwarding engine::
963
964   testpmd> tso set (segsize) (port_id)
965
966.. note::
967
968   Check the NIC datasheet for hardware limits.
969
970tso show
971~~~~~~~~
972
973Display the status of TCP Segmentation Offload::
974
975   testpmd> tso show (port_id)
976
977set port - gro
978~~~~~~~~~~~~~~
979
980Enable or disable GRO in ``csum`` forwarding engine::
981
982   testpmd> set port <port_id> gro on|off
983
984If enabled, the csum forwarding engine will perform GRO on the TCP/IPv4
985packets received from the given port.
986
987If disabled, packets received from the given port won't be performed
988GRO. By default, GRO is disabled for all ports.
989
990.. note::
991
992   When enable GRO for a port, TCP/IPv4 packets received from the port
993   will be performed GRO. After GRO, all merged packets have bad
994   checksums, since the GRO library doesn't re-calculate checksums for
995   the merged packets. Therefore, if users want the merged packets to
996   have correct checksums, please select HW IP checksum calculation and
997   HW TCP checksum calculation for the port which the merged packets are
998   transmitted to.
999
1000show port - gro
1001~~~~~~~~~~~~~~~
1002
1003Display GRO configuration for a given port::
1004
1005   testpmd> show port <port_id> gro
1006
1007set gro flush
1008~~~~~~~~~~~~~
1009
1010Set the cycle to flush the GROed packets from reassembly tables::
1011
1012   testpmd> set gro flush <cycles>
1013
1014When enable GRO, the csum forwarding engine performs GRO on received
1015packets, and the GROed packets are stored in reassembly tables. Users
1016can use this command to determine when the GROed packets are flushed
1017from the reassembly tables.
1018
1019The ``cycles`` is measured in GRO operation times. The csum forwarding
1020engine flushes the GROed packets from the tables every ``cycles`` GRO
1021operations.
1022
1023By default, the value of ``cycles`` is 1, which means flush GROed packets
1024from the reassembly tables as soon as one GRO operation finishes. The value
1025of ``cycles`` should be in the range of 1 to ``GRO_MAX_FLUSH_CYCLES``.
1026
1027Please note that the large value of ``cycles`` may cause the poor TCP/IP
1028stack performance. Because the GROed packets are delayed to arrive the
1029stack, thus causing more duplicated ACKs and TCP retransmissions.
1030
1031set port - gso
1032~~~~~~~~~~~~~~
1033
1034Toggle per-port GSO support in ``csum`` forwarding engine::
1035
1036   testpmd> set port <port_id> gso on|off
1037
1038If enabled, the csum forwarding engine will perform GSO on supported IPv4
1039packets, transmitted on the given port.
1040
1041If disabled, packets transmitted on the given port will not undergo GSO.
1042By default, GSO is disabled for all ports.
1043
1044.. note::
1045
1046   When GSO is enabled on a port, supported IPv4 packets transmitted on that
1047   port undergo GSO. Afterwards, the segmented packets are represented by
1048   multi-segment mbufs; however, the csum forwarding engine doesn't calculation
1049   of checksums for GSO'd segments in SW. As a result, if users want correct
1050   checksums in GSO segments, they should enable HW checksum calculation for
1051   GSO-enabled ports.
1052
1053   For example, HW checksum calculation for VxLAN GSO'd packets may be enabled
1054   by setting the following options in the csum forwarding engine:
1055
1056   testpmd> csum set outer_ip hw <port_id>
1057
1058   testpmd> csum set ip hw <port_id>
1059
1060   testpmd> csum set tcp hw <port_id>
1061
1062set gso segsz
1063~~~~~~~~~~~~~
1064
1065Set the maximum GSO segment size (measured in bytes), which includes the
1066packet header and the packet payload for GSO-enabled ports (global)::
1067
1068   testpmd> set gso segsz <length>
1069
1070show port - gso
1071~~~~~~~~~~~~~~~
1072
1073Display the status of Generic Segmentation Offload for a given port::
1074
1075   testpmd> show port <port_id> gso
1076
1077mac_addr add
1078~~~~~~~~~~~~
1079
1080Add an alternative MAC address to a port::
1081
1082   testpmd> mac_addr add (port_id) (XX:XX:XX:XX:XX:XX)
1083
1084mac_addr remove
1085~~~~~~~~~~~~~~~
1086
1087Remove a MAC address from a port::
1088
1089   testpmd> mac_addr remove (port_id) (XX:XX:XX:XX:XX:XX)
1090
1091mac_addr add (for VF)
1092~~~~~~~~~~~~~~~~~~~~~
1093
1094Add an alternative MAC address for a VF to a port::
1095
1096   testpmd> mac_add add port (port_id) vf (vf_id) (XX:XX:XX:XX:XX:XX)
1097
1098mac_addr set
1099~~~~~~~~~~~~
1100
1101Set the default MAC address for a port::
1102
1103   testpmd> mac_addr set (port_id) (XX:XX:XX:XX:XX:XX)
1104
1105mac_addr set (for VF)
1106~~~~~~~~~~~~~~~~~~~~~
1107
1108Set the MAC address for a VF from the PF::
1109
1110   testpmd> set vf mac addr (port_id) (vf_id) (XX:XX:XX:XX:XX:XX)
1111
1112set eth-peer
1113~~~~~~~~~~~~
1114
1115Set the forwarding peer address for certain port::
1116
1117   testpmd> set eth-peer (port_id) (perr_addr)
1118
1119This is equivalent to the ``--eth-peer`` command-line option.
1120
1121set port-uta
1122~~~~~~~~~~~~
1123
1124Set the unicast hash filter(s) on/off for a port::
1125
1126   testpmd> set port (port_id) uta (XX:XX:XX:XX:XX:XX|all) (on|off)
1127
1128set promisc
1129~~~~~~~~~~~
1130
1131Set the promiscuous mode on for a port or for all ports.
1132In promiscuous mode packets are not dropped if they aren't for the specified MAC address::
1133
1134   testpmd> set promisc (port_id|all) (on|off)
1135
1136set allmulti
1137~~~~~~~~~~~~
1138
1139Set the allmulti mode for a port or for all ports::
1140
1141   testpmd> set allmulti (port_id|all) (on|off)
1142
1143Same as the ifconfig (8) option. Controls how multicast packets are handled.
1144
1145set promisc (for VF)
1146~~~~~~~~~~~~~~~~~~~~
1147
1148Set the unicast promiscuous mode for a VF from PF.
1149It's supported by Intel i40e NICs now.
1150In promiscuous mode packets are not dropped if they aren't for the specified MAC address::
1151
1152   testpmd> set vf promisc (port_id) (vf_id) (on|off)
1153
1154set allmulticast (for VF)
1155~~~~~~~~~~~~~~~~~~~~~~~~~
1156
1157Set the multicast promiscuous mode for a VF from PF.
1158It's supported by Intel i40e NICs now.
1159In promiscuous mode packets are not dropped if they aren't for the specified MAC address::
1160
1161   testpmd> set vf allmulti (port_id) (vf_id) (on|off)
1162
1163set tx max bandwidth (for VF)
1164~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
1165
1166Set TX max absolute bandwidth (Mbps) for a VF from PF::
1167
1168   testpmd> set vf tx max-bandwidth (port_id) (vf_id) (max_bandwidth)
1169
1170set tc tx min bandwidth (for VF)
1171~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
1172
1173Set all TCs' TX min relative bandwidth (%) for a VF from PF::
1174
1175   testpmd> set vf tc tx min-bandwidth (port_id) (vf_id) (bw1, bw2, ...)
1176
1177set tc tx max bandwidth (for VF)
1178~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
1179
1180Set a TC's TX max absolute bandwidth (Mbps) for a VF from PF::
1181
1182   testpmd> set vf tc tx max-bandwidth (port_id) (vf_id) (tc_no) (max_bandwidth)
1183
1184set tc strict link priority mode
1185~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
1186
1187Set some TCs' strict link priority mode on a physical port::
1188
1189   testpmd> set tx strict-link-priority (port_id) (tc_bitmap)
1190
1191set tc tx min bandwidth
1192~~~~~~~~~~~~~~~~~~~~~~~
1193
1194Set all TCs' TX min relative bandwidth (%) globally for all PF and VFs::
1195
1196   testpmd> set tc tx min-bandwidth (port_id) (bw1, bw2, ...)
1197
1198set flow_ctrl rx
1199~~~~~~~~~~~~~~~~
1200
1201Set the link flow control parameter on a port::
1202
1203   testpmd> set flow_ctrl rx (on|off) tx (on|off) (high_water) (low_water) \
1204            (pause_time) (send_xon) mac_ctrl_frame_fwd (on|off) \
1205	    autoneg (on|off) (port_id)
1206
1207Where:
1208
1209* ``high_water`` (integer): High threshold value to trigger XOFF.
1210
1211* ``low_water`` (integer): Low threshold value to trigger XON.
1212
1213* ``pause_time`` (integer): Pause quota in the Pause frame.
1214
1215* ``send_xon`` (0/1): Send XON frame.
1216
1217* ``mac_ctrl_frame_fwd``: Enable receiving MAC control frames.
1218
1219* ``autoneg``: Change the auto-negotiation parameter.
1220
1221set pfc_ctrl rx
1222~~~~~~~~~~~~~~~
1223
1224Set the priority flow control parameter on a port::
1225
1226   testpmd> set pfc_ctrl rx (on|off) tx (on|off) (high_water) (low_water) \
1227            (pause_time) (priority) (port_id)
1228
1229Where:
1230
1231* ``high_water`` (integer): High threshold value.
1232
1233* ``low_water`` (integer): Low threshold value.
1234
1235* ``pause_time`` (integer): Pause quota in the Pause frame.
1236
1237* ``priority`` (0-7): VLAN User Priority.
1238
1239set stat_qmap
1240~~~~~~~~~~~~~
1241
1242Set statistics mapping (qmapping 0..15) for RX/TX queue on port::
1243
1244   testpmd> set stat_qmap (tx|rx) (port_id) (queue_id) (qmapping)
1245
1246For example, to set rx queue 2 on port 0 to mapping 5::
1247
1248   testpmd>set stat_qmap rx 0 2 5
1249
1250set xstats-hide-zero
1251~~~~~~~~~~~~~~~~~~~~
1252
1253Set the option to hide zero values for xstats display::
1254
1255	testpmd> set xstats-hide-zero on|off
1256
1257.. note::
1258
1259	By default, the zero values are displayed for xstats.
1260
1261set port - rx/tx (for VF)
1262~~~~~~~~~~~~~~~~~~~~~~~~~
1263
1264Set VF receive/transmit from a port::
1265
1266   testpmd> set port (port_id) vf (vf_id) (rx|tx) (on|off)
1267
1268set port - mac address filter (for VF)
1269~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
1270
1271Add/Remove unicast or multicast MAC addr filter for a VF::
1272
1273   testpmd> set port (port_id) vf (vf_id) (mac_addr) \
1274            (exact-mac|exact-mac-vlan|hashmac|hashmac-vlan) (on|off)
1275
1276set port - rx mode(for VF)
1277~~~~~~~~~~~~~~~~~~~~~~~~~~
1278
1279Set the VF receive mode of a port::
1280
1281   testpmd> set port (port_id) vf (vf_id) \
1282            rxmode (AUPE|ROPE|BAM|MPE) (on|off)
1283
1284The available receive modes are:
1285
1286* ``AUPE``: Accepts untagged VLAN.
1287
1288* ``ROPE``: Accepts unicast hash.
1289
1290* ``BAM``: Accepts broadcast packets.
1291
1292* ``MPE``: Accepts all multicast packets.
1293
1294set port - tx_rate (for Queue)
1295~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
1296
1297Set TX rate limitation for a queue on a port::
1298
1299   testpmd> set port (port_id) queue (queue_id) rate (rate_value)
1300
1301set port - tx_rate (for VF)
1302~~~~~~~~~~~~~~~~~~~~~~~~~~~
1303
1304Set TX rate limitation for queues in VF on a port::
1305
1306   testpmd> set port (port_id) vf (vf_id) rate (rate_value) queue_mask (queue_mask)
1307
1308set port - mirror rule
1309~~~~~~~~~~~~~~~~~~~~~~
1310
1311Set pool or vlan type mirror rule for a port::
1312
1313   testpmd> set port (port_id) mirror-rule (rule_id) \
1314            (pool-mirror-up|pool-mirror-down|vlan-mirror) \
1315            (poolmask|vlanid[,vlanid]*) dst-pool (pool_id) (on|off)
1316
1317Set link mirror rule for a port::
1318
1319   testpmd> set port (port_id) mirror-rule (rule_id) \
1320           (uplink-mirror|downlink-mirror) dst-pool (pool_id) (on|off)
1321
1322For example to enable mirror traffic with vlan 0,1 to pool 0::
1323
1324   set port 0 mirror-rule 0 vlan-mirror 0,1 dst-pool 0 on
1325
1326reset port - mirror rule
1327~~~~~~~~~~~~~~~~~~~~~~~~
1328
1329Reset a mirror rule for a port::
1330
1331   testpmd> reset port (port_id) mirror-rule (rule_id)
1332
1333set flush_rx
1334~~~~~~~~~~~~
1335
1336Set the flush on RX streams before forwarding.
1337The default is flush ``on``.
1338Mainly used with PCAP drivers to turn off the default behavior of flushing the first 512 packets on RX streams::
1339
1340   testpmd> set flush_rx off
1341
1342set bypass mode
1343~~~~~~~~~~~~~~~
1344
1345Set the bypass mode for the lowest port on bypass enabled NIC::
1346
1347   testpmd> set bypass mode (normal|bypass|isolate) (port_id)
1348
1349set bypass event
1350~~~~~~~~~~~~~~~~
1351
1352Set the event required to initiate specified bypass mode for the lowest port on a bypass enabled::
1353
1354   testpmd> set bypass event (timeout|os_on|os_off|power_on|power_off) \
1355            mode (normal|bypass|isolate) (port_id)
1356
1357Where:
1358
1359* ``timeout``: Enable bypass after watchdog timeout.
1360
1361* ``os_on``: Enable bypass when OS/board is powered on.
1362
1363* ``os_off``: Enable bypass when OS/board is powered off.
1364
1365* ``power_on``: Enable bypass when power supply is turned on.
1366
1367* ``power_off``: Enable bypass when power supply is turned off.
1368
1369
1370set bypass timeout
1371~~~~~~~~~~~~~~~~~~
1372
1373Set the bypass watchdog timeout to ``n`` seconds where 0 = instant::
1374
1375   testpmd> set bypass timeout (0|1.5|2|3|4|8|16|32)
1376
1377show bypass config
1378~~~~~~~~~~~~~~~~~~
1379
1380Show the bypass configuration for a bypass enabled NIC using the lowest port on the NIC::
1381
1382   testpmd> show bypass config (port_id)
1383
1384set link up
1385~~~~~~~~~~~
1386
1387Set link up for a port::
1388
1389   testpmd> set link-up port (port id)
1390
1391set link down
1392~~~~~~~~~~~~~
1393
1394Set link down for a port::
1395
1396   testpmd> set link-down port (port id)
1397
1398E-tag set
1399~~~~~~~~~
1400
1401Enable E-tag insertion for a VF on a port::
1402
1403   testpmd> E-tag set insertion on port-tag-id (value) port (port_id) vf (vf_id)
1404
1405Disable E-tag insertion for a VF on a port::
1406
1407   testpmd> E-tag set insertion off port (port_id) vf (vf_id)
1408
1409Enable/disable E-tag stripping on a port::
1410
1411   testpmd> E-tag set stripping (on|off) port (port_id)
1412
1413Enable/disable E-tag based forwarding on a port::
1414
1415   testpmd> E-tag set forwarding (on|off) port (port_id)
1416
1417Add an E-tag forwarding filter on a port::
1418
1419   testpmd> E-tag set filter add e-tag-id (value) dst-pool (pool_id) port (port_id)
1420
1421Delete an E-tag forwarding filter on a port::
1422   testpmd> E-tag set filter del e-tag-id (value) port (port_id)
1423
1424ddp add
1425~~~~~~~
1426
1427Load a dynamic device personalization (DDP) profile and store backup profile::
1428
1429   testpmd> ddp add (port_id) (profile_path[,backup_profile_path])
1430
1431ddp del
1432~~~~~~~
1433
1434Delete a dynamic device personalization profile and restore backup profile::
1435
1436   testpmd> ddp del (port_id) (backup_profile_path)
1437
1438ptype mapping
1439~~~~~~~~~~~~~
1440
1441List all items from the ptype mapping table::
1442
1443   testpmd> ptype mapping get (port_id) (valid_only)
1444
1445Where:
1446
1447* ``valid_only``: A flag indicates if only list valid items(=1) or all itemss(=0).
1448
1449Replace a specific or a group of software defined ptype with a new one::
1450
1451   testpmd> ptype mapping replace  (port_id) (target) (mask) (pkt_type)
1452
1453where:
1454
1455* ``target``: A specific software ptype or a mask to represent a group of software ptypes.
1456
1457* ``mask``: A flag indicate if "target" is a specific software ptype(=0) or a ptype mask(=1).
1458
1459* ``pkt_type``: The new software ptype to replace the old ones.
1460
1461Update hardware defined ptype to software defined packet type mapping table::
1462
1463   testpmd> ptype mapping update (port_id) (hw_ptype) (sw_ptype)
1464
1465where:
1466
1467* ``hw_ptype``: hardware ptype as the index of the ptype mapping table.
1468
1469* ``sw_ptype``: software ptype as the value of the ptype mapping table.
1470
1471Reset ptype mapping table::
1472
1473   testpmd> ptype mapping reset (port_id)
1474
1475config per port Rx offloading
1476~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
1477
1478Enable or disable a per port Rx offloading on all Rx queues of a port::
1479
1480   testpmd> port config (port_id) rx_offload (offloading) on|off
1481
1482* ``offloading``: can be any of these offloading capability:
1483                  vlan_strip, ipv4_cksum, udp_cksum, tcp_cksum, tcp_lro,
1484                  qinq_strip, outer_ipv4_cksum, macsec_strip,
1485                  header_split, vlan_filter, vlan_extend, jumbo_frame,
1486                  crc_strip, scatter, timestamp, security
1487
1488This command should be run when the port is stopped, or else it will fail.
1489
1490config per queue Rx offloading
1491~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
1492
1493Enable or disable a per queue Rx offloading only on a specific Rx queue::
1494
1495   testpmd> port (port_id) rxq (queue_id) rx_offload (offloading) on|off
1496
1497* ``offloading``: can be any of these offloading capability:
1498                  vlan_strip, ipv4_cksum, udp_cksum, tcp_cksum, tcp_lro,
1499                  qinq_strip, outer_ipv4_cksum, macsec_strip,
1500                  header_split, vlan_filter, vlan_extend, jumbo_frame,
1501                  crc_strip, scatter, timestamp, security
1502
1503This command should be run when the port is stopped, or else it will fail.
1504
1505config per port Tx offloading
1506~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
1507
1508Enable or disable a per port Tx offloading on all Tx queues of a port::
1509
1510   testpmd> port config (port_id) tx_offload (offloading) on|off
1511
1512* ``offloading``: can be any of these offloading capability:
1513                  vlan_insert, ipv4_cksum, udp_cksum, udp_cksum,
1514                  sctp_cksum, tcp_tso, udp_tso, outer_ipv4_cksum,
1515                  qinq_insert, vxlan_tnl_tso, gre_tnl_tso,
1516                  ipip_tnl_tso, geneve_tnl_tso, macsec_insert,
1517                  mt_lockfree, multi_segs, fast_free, security
1518
1519This command should be run when the port is stopped, or else it will fail.
1520
1521config per queue Tx offloading
1522~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
1523
1524Enable or disable a per queue Tx offloading only on a specific Tx queue::
1525
1526   testpmd> port (port_id) txq (queue_id) tx_offload (offloading) on|off
1527
1528* ``offloading``: can be any of these offloading capability:
1529                  vlan_insert, ipv4_cksum, udp_cksum, udp_cksum,
1530                  sctp_cksum, tcp_tso, udp_tso, outer_ipv4_cksum,
1531                  qinq_insert, vxlan_tnl_tso, gre_tnl_tso,
1532                  ipip_tnl_tso, geneve_tnl_tso, macsec_insert,
1533                  mt_lockfree, multi_segs, fast_free, security
1534
1535This command should be run when the port is stopped, or else it will fail.
1536
1537
1538Port Functions
1539--------------
1540
1541The following sections show functions for configuring ports.
1542
1543.. note::
1544
1545   Port configuration changes only become active when forwarding is started/restarted.
1546
1547port attach
1548~~~~~~~~~~~
1549
1550Attach a port specified by pci address or virtual device args::
1551
1552   testpmd> port attach (identifier)
1553
1554To attach a new pci device, the device should be recognized by kernel first.
1555Then it should be moved under DPDK management.
1556Finally the port can be attached to testpmd.
1557
1558For example, to move a pci device using ixgbe under DPDK management:
1559
1560.. code-block:: console
1561
1562   # Check the status of the available devices.
1563   ./usertools/dpdk-devbind.py --status
1564
1565   Network devices using DPDK-compatible driver
1566   ============================================
1567   <none>
1568
1569   Network devices using kernel driver
1570   ===================================
1571   0000:0a:00.0 '82599ES 10-Gigabit' if=eth2 drv=ixgbe unused=
1572
1573
1574   # Bind the device to igb_uio.
1575   sudo ./usertools/dpdk-devbind.py -b igb_uio 0000:0a:00.0
1576
1577
1578   # Recheck the status of the devices.
1579   ./usertools/dpdk-devbind.py --status
1580   Network devices using DPDK-compatible driver
1581   ============================================
1582   0000:0a:00.0 '82599ES 10-Gigabit' drv=igb_uio unused=
1583
1584To attach a port created by virtual device, above steps are not needed.
1585
1586For example, to attach a port whose pci address is 0000:0a:00.0.
1587
1588.. code-block:: console
1589
1590   testpmd> port attach 0000:0a:00.0
1591   Attaching a new port...
1592   EAL: PCI device 0000:0a:00.0 on NUMA socket -1
1593   EAL:   probe driver: 8086:10fb rte_ixgbe_pmd
1594   EAL:   PCI memory mapped at 0x7f83bfa00000
1595   EAL:   PCI memory mapped at 0x7f83bfa80000
1596   PMD: eth_ixgbe_dev_init(): MAC: 2, PHY: 18, SFP+: 5
1597   PMD: eth_ixgbe_dev_init(): port 0 vendorID=0x8086 deviceID=0x10fb
1598   Port 0 is attached. Now total ports is 1
1599   Done
1600
1601For example, to attach a port created by pcap PMD.
1602
1603.. code-block:: console
1604
1605   testpmd> port attach net_pcap0
1606   Attaching a new port...
1607   PMD: Initializing pmd_pcap for net_pcap0
1608   PMD: Creating pcap-backed ethdev on numa socket 0
1609   Port 0 is attached. Now total ports is 1
1610   Done
1611
1612In this case, identifier is ``net_pcap0``.
1613This identifier format is the same as ``--vdev`` format of DPDK applications.
1614
1615For example, to re-attach a bonded port which has been previously detached,
1616the mode and slave parameters must be given.
1617
1618.. code-block:: console
1619
1620   testpmd> port attach net_bond_0,mode=0,slave=1
1621   Attaching a new port...
1622   EAL: Initializing pmd_bond for net_bond_0
1623   EAL: Create bonded device net_bond_0 on port 0 in mode 0 on socket 0.
1624   Port 0 is attached. Now total ports is 1
1625   Done
1626
1627
1628port detach
1629~~~~~~~~~~~
1630
1631Detach a specific port::
1632
1633   testpmd> port detach (port_id)
1634
1635Before detaching a port, the port should be stopped and closed.
1636
1637For example, to detach a pci device port 0.
1638
1639.. code-block:: console
1640
1641   testpmd> port stop 0
1642   Stopping ports...
1643   Done
1644   testpmd> port close 0
1645   Closing ports...
1646   Done
1647
1648   testpmd> port detach 0
1649   Detaching a port...
1650   EAL: PCI device 0000:0a:00.0 on NUMA socket -1
1651   EAL:   remove driver: 8086:10fb rte_ixgbe_pmd
1652   EAL:   PCI memory unmapped at 0x7f83bfa00000
1653   EAL:   PCI memory unmapped at 0x7f83bfa80000
1654   Done
1655
1656
1657For example, to detach a virtual device port 0.
1658
1659.. code-block:: console
1660
1661   testpmd> port stop 0
1662   Stopping ports...
1663   Done
1664   testpmd> port close 0
1665   Closing ports...
1666   Done
1667
1668   testpmd> port detach 0
1669   Detaching a port...
1670   PMD: Closing pcap ethdev on numa socket 0
1671   Port 'net_pcap0' is detached. Now total ports is 0
1672   Done
1673
1674To remove a pci device completely from the system, first detach the port from testpmd.
1675Then the device should be moved under kernel management.
1676Finally the device can be removed using kernel pci hotplug functionality.
1677
1678For example, to move a pci device under kernel management:
1679
1680.. code-block:: console
1681
1682   sudo ./usertools/dpdk-devbind.py -b ixgbe 0000:0a:00.0
1683
1684   ./usertools/dpdk-devbind.py --status
1685
1686   Network devices using DPDK-compatible driver
1687   ============================================
1688   <none>
1689
1690   Network devices using kernel driver
1691   ===================================
1692   0000:0a:00.0 '82599ES 10-Gigabit' if=eth2 drv=ixgbe unused=igb_uio
1693
1694To remove a port created by a virtual device, above steps are not needed.
1695
1696port start
1697~~~~~~~~~~
1698
1699Start all ports or a specific port::
1700
1701   testpmd> port start (port_id|all)
1702
1703port stop
1704~~~~~~~~~
1705
1706Stop all ports or a specific port::
1707
1708   testpmd> port stop (port_id|all)
1709
1710port close
1711~~~~~~~~~~
1712
1713Close all ports or a specific port::
1714
1715   testpmd> port close (port_id|all)
1716
1717port config - queue ring size
1718~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
1719
1720Configure a rx/tx queue ring size::
1721
1722   testpmd> port (port_id) (rxq|txq) (queue_id) ring_size (value)
1723
1724Only take effect after command that (re-)start the port or command that setup specific queue.
1725
1726port start/stop queue
1727~~~~~~~~~~~~~~~~~~~~~
1728
1729Start/stop a rx/tx queue on a specific port::
1730
1731   testpmd> port (port_id) (rxq|txq) (queue_id) (start|stop)
1732
1733port setup queue
1734~~~~~~~~~~~~~~~~~~~~~
1735
1736Setup a rx/tx queue on a specific port::
1737
1738   testpmd> port (port_id) (rxq|txq) (queue_id) setup
1739
1740Only take effect when port is started.
1741
1742port config - speed
1743~~~~~~~~~~~~~~~~~~~
1744
1745Set the speed and duplex mode for all ports or a specific port::
1746
1747   testpmd> port config (port_id|all) speed (10|100|1000|10000|25000|40000|50000|100000|auto) \
1748            duplex (half|full|auto)
1749
1750port config - queues/descriptors
1751~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
1752
1753Set number of queues/descriptors for rxq, txq, rxd and txd::
1754
1755   testpmd> port config all (rxq|txq|rxd|txd) (value)
1756
1757This is equivalent to the ``--rxq``, ``--txq``, ``--rxd`` and ``--txd`` command-line options.
1758
1759port config - max-pkt-len
1760~~~~~~~~~~~~~~~~~~~~~~~~~
1761
1762Set the maximum packet length::
1763
1764   testpmd> port config all max-pkt-len (value)
1765
1766This is equivalent to the ``--max-pkt-len`` command-line option.
1767
1768port config - CRC Strip
1769~~~~~~~~~~~~~~~~~~~~~~~
1770
1771Set hardware CRC stripping on or off for all ports::
1772
1773   testpmd> port config all crc-strip (on|off)
1774
1775CRC stripping is on by default.
1776
1777The ``off`` option is equivalent to the ``--disable-crc-strip`` command-line option.
1778
1779port config - scatter
1780~~~~~~~~~~~~~~~~~~~~~~~
1781
1782Set RX scatter mode on or off for all ports::
1783
1784   testpmd> port config all scatter (on|off)
1785
1786RX scatter mode is off by default.
1787
1788The ``on`` option is equivalent to the ``--enable-scatter`` command-line option.
1789
1790port config - RX Checksum
1791~~~~~~~~~~~~~~~~~~~~~~~~~
1792
1793Set hardware RX checksum offload to on or off for all ports::
1794
1795   testpmd> port config all rx-cksum (on|off)
1796
1797Checksum offload is off by default.
1798
1799The ``on`` option is equivalent to the ``--enable-rx-cksum`` command-line option.
1800
1801port config - VLAN
1802~~~~~~~~~~~~~~~~~~
1803
1804Set hardware VLAN on or off for all ports::
1805
1806   testpmd> port config all hw-vlan (on|off)
1807
1808Hardware VLAN is off by default.
1809
1810The ``on`` option is equivalent to the ``--enable-hw-vlan`` command-line option.
1811
1812port config - VLAN filter
1813~~~~~~~~~~~~~~~~~~~~~~~~~
1814
1815Set hardware VLAN filter on or off for all ports::
1816
1817   testpmd> port config all hw-vlan-filter (on|off)
1818
1819Hardware VLAN filter is off by default.
1820
1821The ``on`` option is equivalent to the ``--enable-hw-vlan-filter`` command-line option.
1822
1823port config - VLAN strip
1824~~~~~~~~~~~~~~~~~~~~~~~~
1825
1826Set hardware VLAN strip on or off for all ports::
1827
1828   testpmd> port config all hw-vlan-strip (on|off)
1829
1830Hardware VLAN strip is off by default.
1831
1832The ``on`` option is equivalent to the ``--enable-hw-vlan-strip`` command-line option.
1833
1834port config - VLAN extend
1835~~~~~~~~~~~~~~~~~~~~~~~~~
1836
1837Set hardware VLAN extend on or off for all ports::
1838
1839   testpmd> port config all hw-vlan-extend (on|off)
1840
1841Hardware VLAN extend is off by default.
1842
1843The ``on`` option is equivalent to the ``--enable-hw-vlan-extend`` command-line option.
1844
1845port config - Drop Packets
1846~~~~~~~~~~~~~~~~~~~~~~~~~~
1847
1848Set packet drop for packets with no descriptors on or off for all ports::
1849
1850   testpmd> port config all drop-en (on|off)
1851
1852Packet dropping for packets with no descriptors is off by default.
1853
1854The ``on`` option is equivalent to the ``--enable-drop-en`` command-line option.
1855
1856port config - RSS
1857~~~~~~~~~~~~~~~~~
1858
1859Set the RSS (Receive Side Scaling) mode on or off::
1860
1861   testpmd> port config all rss (all|default|ip|tcp|udp|sctp|ether|port|vxlan|geneve|nvgre|none)
1862
1863RSS is on by default.
1864
1865The ``all`` option is equivalent to ip|tcp|udp|sctp|ether.
1866The ``default`` option enables all supported RSS types reported by device info.
1867The ``none`` option is equivalent to the ``--disable-rss`` command-line option.
1868
1869port config - RSS Reta
1870~~~~~~~~~~~~~~~~~~~~~~
1871
1872Set the RSS (Receive Side Scaling) redirection table::
1873
1874   testpmd> port config all rss reta (hash,queue)[,(hash,queue)]
1875
1876port config - DCB
1877~~~~~~~~~~~~~~~~~
1878
1879Set the DCB mode for an individual port::
1880
1881   testpmd> port config (port_id) dcb vt (on|off) (traffic_class) pfc (on|off)
1882
1883The traffic class should be 4 or 8.
1884
1885port config - Burst
1886~~~~~~~~~~~~~~~~~~~
1887
1888Set the number of packets per burst::
1889
1890   testpmd> port config all burst (value)
1891
1892This is equivalent to the ``--burst`` command-line option.
1893
1894port config - Threshold
1895~~~~~~~~~~~~~~~~~~~~~~~
1896
1897Set thresholds for TX/RX queues::
1898
1899   testpmd> port config all (threshold) (value)
1900
1901Where the threshold type can be:
1902
1903* ``txpt:`` Set the prefetch threshold register of the TX rings, 0 <= value <= 255.
1904
1905* ``txht:`` Set the host threshold register of the TX rings, 0 <= value <= 255.
1906
1907* ``txwt:`` Set the write-back threshold register of the TX rings, 0 <= value <= 255.
1908
1909* ``rxpt:`` Set the prefetch threshold register of the RX rings, 0 <= value <= 255.
1910
1911* ``rxht:`` Set the host threshold register of the RX rings, 0 <= value <= 255.
1912
1913* ``rxwt:`` Set the write-back threshold register of the RX rings, 0 <= value <= 255.
1914
1915* ``txfreet:`` Set the transmit free threshold of the TX rings, 0 <= value <= txd.
1916
1917* ``rxfreet:`` Set the transmit free threshold of the RX rings, 0 <= value <= rxd.
1918
1919* ``txrst:`` Set the transmit RS bit threshold of TX rings, 0 <= value <= txd.
1920
1921These threshold options are also available from the command-line.
1922
1923port config - E-tag
1924~~~~~~~~~~~~~~~~~~~
1925
1926Set the value of ether-type for E-tag::
1927
1928   testpmd> port config (port_id|all) l2-tunnel E-tag ether-type (value)
1929
1930Enable/disable the E-tag support::
1931
1932   testpmd> port config (port_id|all) l2-tunnel E-tag (enable|disable)
1933
1934port config pctype mapping
1935~~~~~~~~~~~~~~~~~~~~~~~~~~
1936
1937Reset pctype mapping table::
1938
1939   testpmd> port config (port_id) pctype mapping reset
1940
1941Update hardware defined pctype to software defined flow type mapping table::
1942
1943   testpmd> port config (port_id) pctype mapping update (pctype_id_0[,pctype_id_1]*) (flow_type_id)
1944
1945where:
1946
1947* ``pctype_id_x``: hardware pctype id as index of bit in bitmask value of the pctype mapping table.
1948
1949* ``flow_type_id``: software flow type id as the index of the pctype mapping table.
1950
1951port config input set
1952~~~~~~~~~~~~~~~~~~~~~
1953
1954Config RSS/FDIR/FDIR flexible payload input set for some pctype::
1955   testpmd> port config (port_id) pctype (pctype_id) \
1956            (hash_inset|fdir_inset|fdir_flx_inset) \
1957	    (get|set|clear) field (field_idx)
1958
1959Clear RSS/FDIR/FDIR flexible payload input set for some pctype::
1960   testpmd> port config (port_id) pctype (pctype_id) \
1961            (hash_inset|fdir_inset|fdir_flx_inset) clear all
1962
1963where:
1964
1965* ``pctype_id``: hardware packet classification types.
1966* ``field_idx``: hardware field index.
1967
1968port config udp_tunnel_port
1969~~~~~~~~~~~~~~~~~~~~~~~~~~~
1970
1971Add/remove UDP tunnel port for VXLAN/GENEVE tunneling protocols::
1972    testpmd> port config (port_id) udp_tunnel_port add|rm vxlan|geneve (udp_port)
1973
1974Link Bonding Functions
1975----------------------
1976
1977The Link Bonding functions make it possible to dynamically create and
1978manage link bonding devices from within testpmd interactive prompt.
1979
1980create bonded device
1981~~~~~~~~~~~~~~~~~~~~
1982
1983Create a new bonding device::
1984
1985   testpmd> create bonded device (mode) (socket)
1986
1987For example, to create a bonded device in mode 1 on socket 0::
1988
1989   testpmd> create bonded 1 0
1990   created new bonded device (port X)
1991
1992add bonding slave
1993~~~~~~~~~~~~~~~~~
1994
1995Adds Ethernet device to a Link Bonding device::
1996
1997   testpmd> add bonding slave (slave id) (port id)
1998
1999For example, to add Ethernet device (port 6) to a Link Bonding device (port 10)::
2000
2001   testpmd> add bonding slave 6 10
2002
2003
2004remove bonding slave
2005~~~~~~~~~~~~~~~~~~~~
2006
2007Removes an Ethernet slave device from a Link Bonding device::
2008
2009   testpmd> remove bonding slave (slave id) (port id)
2010
2011For example, to remove Ethernet slave device (port 6) to a Link Bonding device (port 10)::
2012
2013   testpmd> remove bonding slave 6 10
2014
2015set bonding mode
2016~~~~~~~~~~~~~~~~
2017
2018Set the Link Bonding mode of a Link Bonding device::
2019
2020   testpmd> set bonding mode (value) (port id)
2021
2022For example, to set the bonding mode of a Link Bonding device (port 10) to broadcast (mode 3)::
2023
2024   testpmd> set bonding mode 3 10
2025
2026set bonding primary
2027~~~~~~~~~~~~~~~~~~~
2028
2029Set an Ethernet slave device as the primary device on a Link Bonding device::
2030
2031   testpmd> set bonding primary (slave id) (port id)
2032
2033For example, to set the Ethernet slave device (port 6) as the primary port of a Link Bonding device (port 10)::
2034
2035   testpmd> set bonding primary 6 10
2036
2037set bonding mac
2038~~~~~~~~~~~~~~~
2039
2040Set the MAC address of a Link Bonding device::
2041
2042   testpmd> set bonding mac (port id) (mac)
2043
2044For example, to set the MAC address of a Link Bonding device (port 10) to 00:00:00:00:00:01::
2045
2046   testpmd> set bonding mac 10 00:00:00:00:00:01
2047
2048set bonding xmit_balance_policy
2049~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
2050
2051Set the transmission policy for a Link Bonding device when it is in Balance XOR mode::
2052
2053   testpmd> set bonding xmit_balance_policy (port_id) (l2|l23|l34)
2054
2055For example, set a Link Bonding device (port 10) to use a balance policy of layer 3+4 (IP addresses & UDP ports)::
2056
2057   testpmd> set bonding xmit_balance_policy 10 l34
2058
2059
2060set bonding mon_period
2061~~~~~~~~~~~~~~~~~~~~~~
2062
2063Set the link status monitoring polling period in milliseconds for a bonding device.
2064
2065This adds support for PMD slave devices which do not support link status interrupts.
2066When the mon_period is set to a value greater than 0 then all PMD's which do not support
2067link status ISR will be queried every polling interval to check if their link status has changed::
2068
2069   testpmd> set bonding mon_period (port_id) (value)
2070
2071For example, to set the link status monitoring polling period of bonded device (port 5) to 150ms::
2072
2073   testpmd> set bonding mon_period 5 150
2074
2075
2076set bonding lacp dedicated_queue
2077~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
2078
2079Enable dedicated tx/rx queues on bonding devices slaves to handle LACP control plane traffic
2080when in mode 4 (link-aggregration-802.3ad)::
2081
2082   testpmd> set bonding lacp dedicated_queues (port_id) (enable|disable)
2083
2084
2085set bonding agg_mode
2086~~~~~~~~~~~~~~~~~~~~
2087
2088Enable one of the specific aggregators mode when in mode 4 (link-aggregration-802.3ad)::
2089
2090   testpmd> set bonding agg_mode (port_id) (bandwidth|count|stable)
2091
2092
2093show bonding config
2094~~~~~~~~~~~~~~~~~~~
2095
2096Show the current configuration of a Link Bonding device::
2097
2098   testpmd> show bonding config (port id)
2099
2100For example,
2101to show the configuration a Link Bonding device (port 9) with 3 slave devices (1, 3, 4)
2102in balance mode with a transmission policy of layer 2+3::
2103
2104   testpmd> show bonding config 9
2105        Bonding mode: 2
2106        Balance Xmit Policy: BALANCE_XMIT_POLICY_LAYER23
2107        Slaves (3): [1 3 4]
2108        Active Slaves (3): [1 3 4]
2109        Primary: [3]
2110
2111
2112Register Functions
2113------------------
2114
2115The Register Functions can be used to read from and write to registers on the network card referenced by a port number.
2116This is mainly useful for debugging purposes.
2117Reference should be made to the appropriate datasheet for the network card for details on the register addresses
2118and fields that can be accessed.
2119
2120read reg
2121~~~~~~~~
2122
2123Display the value of a port register::
2124
2125   testpmd> read reg (port_id) (address)
2126
2127For example, to examine the Flow Director control register (FDIRCTL, 0x0000EE000) on an Intel 82599 10 GbE Controller::
2128
2129   testpmd> read reg 0 0xEE00
2130   port 0 PCI register at offset 0xEE00: 0x4A060029 (1241907241)
2131
2132read regfield
2133~~~~~~~~~~~~~
2134
2135Display a port register bit field::
2136
2137   testpmd> read regfield (port_id) (address) (bit_x) (bit_y)
2138
2139For example, reading the lowest two bits from the register in the example above::
2140
2141   testpmd> read regfield 0 0xEE00 0 1
2142   port 0 PCI register at offset 0xEE00: bits[0, 1]=0x1 (1)
2143
2144read regbit
2145~~~~~~~~~~~
2146
2147Display a single port register bit::
2148
2149   testpmd> read regbit (port_id) (address) (bit_x)
2150
2151For example, reading the lowest bit from the register in the example above::
2152
2153   testpmd> read regbit 0 0xEE00 0
2154   port 0 PCI register at offset 0xEE00: bit 0=1
2155
2156write reg
2157~~~~~~~~~
2158
2159Set the value of a port register::
2160
2161   testpmd> write reg (port_id) (address) (value)
2162
2163For example, to clear a register::
2164
2165   testpmd> write reg 0 0xEE00 0x0
2166   port 0 PCI register at offset 0xEE00: 0x00000000 (0)
2167
2168write regfield
2169~~~~~~~~~~~~~~
2170
2171Set bit field of a port register::
2172
2173   testpmd> write regfield (port_id) (address) (bit_x) (bit_y) (value)
2174
2175For example, writing to the register cleared in the example above::
2176
2177   testpmd> write regfield 0 0xEE00 0 1 2
2178   port 0 PCI register at offset 0xEE00: 0x00000002 (2)
2179
2180write regbit
2181~~~~~~~~~~~~
2182
2183Set single bit value of a port register::
2184
2185   testpmd> write regbit (port_id) (address) (bit_x) (value)
2186
2187For example, to set the high bit in the register from the example above::
2188
2189   testpmd> write regbit 0 0xEE00 31 1
2190   port 0 PCI register at offset 0xEE00: 0x8000000A (2147483658)
2191
2192Traffic Metering and Policing
2193-----------------------------
2194
2195The following section shows functions for configuring traffic metering and
2196policing on the ethernet device through the use of generic ethdev API.
2197
2198show port traffic management capability
2199~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
2200
2201Show traffic metering and policing capability of the port::
2202
2203   testpmd> show port meter cap (port_id)
2204
2205add port meter profile (srTCM rfc2967)
2206~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
2207
2208Add meter profile (srTCM rfc2697) to the ethernet device::
2209
2210   testpmd> add port meter profile srtcm_rfc2697 (port_id) (profile_id) \
2211   (cir) (cbs) (ebs)
2212
2213where:
2214
2215* ``profile_id``: ID for the meter profile.
2216* ``cir``: Committed Information Rate (CIR) (bytes/second).
2217* ``cbs``: Committed Burst Size (CBS) (bytes).
2218* ``ebs``: Excess Burst Size (EBS) (bytes).
2219
2220add port meter profile (trTCM rfc2968)
2221~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
2222
2223Add meter profile (srTCM rfc2698) to the ethernet device::
2224
2225   testpmd> add port meter profile trtcm_rfc2698 (port_id) (profile_id) \
2226   (cir) (pir) (cbs) (pbs)
2227
2228where:
2229
2230* ``profile_id``: ID for the meter profile.
2231* ``cir``: Committed information rate (bytes/second).
2232* ``pir``: Peak information rate (bytes/second).
2233* ``cbs``: Committed burst size (bytes).
2234* ``pbs``: Peak burst size (bytes).
2235
2236add port meter profile (trTCM rfc4115)
2237~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
2238
2239Add meter profile (trTCM rfc4115) to the ethernet device::
2240
2241   testpmd> add port meter profile trtcm_rfc4115 (port_id) (profile_id) \
2242   (cir) (eir) (cbs) (ebs)
2243
2244where:
2245
2246* ``profile_id``: ID for the meter profile.
2247* ``cir``: Committed information rate (bytes/second).
2248* ``eir``: Excess information rate (bytes/second).
2249* ``cbs``: Committed burst size (bytes).
2250* ``ebs``: Excess burst size (bytes).
2251
2252delete port meter profile
2253~~~~~~~~~~~~~~~~~~~~~~~~~
2254
2255Delete meter profile from the ethernet device::
2256
2257   testpmd> del port meter profile (port_id) (profile_id)
2258
2259create port meter
2260~~~~~~~~~~~~~~~~~
2261
2262Create new meter object for the ethernet device::
2263
2264   testpmd> create port meter (port_id) (mtr_id) (profile_id) \
2265   (meter_enable) (g_action) (y_action) (r_action) (stats_mask) (shared) \
2266   (use_pre_meter_color) [(dscp_tbl_entry0) (dscp_tbl_entry1)...\
2267   (dscp_tbl_entry63)]
2268
2269where:
2270
2271* ``mtr_id``: meter object ID.
2272* ``profile_id``: ID for the meter profile.
2273* ``meter_enable``: When this parameter has a non-zero value, the meter object
2274  gets enabled at the time of creation, otherwise remains disabled.
2275* ``g_action``: Policer action for the packet with green color.
2276* ``y_action``: Policer action for the packet with yellow color.
2277* ``r_action``: Policer action for the packet with red color.
2278* ``stats_mask``: Mask of statistics counter types to be enabled for the
2279  meter object.
2280* ``shared``:  When this parameter has a non-zero value, the meter object is
2281  shared by multiple flows. Otherwise, meter object is used by single flow.
2282* ``use_pre_meter_color``: When this parameter has a non-zero value, the
2283  input color for the current meter object is determined by the latest meter
2284  object in the same flow. Otherwise, the current meter object uses the
2285  *dscp_table* to determine the input color.
2286* ``dscp_tbl_entryx``: DSCP table entry x providing meter providing input
2287  color, 0 <= x <= 63.
2288
2289enable port meter
2290~~~~~~~~~~~~~~~~~
2291
2292Enable meter for the ethernet device::
2293
2294   testpmd> enable port meter (port_id) (mtr_id)
2295
2296disable port meter
2297~~~~~~~~~~~~~~~~~~
2298
2299Disable meter for the ethernet device::
2300
2301   testpmd> disable port meter (port_id) (mtr_id)
2302
2303delete port meter
2304~~~~~~~~~~~~~~~~~
2305
2306Delete meter for the ethernet device::
2307
2308   testpmd> del port meter (port_id) (mtr_id)
2309
2310Set port meter profile
2311~~~~~~~~~~~~~~~~~~~~~~
2312
2313Set meter profile for the ethernet device::
2314
2315   testpmd> set port meter profile (port_id) (mtr_id) (profile_id)
2316
2317set port meter dscp table
2318~~~~~~~~~~~~~~~~~~~~~~~~~
2319
2320Set meter dscp table for the ethernet device::
2321
2322   testpmd> set port meter dscp table (port_id) (mtr_id) [(dscp_tbl_entry0) \
2323   (dscp_tbl_entry1)...(dscp_tbl_entry63)]
2324
2325set port meter policer action
2326~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
2327
2328Set meter policer action for the ethernet device::
2329
2330   testpmd> set port meter policer action (port_id) (mtr_id) (action_mask) \
2331   (action0) [(action1) (action1)]
2332
2333where:
2334
2335* ``action_mask``: Bit mask indicating which policer actions need to be
2336  updated. One or more policer actions can be updated in a single function
2337  invocation. To update the policer action associated with color C, bit
2338  (1 << C) needs to be set in *action_mask* and element at position C
2339  in the *actions* array needs to be valid.
2340* ``actionx``: Policer action for the color x,
2341  RTE_MTR_GREEN <= x < RTE_MTR_COLORS
2342
2343set port meter stats mask
2344~~~~~~~~~~~~~~~~~~~~~~~~~
2345
2346Set meter stats mask for the ethernet device::
2347
2348   testpmd> set port meter stats mask (port_id) (mtr_id) (stats_mask)
2349
2350where:
2351
2352* ``stats_mask``: Bit mask indicating statistics counter types to be enabled.
2353
2354show port meter stats
2355~~~~~~~~~~~~~~~~~~~~~
2356
2357Show meter stats of the ethernet device::
2358
2359   testpmd> show port meter stats (port_id) (mtr_id) (clear)
2360
2361where:
2362
2363* ``clear``: Flag that indicates whether the statistics counters should
2364  be cleared (i.e. set to zero) immediately after they have been read or not.
2365
2366Traffic Management
2367------------------
2368
2369The following section shows functions for configuring traffic management on
2370on the ethernet device through the use of generic TM API.
2371
2372show port traffic management capability
2373~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
2374
2375Show traffic management capability of the port::
2376
2377   testpmd> show port tm cap (port_id)
2378
2379show port traffic management capability (hierarchy level)
2380~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
2381
2382Show traffic management hierarchy level capability of the port::
2383
2384   testpmd> show port tm level cap (port_id) (level_id)
2385
2386show port traffic management capability (hierarchy node level)
2387~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
2388
2389Show the traffic management hierarchy node capability of the port::
2390
2391   testpmd> show port tm node cap (port_id) (node_id)
2392
2393show port traffic management hierarchy node type
2394~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
2395
2396Show the port traffic management hierarchy node type::
2397
2398   testpmd> show port tm node type (port_id) (node_id)
2399
2400show port traffic management hierarchy node stats
2401~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
2402
2403Show the port traffic management hierarchy node statistics::
2404
2405   testpmd> show port tm node stats (port_id) (node_id) (clear)
2406
2407where:
2408
2409* ``clear``: When this parameter has a non-zero value, the statistics counters
2410  are cleared (i.e. set to zero) immediately after they have been read,
2411  otherwise the statistics counters are left untouched.
2412
2413Add port traffic management private shaper profile
2414~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
2415
2416Add the port traffic management private shaper profile::
2417
2418   testpmd> add port tm node shaper profile (port_id) (shaper_profile_id) \
2419   (tb_rate) (tb_size) (packet_length_adjust)
2420
2421where:
2422
2423* ``shaper_profile id``: Shaper profile ID for the new profile.
2424* ``tb_rate``: Token bucket rate (bytes per second).
2425* ``tb_size``: Token bucket size (bytes).
2426* ``packet_length_adjust``: The value (bytes) to be added to the length of
2427  each packet for the purpose of shaping. This parameter value can be used to
2428  correct the packet length with the framing overhead bytes that are consumed
2429  on the wire.
2430
2431Delete port traffic management private shaper profile
2432~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
2433
2434Delete the port traffic management private shaper::
2435
2436   testpmd> del port tm node shaper profile (port_id) (shaper_profile_id)
2437
2438where:
2439
2440* ``shaper_profile id``: Shaper profile ID that needs to be deleted.
2441
2442Add port traffic management shared shaper
2443~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
2444
2445Create the port traffic management shared shaper::
2446
2447   testpmd> add port tm node shared shaper (port_id) (shared_shaper_id) \
2448   (shaper_profile_id)
2449
2450where:
2451
2452* ``shared_shaper_id``: Shared shaper ID to be created.
2453* ``shaper_profile id``: Shaper profile ID for shared shaper.
2454
2455Set port traffic management shared shaper
2456~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
2457
2458Update the port traffic management shared shaper::
2459
2460   testpmd> set port tm node shared shaper (port_id) (shared_shaper_id) \
2461   (shaper_profile_id)
2462
2463where:
2464
2465* ``shared_shaper_id``: Shared shaper ID to be update.
2466* ``shaper_profile id``: Shaper profile ID for shared shaper.
2467
2468Delete port traffic management shared shaper
2469~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
2470
2471Delete the port traffic management shared shaper::
2472
2473   testpmd> del port tm node shared shaper (port_id) (shared_shaper_id)
2474
2475where:
2476
2477* ``shared_shaper_id``: Shared shaper ID to be deleted.
2478
2479Set port traffic management hiearchy node private shaper
2480~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
2481
2482set the port traffic management hierarchy node private shaper::
2483
2484   testpmd> set port tm node shaper profile (port_id) (node_id) \
2485   (shaper_profile_id)
2486
2487where:
2488
2489* ``shaper_profile id``: Private shaper profile ID to be enabled on the
2490  hierarchy node.
2491
2492Add port traffic management WRED profile
2493~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
2494
2495Create a new WRED profile::
2496
2497   testpmd> add port tm node wred profile (port_id) (wred_profile_id) \
2498   (color_g) (min_th_g) (max_th_g) (maxp_inv_g) (wq_log2_g) \
2499   (color_y) (min_th_y) (max_th_y) (maxp_inv_y) (wq_log2_y) \
2500   (color_r) (min_th_r) (max_th_r) (maxp_inv_r) (wq_log2_r)
2501
2502where:
2503
2504* ``wred_profile id``: Identifier for the newly create WRED profile
2505* ``color_g``: Packet color (green)
2506* ``min_th_g``: Minimum queue threshold for packet with green color
2507* ``max_th_g``: Minimum queue threshold for packet with green color
2508* ``maxp_inv_g``: Inverse of packet marking probability maximum value (maxp)
2509* ``wq_log2_g``: Negated log2 of queue weight (wq)
2510* ``color_y``: Packet color (yellow)
2511* ``min_th_y``: Minimum queue threshold for packet with yellow color
2512* ``max_th_y``: Minimum queue threshold for packet with yellow color
2513* ``maxp_inv_y``: Inverse of packet marking probability maximum value (maxp)
2514* ``wq_log2_y``: Negated log2 of queue weight (wq)
2515* ``color_r``: Packet color (red)
2516* ``min_th_r``: Minimum queue threshold for packet with yellow color
2517* ``max_th_r``: Minimum queue threshold for packet with yellow color
2518* ``maxp_inv_r``: Inverse of packet marking probability maximum value (maxp)
2519* ``wq_log2_r``: Negated log2 of queue weight (wq)
2520
2521Delete port traffic management WRED profile
2522~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
2523
2524Delete the WRED profile::
2525
2526   testpmd> del port tm node wred profile (port_id) (wred_profile_id)
2527
2528Add port traffic management hierarchy nonleaf node
2529~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
2530
2531Add nonleaf node to port traffic management hiearchy::
2532
2533   testpmd> add port tm nonleaf node (port_id) (node_id) (parent_node_id) \
2534   (priority) (weight) (level_id) (shaper_profile_id) \
2535   (n_sp_priorities) (stats_mask) (n_shared_shapers) \
2536   [(shared_shaper_0) (shared_shaper_1) ...] \
2537
2538where:
2539
2540* ``parent_node_id``: Node ID of the parent.
2541* ``priority``: Node priority (highest node priority is zero). This is used by
2542  the SP algorithm running on the parent node for scheduling this node.
2543* ``weight``: Node weight (lowest weight is one). The node weight is relative
2544  to the weight sum of all siblings that have the same priority. It is used by
2545  the WFQ algorithm running on the parent node for scheduling this node.
2546* ``level_id``: Hiearchy level of the node.
2547* ``shaper_profile_id``: Shaper profile ID of the private shaper to be used by
2548  the node.
2549* ``n_sp_priorities``: Number of strict priorities.
2550* ``stats_mask``: Mask of statistics counter types to be enabled for this node.
2551* ``n_shared_shapers``: Number of shared shapers.
2552* ``shared_shaper_id``: Shared shaper id.
2553
2554Add port traffic management hierarchy leaf node
2555~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
2556
2557Add leaf node to port traffic management hiearchy::
2558
2559   testpmd> add port tm leaf node (port_id) (node_id) (parent_node_id) \
2560   (priority) (weight) (level_id) (shaper_profile_id) \
2561   (cman_mode) (wred_profile_id) (stats_mask) (n_shared_shapers) \
2562   [(shared_shaper_id) (shared_shaper_id) ...] \
2563
2564where:
2565
2566* ``parent_node_id``: Node ID of the parent.
2567* ``priority``: Node priority (highest node priority is zero). This is used by
2568  the SP algorithm running on the parent node for scheduling this node.
2569* ``weight``: Node weight (lowest weight is one). The node weight is relative
2570  to the weight sum of all siblings that have the same priority. It is used by
2571  the WFQ algorithm running on the parent node for scheduling this node.
2572* ``level_id``: Hiearchy level of the node.
2573* ``shaper_profile_id``: Shaper profile ID of the private shaper to be used by
2574  the node.
2575* ``cman_mode``: Congestion management mode to be enabled for this node.
2576* ``wred_profile_id``: WRED profile id to be enabled for this node.
2577* ``stats_mask``: Mask of statistics counter types to be enabled for this node.
2578* ``n_shared_shapers``: Number of shared shapers.
2579* ``shared_shaper_id``: Shared shaper id.
2580
2581Delete port traffic management hierarchy node
2582~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
2583
2584Delete node from port traffic management hiearchy::
2585
2586   testpmd> del port tm node (port_id) (node_id)
2587
2588Update port traffic management hierarchy parent node
2589~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
2590
2591Update port traffic management hierarchy parent node::
2592
2593   testpmd> set port tm node parent (port_id) (node_id) (parent_node_id) \
2594   (priority) (weight)
2595
2596This function can only be called after the hierarchy commit invocation. Its
2597success depends on the port support for this operation, as advertised through
2598the port capability set. This function is valid for all nodes of the traffic
2599management hierarchy except root node.
2600
2601Suspend port traffic management hierarchy node
2602~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
2603
2604   testpmd> suspend port tm node (port_id) (node_id)
2605
2606Resume port traffic management hierarchy node
2607~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
2608
2609   testpmd> resume port tm node (port_id) (node_id)
2610
2611Commit port traffic management hierarchy
2612~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
2613
2614Commit the traffic management hierarchy on the port::
2615
2616   testpmd> port tm hierarchy commit (port_id) (clean_on_fail)
2617
2618where:
2619
2620* ``clean_on_fail``: When set to non-zero, hierarchy is cleared on function
2621  call failure. On the other hand, hierarchy is preserved when this parameter
2622  is equal to zero.
2623
2624Set port traffic management default hierarchy (tm forwarding mode)
2625~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
2626
2627set the traffic management default hierarchy on the port::
2628
2629   testpmd> set port tm hierarchy default (port_id)
2630
2631Filter Functions
2632----------------
2633
2634This section details the available filter functions that are available.
2635
2636Note these functions interface the deprecated legacy filtering framework,
2637superseded by *rte_flow*. See `Flow rules management`_.
2638
2639ethertype_filter
2640~~~~~~~~~~~~~~~~~~~~
2641
2642Add or delete a L2 Ethertype filter, which identify packets by their L2 Ethertype mainly assign them to a receive queue::
2643
2644   ethertype_filter (port_id) (add|del) (mac_addr|mac_ignr) (mac_address) \
2645                    ethertype (ether_type) (drop|fwd) queue (queue_id)
2646
2647The available information parameters are:
2648
2649* ``port_id``: The port which the Ethertype filter assigned on.
2650
2651* ``mac_addr``: Compare destination mac address.
2652
2653* ``mac_ignr``: Ignore destination mac address match.
2654
2655* ``mac_address``: Destination mac address to match.
2656
2657* ``ether_type``: The EtherType value want to match,
2658  for example 0x0806 for ARP packet. 0x0800 (IPv4) and 0x86DD (IPv6) are invalid.
2659
2660* ``queue_id``: The receive queue associated with this EtherType filter.
2661  It is meaningless when deleting or dropping.
2662
2663Example, to add/remove an ethertype filter rule::
2664
2665   testpmd> ethertype_filter 0 add mac_ignr 00:11:22:33:44:55 \
2666                             ethertype 0x0806 fwd queue 3
2667
2668   testpmd> ethertype_filter 0 del mac_ignr 00:11:22:33:44:55 \
2669                             ethertype 0x0806 fwd queue 3
2670
26712tuple_filter
2672~~~~~~~~~~~~~~~~~
2673
2674Add or delete a 2-tuple filter,
2675which identifies packets by specific protocol and destination TCP/UDP port
2676and forwards packets into one of the receive queues::
2677
2678   2tuple_filter (port_id) (add|del) dst_port (dst_port_value) \
2679                 protocol (protocol_value) mask (mask_value) \
2680                 tcp_flags (tcp_flags_value) priority (prio_value) \
2681                 queue (queue_id)
2682
2683The available information parameters are:
2684
2685* ``port_id``: The port which the 2-tuple filter assigned on.
2686
2687* ``dst_port_value``: Destination port in L4.
2688
2689* ``protocol_value``: IP L4 protocol.
2690
2691* ``mask_value``: Participates in the match or not by bit for field above, 1b means participate.
2692
2693* ``tcp_flags_value``: TCP control bits. The non-zero value is invalid, when the pro_value is not set to 0x06 (TCP).
2694
2695* ``prio_value``: Priority of this filter.
2696
2697* ``queue_id``: The receive queue associated with this 2-tuple filter.
2698
2699Example, to add/remove an 2tuple filter rule::
2700
2701   testpmd> 2tuple_filter 0 add dst_port 32 protocol 0x06 mask 0x03 \
2702                          tcp_flags 0x02 priority 3 queue 3
2703
2704   testpmd> 2tuple_filter 0 del dst_port 32 protocol 0x06 mask 0x03 \
2705                          tcp_flags 0x02 priority 3 queue 3
2706
27075tuple_filter
2708~~~~~~~~~~~~~~~~~
2709
2710Add or delete a 5-tuple filter,
2711which consists of a 5-tuple (protocol, source and destination IP addresses, source and destination TCP/UDP/SCTP port)
2712and routes packets into one of the receive queues::
2713
2714   5tuple_filter (port_id) (add|del) dst_ip (dst_address) src_ip \
2715                 (src_address) dst_port (dst_port_value) \
2716                 src_port (src_port_value) protocol (protocol_value) \
2717                 mask (mask_value) tcp_flags (tcp_flags_value) \
2718                 priority (prio_value) queue (queue_id)
2719
2720The available information parameters are:
2721
2722* ``port_id``: The port which the 5-tuple filter assigned on.
2723
2724* ``dst_address``: Destination IP address.
2725
2726* ``src_address``: Source IP address.
2727
2728* ``dst_port_value``: TCP/UDP destination port.
2729
2730* ``src_port_value``: TCP/UDP source port.
2731
2732* ``protocol_value``: L4 protocol.
2733
2734* ``mask_value``: Participates in the match or not by bit for field above, 1b means participate
2735
2736* ``tcp_flags_value``: TCP control bits. The non-zero value is invalid, when the protocol_value is not set to 0x06 (TCP).
2737
2738* ``prio_value``: The priority of this filter.
2739
2740* ``queue_id``: The receive queue associated with this 5-tuple filter.
2741
2742Example, to add/remove an 5tuple filter rule::
2743
2744   testpmd> 5tuple_filter 0 add dst_ip 2.2.2.5 src_ip 2.2.2.4 \
2745            dst_port 64 src_port 32 protocol 0x06 mask 0x1F \
2746            flags 0x0 priority 3 queue 3
2747
2748   testpmd> 5tuple_filter 0 del dst_ip 2.2.2.5 src_ip 2.2.2.4 \
2749            dst_port 64 src_port 32 protocol 0x06 mask 0x1F \
2750            flags 0x0 priority 3 queue 3
2751
2752syn_filter
2753~~~~~~~~~~
2754
2755Using the  SYN filter, TCP packets whose *SYN* flag is set can be forwarded to a separate queue::
2756
2757   syn_filter (port_id) (add|del) priority (high|low) queue (queue_id)
2758
2759The available information parameters are:
2760
2761* ``port_id``: The port which the SYN filter assigned on.
2762
2763* ``high``: This SYN filter has higher priority than other filters.
2764
2765* ``low``: This SYN filter has lower priority than other filters.
2766
2767* ``queue_id``: The receive queue associated with this SYN filter
2768
2769Example::
2770
2771   testpmd> syn_filter 0 add priority high queue 3
2772
2773flex_filter
2774~~~~~~~~~~~
2775
2776With flex filter, packets can be recognized by any arbitrary pattern within the first 128 bytes of the packet
2777and routed into one of the receive queues::
2778
2779   flex_filter (port_id) (add|del) len (len_value) bytes (bytes_value) \
2780               mask (mask_value) priority (prio_value) queue (queue_id)
2781
2782The available information parameters are:
2783
2784* ``port_id``: The port which the Flex filter is assigned on.
2785
2786* ``len_value``: Filter length in bytes, no greater than 128.
2787
2788* ``bytes_value``: A string in hexadecimal, means the value the flex filter needs to match.
2789
2790* ``mask_value``: A string in hexadecimal, bit 1 means corresponding byte participates in the match.
2791
2792* ``prio_value``: The priority of this filter.
2793
2794* ``queue_id``: The receive queue associated with this Flex filter.
2795
2796Example::
2797
2798   testpmd> flex_filter 0 add len 16 bytes 0x00000000000000000000000008060000 \
2799                          mask 000C priority 3 queue 3
2800
2801   testpmd> flex_filter 0 del len 16 bytes 0x00000000000000000000000008060000 \
2802                          mask 000C priority 3 queue 3
2803
2804
2805.. _testpmd_flow_director:
2806
2807flow_director_filter
2808~~~~~~~~~~~~~~~~~~~~
2809
2810The Flow Director works in receive mode to identify specific flows or sets of flows and route them to specific queues.
2811
2812Four types of filtering are supported which are referred to as Perfect Match, Signature, Perfect-mac-vlan and
2813Perfect-tunnel filters, the match mode is set by the ``--pkt-filter-mode`` command-line parameter:
2814
2815* Perfect match filters.
2816  The hardware checks a match between the masked fields of the received packets and the programmed filters.
2817  The masked fields are for IP flow.
2818
2819* Signature filters.
2820  The hardware checks a match between a hash-based signature of the masked fields of the received packet.
2821
2822* Perfect-mac-vlan match filters.
2823  The hardware checks a match between the masked fields of the received packets and the programmed filters.
2824  The masked fields are for MAC VLAN flow.
2825
2826* Perfect-tunnel match filters.
2827  The hardware checks a match between the masked fields of the received packets and the programmed filters.
2828  The masked fields are for tunnel flow.
2829
2830* Perfect-raw-flow-type match filters.
2831  The hardware checks a match between the masked fields of the received packets and pre-loaded raw (template) packet.
2832  The masked fields are specified by input sets.
2833
2834The Flow Director filters can match the different fields for different type of packet: flow type, specific input set
2835per flow type and the flexible payload.
2836
2837The Flow Director can also mask out parts of all of these fields so that filters
2838are only applied to certain fields or parts of the fields.
2839
2840Note that for raw flow type mode the source and destination fields in the
2841raw packet buffer need to be presented in a reversed order with respect
2842to the expected received packets.
2843For example: IP source and destination addresses or TCP/UDP/SCTP
2844source and destination ports
2845
2846Different NICs may have different capabilities, command show port fdir (port_id) can be used to acquire the information.
2847
2848# Commands to add flow director filters of different flow types::
2849
2850   flow_director_filter (port_id) mode IP (add|del|update) \
2851                        flow (ipv4-other|ipv4-frag|ipv6-other|ipv6-frag) \
2852                        src (src_ip_address) dst (dst_ip_address) \
2853                        tos (tos_value) proto (proto_value) ttl (ttl_value) \
2854                        vlan (vlan_value) flexbytes (flexbytes_value) \
2855                        (drop|fwd) pf|vf(vf_id) queue (queue_id) \
2856                        fd_id (fd_id_value)
2857
2858   flow_director_filter (port_id) mode IP (add|del|update) \
2859                        flow (ipv4-tcp|ipv4-udp|ipv6-tcp|ipv6-udp) \
2860                        src (src_ip_address) (src_port) \
2861                        dst (dst_ip_address) (dst_port) \
2862                        tos (tos_value) ttl (ttl_value) \
2863                        vlan (vlan_value) flexbytes (flexbytes_value) \
2864                        (drop|fwd) queue pf|vf(vf_id) (queue_id) \
2865                        fd_id (fd_id_value)
2866
2867   flow_director_filter (port_id) mode IP (add|del|update) \
2868                        flow (ipv4-sctp|ipv6-sctp) \
2869                        src (src_ip_address) (src_port) \
2870                        dst (dst_ip_address) (dst_port) \
2871                        tos (tos_value) ttl (ttl_value) \
2872                        tag (verification_tag) vlan (vlan_value) \
2873                        flexbytes (flexbytes_value) (drop|fwd) \
2874                        pf|vf(vf_id) queue (queue_id) fd_id (fd_id_value)
2875
2876   flow_director_filter (port_id) mode IP (add|del|update) flow l2_payload \
2877                        ether (ethertype) flexbytes (flexbytes_value) \
2878                        (drop|fwd) pf|vf(vf_id) queue (queue_id)
2879                        fd_id (fd_id_value)
2880
2881   flow_director_filter (port_id) mode MAC-VLAN (add|del|update) \
2882                        mac (mac_address) vlan (vlan_value) \
2883                        flexbytes (flexbytes_value) (drop|fwd) \
2884                        queue (queue_id) fd_id (fd_id_value)
2885
2886   flow_director_filter (port_id) mode Tunnel (add|del|update) \
2887                        mac (mac_address) vlan (vlan_value) \
2888                        tunnel (NVGRE|VxLAN) tunnel-id (tunnel_id_value) \
2889                        flexbytes (flexbytes_value) (drop|fwd) \
2890                        queue (queue_id) fd_id (fd_id_value)
2891
2892   flow_director_filter (port_id) mode raw (add|del|update) flow (flow_id) \
2893                        (drop|fwd) queue (queue_id) fd_id (fd_id_value) \
2894                        packet (packet file name)
2895
2896For example, to add an ipv4-udp flow type filter::
2897
2898   testpmd> flow_director_filter 0 mode IP add flow ipv4-udp src 2.2.2.3 32 \
2899            dst 2.2.2.5 33 tos 2 ttl 40 vlan 0x1 flexbytes (0x88,0x48) \
2900            fwd pf queue 1 fd_id 1
2901
2902For example, add an ipv4-other flow type filter::
2903
2904   testpmd> flow_director_filter 0 mode IP add flow ipv4-other src 2.2.2.3 \
2905             dst 2.2.2.5 tos 2 proto 20 ttl 40 vlan 0x1 \
2906             flexbytes (0x88,0x48) fwd pf queue 1 fd_id 1
2907
2908flush_flow_director
2909~~~~~~~~~~~~~~~~~~~
2910
2911Flush all flow director filters on a device::
2912
2913   testpmd> flush_flow_director (port_id)
2914
2915Example, to flush all flow director filter on port 0::
2916
2917   testpmd> flush_flow_director 0
2918
2919flow_director_mask
2920~~~~~~~~~~~~~~~~~~
2921
2922Set flow director's input masks::
2923
2924   flow_director_mask (port_id) mode IP vlan (vlan_value) \
2925                      src_mask (ipv4_src) (ipv6_src) (src_port) \
2926                      dst_mask (ipv4_dst) (ipv6_dst) (dst_port)
2927
2928   flow_director_mask (port_id) mode MAC-VLAN vlan (vlan_value)
2929
2930   flow_director_mask (port_id) mode Tunnel vlan (vlan_value) \
2931                      mac (mac_value) tunnel-type (tunnel_type_value) \
2932                      tunnel-id (tunnel_id_value)
2933
2934Example, to set flow director mask on port 0::
2935
2936   testpmd> flow_director_mask 0 mode IP vlan 0xefff \
2937            src_mask 255.255.255.255 \
2938                FFFF:FFFF:FFFF:FFFF:FFFF:FFFF:FFFF:FFFF 0xFFFF \
2939            dst_mask 255.255.255.255 \
2940                FFFF:FFFF:FFFF:FFFF:FFFF:FFFF:FFFF:FFFF 0xFFFF
2941
2942flow_director_flex_mask
2943~~~~~~~~~~~~~~~~~~~~~~~
2944
2945set masks of flow director's flexible payload based on certain flow type::
2946
2947   testpmd> flow_director_flex_mask (port_id) \
2948            flow (none|ipv4-other|ipv4-frag|ipv4-tcp|ipv4-udp|ipv4-sctp| \
2949                  ipv6-other|ipv6-frag|ipv6-tcp|ipv6-udp|ipv6-sctp| \
2950                  l2_payload|all) (mask)
2951
2952Example, to set flow director's flex mask for all flow type on port 0::
2953
2954   testpmd> flow_director_flex_mask 0 flow all \
2955            (0xff,0xff,0,0,0,0,0,0,0,0,0,0,0,0,0,0)
2956
2957
2958flow_director_flex_payload
2959~~~~~~~~~~~~~~~~~~~~~~~~~~
2960
2961Configure flexible payload selection::
2962
2963   flow_director_flex_payload (port_id) (raw|l2|l3|l4) (config)
2964
2965For example, to select the first 16 bytes from the offset 4 (bytes) of packet's payload as flexible payload::
2966
2967   testpmd> flow_director_flex_payload 0 l4 \
2968            (4,5,6,7,8,9,10,11,12,13,14,15,16,17,18,19)
2969
2970get_sym_hash_ena_per_port
2971~~~~~~~~~~~~~~~~~~~~~~~~~
2972
2973Get symmetric hash enable configuration per port::
2974
2975   get_sym_hash_ena_per_port (port_id)
2976
2977For example, to get symmetric hash enable configuration of port 1::
2978
2979   testpmd> get_sym_hash_ena_per_port 1
2980
2981set_sym_hash_ena_per_port
2982~~~~~~~~~~~~~~~~~~~~~~~~~
2983
2984Set symmetric hash enable configuration per port to enable or disable::
2985
2986   set_sym_hash_ena_per_port (port_id) (enable|disable)
2987
2988For example, to set symmetric hash enable configuration of port 1 to enable::
2989
2990   testpmd> set_sym_hash_ena_per_port 1 enable
2991
2992get_hash_global_config
2993~~~~~~~~~~~~~~~~~~~~~~
2994
2995Get the global configurations of hash filters::
2996
2997   get_hash_global_config (port_id)
2998
2999For example, to get the global configurations of hash filters of port 1::
3000
3001   testpmd> get_hash_global_config 1
3002
3003set_hash_global_config
3004~~~~~~~~~~~~~~~~~~~~~~
3005
3006Set the global configurations of hash filters::
3007
3008   set_hash_global_config (port_id) (toeplitz|simple_xor|default) \
3009   (ipv4|ipv4-frag|ipv4-tcp|ipv4-udp|ipv4-sctp|ipv4-other|ipv6|ipv6-frag| \
3010   ipv6-tcp|ipv6-udp|ipv6-sctp|ipv6-other|l2_payload|<flow_id>) \
3011   (enable|disable)
3012
3013For example, to enable simple_xor for flow type of ipv6 on port 2::
3014
3015   testpmd> set_hash_global_config 2 simple_xor ipv6 enable
3016
3017set_hash_input_set
3018~~~~~~~~~~~~~~~~~~
3019
3020Set the input set for hash::
3021
3022   set_hash_input_set (port_id) (ipv4-frag|ipv4-tcp|ipv4-udp|ipv4-sctp| \
3023   ipv4-other|ipv6-frag|ipv6-tcp|ipv6-udp|ipv6-sctp|ipv6-other| \
3024   l2_payload|<flow_id>) (ovlan|ivlan|src-ipv4|dst-ipv4|src-ipv6|dst-ipv6| \
3025   ipv4-tos|ipv4-proto|ipv6-tc|ipv6-next-header|udp-src-port|udp-dst-port| \
3026   tcp-src-port|tcp-dst-port|sctp-src-port|sctp-dst-port|sctp-veri-tag| \
3027   udp-key|gre-key|fld-1st|fld-2nd|fld-3rd|fld-4th|fld-5th|fld-6th|fld-7th| \
3028   fld-8th|none) (select|add)
3029
3030For example, to add source IP to hash input set for flow type of ipv4-udp on port 0::
3031
3032   testpmd> set_hash_input_set 0 ipv4-udp src-ipv4 add
3033
3034set_fdir_input_set
3035~~~~~~~~~~~~~~~~~~
3036
3037The Flow Director filters can match the different fields for different type of packet, i.e. specific input set
3038on per flow type and the flexible payload. This command can be used to change input set for each flow type.
3039
3040Set the input set for flow director::
3041
3042   set_fdir_input_set (port_id) (ipv4-frag|ipv4-tcp|ipv4-udp|ipv4-sctp| \
3043   ipv4-other|ipv6|ipv6-frag|ipv6-tcp|ipv6-udp|ipv6-sctp|ipv6-other| \
3044   l2_payload|<flow_id>) (ivlan|ethertype|src-ipv4|dst-ipv4|src-ipv6|dst-ipv6| \
3045   ipv4-tos|ipv4-proto|ipv4-ttl|ipv6-tc|ipv6-next-header|ipv6-hop-limits| \
3046   tudp-src-port|udp-dst-port|cp-src-port|tcp-dst-port|sctp-src-port| \
3047   sctp-dst-port|sctp-veri-tag|none) (select|add)
3048
3049For example to add source IP to FD input set for flow type of ipv4-udp on port 0::
3050
3051   testpmd> set_fdir_input_set 0 ipv4-udp src-ipv4 add
3052
3053global_config
3054~~~~~~~~~~~~~
3055
3056Set different GRE key length for input set::
3057
3058   global_config (port_id) gre-key-len (number in bytes)
3059
3060For example to set GRE key length for input set to 4 bytes on port 0::
3061
3062   testpmd> global_config 0 gre-key-len 4
3063
3064
3065.. _testpmd_rte_flow:
3066
3067Flow rules management
3068---------------------
3069
3070Control of the generic flow API (*rte_flow*) is fully exposed through the
3071``flow`` command (validation, creation, destruction, queries and operation
3072modes).
3073
3074Considering *rte_flow* overlaps with all `Filter Functions`_, using both
3075features simultaneously may cause undefined side-effects and is therefore
3076not recommended.
3077
3078``flow`` syntax
3079~~~~~~~~~~~~~~~
3080
3081Because the ``flow`` command uses dynamic tokens to handle the large number
3082of possible flow rules combinations, its behavior differs slightly from
3083other commands, in particular:
3084
3085- Pressing *?* or the *<tab>* key displays contextual help for the current
3086  token, not that of the entire command.
3087
3088- Optional and repeated parameters are supported (provided they are listed
3089  in the contextual help).
3090
3091The first parameter stands for the operation mode. Possible operations and
3092their general syntax are described below. They are covered in detail in the
3093following sections.
3094
3095- Check whether a flow rule can be created::
3096
3097   flow validate {port_id}
3098       [group {group_id}] [priority {level}] [ingress] [egress] [transfer]
3099       pattern {item} [/ {item} [...]] / end
3100       actions {action} [/ {action} [...]] / end
3101
3102- Create a flow rule::
3103
3104   flow create {port_id}
3105       [group {group_id}] [priority {level}] [ingress] [egress] [transfer]
3106       pattern {item} [/ {item} [...]] / end
3107       actions {action} [/ {action} [...]] / end
3108
3109- Destroy specific flow rules::
3110
3111   flow destroy {port_id} rule {rule_id} [...]
3112
3113- Destroy all flow rules::
3114
3115   flow flush {port_id}
3116
3117- Query an existing flow rule::
3118
3119   flow query {port_id} {rule_id} {action}
3120
3121- List existing flow rules sorted by priority, filtered by group
3122  identifiers::
3123
3124   flow list {port_id} [group {group_id}] [...]
3125
3126- Restrict ingress traffic to the defined flow rules::
3127
3128   flow isolate {port_id} {boolean}
3129
3130Validating flow rules
3131~~~~~~~~~~~~~~~~~~~~~
3132
3133``flow validate`` reports whether a flow rule would be accepted by the
3134underlying device in its current state but stops short of creating it. It is
3135bound to ``rte_flow_validate()``::
3136
3137   flow validate {port_id}
3138      [group {group_id}] [priority {level}] [ingress] [egress] [transfer]
3139      pattern {item} [/ {item} [...]] / end
3140      actions {action} [/ {action} [...]] / end
3141
3142If successful, it will show::
3143
3144   Flow rule validated
3145
3146Otherwise it will show an error message of the form::
3147
3148   Caught error type [...] ([...]): [...]
3149
3150This command uses the same parameters as ``flow create``, their format is
3151described in `Creating flow rules`_.
3152
3153Check whether redirecting any Ethernet packet received on port 0 to RX queue
3154index 6 is supported::
3155
3156   testpmd> flow validate 0 ingress pattern eth / end
3157      actions queue index 6 / end
3158   Flow rule validated
3159   testpmd>
3160
3161Port 0 does not support TCPv6 rules::
3162
3163   testpmd> flow validate 0 ingress pattern eth / ipv6 / tcp / end
3164      actions drop / end
3165   Caught error type 9 (specific pattern item): Invalid argument
3166   testpmd>
3167
3168Creating flow rules
3169~~~~~~~~~~~~~~~~~~~
3170
3171``flow create`` validates and creates the specified flow rule. It is bound
3172to ``rte_flow_create()``::
3173
3174   flow create {port_id}
3175      [group {group_id}] [priority {level}] [ingress] [egress] [transfer]
3176      pattern {item} [/ {item} [...]] / end
3177      actions {action} [/ {action} [...]] / end
3178
3179If successful, it will return a flow rule ID usable with other commands::
3180
3181   Flow rule #[...] created
3182
3183Otherwise it will show an error message of the form::
3184
3185   Caught error type [...] ([...]): [...]
3186
3187Parameters describe in the following order:
3188
3189- Attributes (*group*, *priority*, *ingress*, *egress*, *transfer* tokens).
3190- A matching pattern, starting with the *pattern* token and terminated by an
3191  *end* pattern item.
3192- Actions, starting with the *actions* token and terminated by an *end*
3193  action.
3194
3195These translate directly to *rte_flow* objects provided as-is to the
3196underlying functions.
3197
3198The shortest valid definition only comprises mandatory tokens::
3199
3200   testpmd> flow create 0 pattern end actions end
3201
3202Note that PMDs may refuse rules that essentially do nothing such as this
3203one.
3204
3205**All unspecified object values are automatically initialized to 0.**
3206
3207Attributes
3208^^^^^^^^^^
3209
3210These tokens affect flow rule attributes (``struct rte_flow_attr``) and are
3211specified before the ``pattern`` token.
3212
3213- ``group {group id}``: priority group.
3214- ``priority {level}``: priority level within group.
3215- ``ingress``: rule applies to ingress traffic.
3216- ``egress``: rule applies to egress traffic.
3217- ``transfer``: apply rule directly to endpoints found in pattern.
3218
3219Each instance of an attribute specified several times overrides the previous
3220value as shown below (group 4 is used)::
3221
3222   testpmd> flow create 0 group 42 group 24 group 4 [...]
3223
3224Note that once enabled, ``ingress`` and ``egress`` cannot be disabled.
3225
3226While not specifying a direction is an error, some rules may allow both
3227simultaneously.
3228
3229Most rules affect RX therefore contain the ``ingress`` token::
3230
3231   testpmd> flow create 0 ingress pattern [...]
3232
3233Matching pattern
3234^^^^^^^^^^^^^^^^
3235
3236A matching pattern starts after the ``pattern`` token. It is made of pattern
3237items and is terminated by a mandatory ``end`` item.
3238
3239Items are named after their type (*RTE_FLOW_ITEM_TYPE_* from ``enum
3240rte_flow_item_type``).
3241
3242The ``/`` token is used as a separator between pattern items as shown
3243below::
3244
3245   testpmd> flow create 0 ingress pattern eth / ipv4 / udp / end [...]
3246
3247Note that protocol items like these must be stacked from lowest to highest
3248layer to make sense. For instance, the following rule is either invalid or
3249unlikely to match any packet::
3250
3251   testpmd> flow create 0 ingress pattern eth / udp / ipv4 / end [...]
3252
3253More information on these restrictions can be found in the *rte_flow*
3254documentation.
3255
3256Several items support additional specification structures, for example
3257``ipv4`` allows specifying source and destination addresses as follows::
3258
3259   testpmd> flow create 0 ingress pattern eth / ipv4 src is 10.1.1.1
3260      dst is 10.2.0.0 / end [...]
3261
3262This rule matches all IPv4 traffic with the specified properties.
3263
3264In this example, ``src`` and ``dst`` are field names of the underlying
3265``struct rte_flow_item_ipv4`` object. All item properties can be specified
3266in a similar fashion.
3267
3268The ``is`` token means that the subsequent value must be matched exactly,
3269and assigns ``spec`` and ``mask`` fields in ``struct rte_flow_item``
3270accordingly. Possible assignment tokens are:
3271
3272- ``is``: match value perfectly (with full bit-mask).
3273- ``spec``: match value according to configured bit-mask.
3274- ``last``: specify upper bound to establish a range.
3275- ``mask``: specify bit-mask with relevant bits set to one.
3276- ``prefix``: generate bit-mask from a prefix length.
3277
3278These yield identical results::
3279
3280   ipv4 src is 10.1.1.1
3281
3282::
3283
3284   ipv4 src spec 10.1.1.1 src mask 255.255.255.255
3285
3286::
3287
3288   ipv4 src spec 10.1.1.1 src prefix 32
3289
3290::
3291
3292   ipv4 src is 10.1.1.1 src last 10.1.1.1 # range with a single value
3293
3294::
3295
3296   ipv4 src is 10.1.1.1 src last 0 # 0 disables range
3297
3298Inclusive ranges can be defined with ``last``::
3299
3300   ipv4 src is 10.1.1.1 src last 10.2.3.4 # 10.1.1.1 to 10.2.3.4
3301
3302Note that ``mask`` affects both ``spec`` and ``last``::
3303
3304   ipv4 src is 10.1.1.1 src last 10.2.3.4 src mask 255.255.0.0
3305      # matches 10.1.0.0 to 10.2.255.255
3306
3307Properties can be modified multiple times::
3308
3309   ipv4 src is 10.1.1.1 src is 10.1.2.3 src is 10.2.3.4 # matches 10.2.3.4
3310
3311::
3312
3313   ipv4 src is 10.1.1.1 src prefix 24 src prefix 16 # matches 10.1.0.0/16
3314
3315Pattern items
3316^^^^^^^^^^^^^
3317
3318This section lists supported pattern items and their attributes, if any.
3319
3320- ``end``: end list of pattern items.
3321
3322- ``void``: no-op pattern item.
3323
3324- ``invert``: perform actions when pattern does not match.
3325
3326- ``any``: match any protocol for the current layer.
3327
3328  - ``num {unsigned}``: number of layers covered.
3329
3330- ``pf``: match traffic from/to the physical function.
3331
3332- ``vf``: match traffic from/to a virtual function ID.
3333
3334  - ``id {unsigned}``: VF ID.
3335
3336- ``phy_port``: match traffic from/to a specific physical port.
3337
3338  - ``index {unsigned}``: physical port index.
3339
3340- ``port_id``: match traffic from/to a given DPDK port ID.
3341
3342  - ``id {unsigned}``: DPDK port ID.
3343
3344- ``mark``: match value set in previously matched flow rule using the mark action.
3345
3346  - ``id {unsigned}``: arbitrary integer value.
3347
3348- ``raw``: match an arbitrary byte string.
3349
3350  - ``relative {boolean}``: look for pattern after the previous item.
3351  - ``search {boolean}``: search pattern from offset (see also limit).
3352  - ``offset {integer}``: absolute or relative offset for pattern.
3353  - ``limit {unsigned}``: search area limit for start of pattern.
3354  - ``pattern {string}``: byte string to look for.
3355
3356- ``eth``: match Ethernet header.
3357
3358  - ``dst {MAC-48}``: destination MAC.
3359  - ``src {MAC-48}``: source MAC.
3360  - ``type {unsigned}``: EtherType or TPID.
3361
3362- ``vlan``: match 802.1Q/ad VLAN tag.
3363
3364  - ``tci {unsigned}``: tag control information.
3365  - ``pcp {unsigned}``: priority code point.
3366  - ``dei {unsigned}``: drop eligible indicator.
3367  - ``vid {unsigned}``: VLAN identifier.
3368  - ``inner_type {unsigned}``: inner EtherType or TPID.
3369
3370- ``ipv4``: match IPv4 header.
3371
3372  - ``tos {unsigned}``: type of service.
3373  - ``ttl {unsigned}``: time to live.
3374  - ``proto {unsigned}``: next protocol ID.
3375  - ``src {ipv4 address}``: source address.
3376  - ``dst {ipv4 address}``: destination address.
3377
3378- ``ipv6``: match IPv6 header.
3379
3380  - ``tc {unsigned}``: traffic class.
3381  - ``flow {unsigned}``: flow label.
3382  - ``proto {unsigned}``: protocol (next header).
3383  - ``hop {unsigned}``: hop limit.
3384  - ``src {ipv6 address}``: source address.
3385  - ``dst {ipv6 address}``: destination address.
3386
3387- ``icmp``: match ICMP header.
3388
3389  - ``type {unsigned}``: ICMP packet type.
3390  - ``code {unsigned}``: ICMP packet code.
3391
3392- ``udp``: match UDP header.
3393
3394  - ``src {unsigned}``: UDP source port.
3395  - ``dst {unsigned}``: UDP destination port.
3396
3397- ``tcp``: match TCP header.
3398
3399  - ``src {unsigned}``: TCP source port.
3400  - ``dst {unsigned}``: TCP destination port.
3401
3402- ``sctp``: match SCTP header.
3403
3404  - ``src {unsigned}``: SCTP source port.
3405  - ``dst {unsigned}``: SCTP destination port.
3406  - ``tag {unsigned}``: validation tag.
3407  - ``cksum {unsigned}``: checksum.
3408
3409- ``vxlan``: match VXLAN header.
3410
3411  - ``vni {unsigned}``: VXLAN identifier.
3412
3413- ``e_tag``: match IEEE 802.1BR E-Tag header.
3414
3415  - ``grp_ecid_b {unsigned}``: GRP and E-CID base.
3416
3417- ``nvgre``: match NVGRE header.
3418
3419  - ``tni {unsigned}``: virtual subnet ID.
3420
3421- ``mpls``: match MPLS header.
3422
3423  - ``label {unsigned}``: MPLS label.
3424
3425- ``gre``: match GRE header.
3426
3427  - ``protocol {unsigned}``: protocol type.
3428
3429- ``fuzzy``: fuzzy pattern match, expect faster than default.
3430
3431  - ``thresh {unsigned}``: accuracy threshold.
3432
3433- ``gtp``, ``gtpc``, ``gtpu``: match GTPv1 header.
3434
3435  - ``teid {unsigned}``: tunnel endpoint identifier.
3436
3437- ``geneve``: match GENEVE header.
3438
3439  - ``vni {unsigned}``: virtual network identifier.
3440  - ``protocol {unsigned}``: protocol type.
3441
3442- ``vxlan-gpe``: match VXLAN-GPE header.
3443
3444  - ``vni {unsigned}``: VXLAN-GPE identifier.
3445
3446- ``arp_eth_ipv4``: match ARP header for Ethernet/IPv4.
3447
3448  - ``sha {MAC-48}``: sender hardware address.
3449  - ``spa {ipv4 address}``: sender IPv4 address.
3450  - ``tha {MAC-48}``: target hardware address.
3451  - ``tpa {ipv4 address}``: target IPv4 address.
3452
3453- ``ipv6_ext``: match presence of any IPv6 extension header.
3454
3455  - ``next_hdr {unsigned}``: next header.
3456
3457- ``icmp6``: match any ICMPv6 header.
3458
3459  - ``type {unsigned}``: ICMPv6 type.
3460  - ``code {unsigned}``: ICMPv6 code.
3461
3462- ``icmp6_nd_ns``: match ICMPv6 neighbor discovery solicitation.
3463
3464  - ``target_addr {ipv6 address}``: target address.
3465
3466- ``icmp6_nd_na``: match ICMPv6 neighbor discovery advertisement.
3467
3468  - ``target_addr {ipv6 address}``: target address.
3469
3470- ``icmp6_nd_opt``: match presence of any ICMPv6 neighbor discovery option.
3471
3472  - ``type {unsigned}``: ND option type.
3473
3474- ``icmp6_nd_opt_sla_eth``: match ICMPv6 neighbor discovery source Ethernet
3475  link-layer address option.
3476
3477  - ``sla {MAC-48}``: source Ethernet LLA.
3478
3479- ``icmp6_nd_opt_sla_eth``: match ICMPv6 neighbor discovery target Ethernet
3480  link-layer address option.
3481
3482  - ``tla {MAC-48}``: target Ethernet LLA.
3483
3484Actions list
3485^^^^^^^^^^^^
3486
3487A list of actions starts after the ``actions`` token in the same fashion as
3488`Matching pattern`_; actions are separated by ``/`` tokens and the list is
3489terminated by a mandatory ``end`` action.
3490
3491Actions are named after their type (*RTE_FLOW_ACTION_TYPE_* from ``enum
3492rte_flow_action_type``).
3493
3494Dropping all incoming UDPv4 packets can be expressed as follows::
3495
3496   testpmd> flow create 0 ingress pattern eth / ipv4 / udp / end
3497      actions drop / end
3498
3499Several actions have configurable properties which must be specified when
3500there is no valid default value. For example, ``queue`` requires a target
3501queue index.
3502
3503This rule redirects incoming UDPv4 traffic to queue index 6::
3504
3505   testpmd> flow create 0 ingress pattern eth / ipv4 / udp / end
3506      actions queue index 6 / end
3507
3508While this one could be rejected by PMDs (unspecified queue index)::
3509
3510   testpmd> flow create 0 ingress pattern eth / ipv4 / udp / end
3511      actions queue / end
3512
3513As defined by *rte_flow*, the list is not ordered, all actions of a given
3514rule are performed simultaneously. These are equivalent::
3515
3516   queue index 6 / void / mark id 42 / end
3517
3518::
3519
3520   void / mark id 42 / queue index 6 / end
3521
3522All actions in a list should have different types, otherwise only the last
3523action of a given type is taken into account::
3524
3525   queue index 4 / queue index 5 / queue index 6 / end # will use queue 6
3526
3527::
3528
3529   drop / drop / drop / end # drop is performed only once
3530
3531::
3532
3533   mark id 42 / queue index 3 / mark id 24 / end # mark will be 24
3534
3535Considering they are performed simultaneously, opposite and overlapping
3536actions can sometimes be combined when the end result is unambiguous::
3537
3538   drop / queue index 6 / end # drop has no effect
3539
3540::
3541
3542   queue index 6 / rss queues 6 7 8 / end # queue has no effect
3543
3544::
3545
3546   drop / passthru / end # drop has no effect
3547
3548Note that PMDs may still refuse such combinations.
3549
3550Actions
3551^^^^^^^
3552
3553This section lists supported actions and their attributes, if any.
3554
3555- ``end``: end list of actions.
3556
3557- ``void``: no-op action.
3558
3559- ``passthru``: let subsequent rule process matched packets.
3560
3561- ``jump``: redirect traffic to group on device.
3562
3563  - ``group {unsigned}``: group to redirect to.
3564
3565- ``mark``: attach 32 bit value to packets.
3566
3567  - ``id {unsigned}``: 32 bit value to return with packets.
3568
3569- ``flag``: flag packets.
3570
3571- ``queue``: assign packets to a given queue index.
3572
3573  - ``index {unsigned}``: queue index to use.
3574
3575- ``drop``: drop packets (note: passthru has priority).
3576
3577- ``count``: enable counters for this rule.
3578
3579- ``rss``: spread packets among several queues.
3580
3581  - ``func {hash function}``: RSS hash function to apply, allowed tokens are
3582    the same as `set_hash_global_config`_.
3583
3584  - ``level {unsigned}``: encapsulation level for ``types``.
3585
3586  - ``types [{RSS hash type} [...]] end``: specific RSS hash types, allowed
3587    tokens are the same as `set_hash_input_set`_, except that an empty list
3588    does not disable RSS but instead requests unspecified "best-effort"
3589    settings.
3590
3591  - ``key {string}``: RSS hash key, overrides ``key_len``.
3592
3593  - ``key_len {unsigned}``: RSS hash key length in bytes, can be used in
3594    conjunction with ``key`` to pad or truncate it.
3595
3596  - ``queues [{unsigned} [...]] end``: queue indices to use.
3597
3598- ``pf``: direct traffic to physical function.
3599
3600- ``vf``: direct traffic to a virtual function ID.
3601
3602  - ``original {boolean}``: use original VF ID if possible.
3603  - ``id {unsigned}``: VF ID.
3604
3605- ``phy_port``: direct packets to physical port index.
3606
3607  - ``original {boolean}``: use original port index if possible.
3608  - ``index {unsigned}``: physical port index.
3609
3610- ``port_id``: direct matching traffic to a given DPDK port ID.
3611
3612  - ``original {boolean}``: use original DPDK port ID if possible.
3613  - ``id {unsigned}``: DPDK port ID.
3614
3615- ``of_set_mpls_ttl``: OpenFlow's ``OFPAT_SET_MPLS_TTL``.
3616
3617  - ``mpls_ttl``: MPLS TTL.
3618
3619- ``of_dec_mpls_ttl``: OpenFlow's ``OFPAT_DEC_MPLS_TTL``.
3620
3621- ``of_set_nw_ttl``: OpenFlow's ``OFPAT_SET_NW_TTL``.
3622
3623  - ``nw_ttl``: IP TTL.
3624
3625- ``of_dec_nw_ttl``: OpenFlow's ``OFPAT_DEC_NW_TTL``.
3626
3627- ``of_copy_ttl_out``: OpenFlow's ``OFPAT_COPY_TTL_OUT``.
3628
3629- ``of_copy_ttl_in``: OpenFlow's ``OFPAT_COPY_TTL_IN``.
3630
3631- ``of_pop_vlan``: OpenFlow's ``OFPAT_POP_VLAN``.
3632
3633- ``of_push_vlan``: OpenFlow's ``OFPAT_PUSH_VLAN``.
3634
3635  - ``ethertype``: Ethertype.
3636
3637- ``of_set_vlan_vid``: OpenFlow's ``OFPAT_SET_VLAN_VID``.
3638
3639  - ``vlan_vid``: VLAN id.
3640
3641- ``of_set_vlan_pcp``: OpenFlow's ``OFPAT_SET_VLAN_PCP``.
3642
3643  - ``vlan_pcp``: VLAN priority.
3644
3645- ``of_pop_mpls``: OpenFlow's ``OFPAT_POP_MPLS``.
3646
3647  - ``ethertype``: Ethertype.
3648
3649- ``of_push_mpls``: OpenFlow's ``OFPAT_PUSH_MPLS``.
3650
3651  - ``ethertype``: Ethertype.
3652
3653Destroying flow rules
3654~~~~~~~~~~~~~~~~~~~~~
3655
3656``flow destroy`` destroys one or more rules from their rule ID (as returned
3657by ``flow create``), this command calls ``rte_flow_destroy()`` as many
3658times as necessary::
3659
3660   flow destroy {port_id} rule {rule_id} [...]
3661
3662If successful, it will show::
3663
3664   Flow rule #[...] destroyed
3665
3666It does not report anything for rule IDs that do not exist. The usual error
3667message is shown when a rule cannot be destroyed::
3668
3669   Caught error type [...] ([...]): [...]
3670
3671``flow flush`` destroys all rules on a device and does not take extra
3672arguments. It is bound to ``rte_flow_flush()``::
3673
3674   flow flush {port_id}
3675
3676Any errors are reported as above.
3677
3678Creating several rules and destroying them::
3679
3680   testpmd> flow create 0 ingress pattern eth / ipv6 / end
3681      actions queue index 2 / end
3682   Flow rule #0 created
3683   testpmd> flow create 0 ingress pattern eth / ipv4 / end
3684      actions queue index 3 / end
3685   Flow rule #1 created
3686   testpmd> flow destroy 0 rule 0 rule 1
3687   Flow rule #1 destroyed
3688   Flow rule #0 destroyed
3689   testpmd>
3690
3691The same result can be achieved using ``flow flush``::
3692
3693   testpmd> flow create 0 ingress pattern eth / ipv6 / end
3694      actions queue index 2 / end
3695   Flow rule #0 created
3696   testpmd> flow create 0 ingress pattern eth / ipv4 / end
3697      actions queue index 3 / end
3698   Flow rule #1 created
3699   testpmd> flow flush 0
3700   testpmd>
3701
3702Non-existent rule IDs are ignored::
3703
3704   testpmd> flow create 0 ingress pattern eth / ipv6 / end
3705      actions queue index 2 / end
3706   Flow rule #0 created
3707   testpmd> flow create 0 ingress pattern eth / ipv4 / end
3708      actions queue index 3 / end
3709   Flow rule #1 created
3710   testpmd> flow destroy 0 rule 42 rule 10 rule 2
3711   testpmd>
3712   testpmd> flow destroy 0 rule 0
3713   Flow rule #0 destroyed
3714   testpmd>
3715
3716Querying flow rules
3717~~~~~~~~~~~~~~~~~~~
3718
3719``flow query`` queries a specific action of a flow rule having that
3720ability. Such actions collect information that can be reported using this
3721command. It is bound to ``rte_flow_query()``::
3722
3723   flow query {port_id} {rule_id} {action}
3724
3725If successful, it will display either the retrieved data for known actions
3726or the following message::
3727
3728   Cannot display result for action type [...] ([...])
3729
3730Otherwise, it will complain either that the rule does not exist or that some
3731error occurred::
3732
3733   Flow rule #[...] not found
3734
3735::
3736
3737   Caught error type [...] ([...]): [...]
3738
3739Currently only the ``count`` action is supported. This action reports the
3740number of packets that hit the flow rule and the total number of bytes. Its
3741output has the following format::
3742
3743   count:
3744    hits_set: [...] # whether "hits" contains a valid value
3745    bytes_set: [...] # whether "bytes" contains a valid value
3746    hits: [...] # number of packets
3747    bytes: [...] # number of bytes
3748
3749Querying counters for TCPv6 packets redirected to queue 6::
3750
3751   testpmd> flow create 0 ingress pattern eth / ipv6 / tcp / end
3752      actions queue index 6 / count / end
3753   Flow rule #4 created
3754   testpmd> flow query 0 4 count
3755   count:
3756    hits_set: 1
3757    bytes_set: 0
3758    hits: 386446
3759    bytes: 0
3760   testpmd>
3761
3762Listing flow rules
3763~~~~~~~~~~~~~~~~~~
3764
3765``flow list`` lists existing flow rules sorted by priority and optionally
3766filtered by group identifiers::
3767
3768   flow list {port_id} [group {group_id}] [...]
3769
3770This command only fails with the following message if the device does not
3771exist::
3772
3773   Invalid port [...]
3774
3775Output consists of a header line followed by a short description of each
3776flow rule, one per line. There is no output at all when no flow rules are
3777configured on the device::
3778
3779   ID      Group   Prio    Attr    Rule
3780   [...]   [...]   [...]   [...]   [...]
3781
3782``Attr`` column flags:
3783
3784- ``i`` for ``ingress``.
3785- ``e`` for ``egress``.
3786
3787Creating several flow rules and listing them::
3788
3789   testpmd> flow create 0 ingress pattern eth / ipv4 / end
3790      actions queue index 6 / end
3791   Flow rule #0 created
3792   testpmd> flow create 0 ingress pattern eth / ipv6 / end
3793      actions queue index 2 / end
3794   Flow rule #1 created
3795   testpmd> flow create 0 priority 5 ingress pattern eth / ipv4 / udp / end
3796      actions rss queues 6 7 8 end / end
3797   Flow rule #2 created
3798   testpmd> flow list 0
3799   ID      Group   Prio    Attr    Rule
3800   0       0       0       i-      ETH IPV4 => QUEUE
3801   1       0       0       i-      ETH IPV6 => QUEUE
3802   2       0       5       i-      ETH IPV4 UDP => RSS
3803   testpmd>
3804
3805Rules are sorted by priority (i.e. group ID first, then priority level)::
3806
3807   testpmd> flow list 1
3808   ID      Group   Prio    Attr    Rule
3809   0       0       0       i-      ETH => COUNT
3810   6       0       500     i-      ETH IPV6 TCP => DROP COUNT
3811   5       0       1000    i-      ETH IPV6 ICMP => QUEUE
3812   1       24      0       i-      ETH IPV4 UDP => QUEUE
3813   4       24      10      i-      ETH IPV4 TCP => DROP
3814   3       24      20      i-      ETH IPV4 => DROP
3815   2       24      42      i-      ETH IPV4 UDP => QUEUE
3816   7       63      0       i-      ETH IPV6 UDP VXLAN => MARK QUEUE
3817   testpmd>
3818
3819Output can be limited to specific groups::
3820
3821   testpmd> flow list 1 group 0 group 63
3822   ID      Group   Prio    Attr    Rule
3823   0       0       0       i-      ETH => COUNT
3824   6       0       500     i-      ETH IPV6 TCP => DROP COUNT
3825   5       0       1000    i-      ETH IPV6 ICMP => QUEUE
3826   7       63      0       i-      ETH IPV6 UDP VXLAN => MARK QUEUE
3827   testpmd>
3828
3829Toggling isolated mode
3830~~~~~~~~~~~~~~~~~~~~~~
3831
3832``flow isolate`` can be used to tell the underlying PMD that ingress traffic
3833must only be injected from the defined flow rules; that no default traffic
3834is expected outside those rules and the driver is free to assign more
3835resources to handle them. It is bound to ``rte_flow_isolate()``::
3836
3837 flow isolate {port_id} {boolean}
3838
3839If successful, enabling or disabling isolated mode shows either::
3840
3841 Ingress traffic on port [...]
3842    is now restricted to the defined flow rules
3843
3844Or::
3845
3846 Ingress traffic on port [...]
3847    is not restricted anymore to the defined flow rules
3848
3849Otherwise, in case of error::
3850
3851   Caught error type [...] ([...]): [...]
3852
3853Mainly due to its side effects, PMDs supporting this mode may not have the
3854ability to toggle it more than once without reinitializing affected ports
3855first (e.g. by exiting testpmd).
3856
3857Enabling isolated mode::
3858
3859 testpmd> flow isolate 0 true
3860 Ingress traffic on port 0 is now restricted to the defined flow rules
3861 testpmd>
3862
3863Disabling isolated mode::
3864
3865 testpmd> flow isolate 0 false
3866 Ingress traffic on port 0 is not restricted anymore to the defined flow rules
3867 testpmd>
3868
3869Sample QinQ flow rules
3870~~~~~~~~~~~~~~~~~~~~~~
3871
3872Before creating QinQ rule(s) the following commands should be issued to enable QinQ::
3873
3874   testpmd> port stop 0
3875   testpmd> vlan set qinq on 0
3876
3877The above command sets the inner and outer TPID's to 0x8100.
3878
3879To change the TPID's the following commands should be used::
3880
3881   testpmd> vlan set outer tpid 0xa100 0
3882   testpmd> vlan set inner tpid 0x9100 0
3883   testpmd> port start 0
3884
3885Validate and create a QinQ rule on port 0 to steer traffic to a VF queue in a VM.
3886
3887::
3888
3889   testpmd> flow validate 0 ingress pattern eth / vlan tci is 123 /
3890       vlan tci is 456 / end actions vf id 1 / queue index 0 / end
3891   Flow rule #0 validated
3892
3893   testpmd> flow create 0 ingress pattern eth / vlan tci is 4 /
3894       vlan tci is 456 / end actions vf id 123 / queue index 0 / end
3895   Flow rule #0 created
3896
3897   testpmd> flow list 0
3898   ID      Group   Prio    Attr    Rule
3899   0       0       0       i-      ETH VLAN VLAN=>VF QUEUE
3900
3901Validate and create a QinQ rule on port 0 to steer traffic to a queue on the host.
3902
3903::
3904
3905   testpmd> flow validate 0 ingress pattern eth / vlan tci is 321 /
3906        vlan tci is 654 / end actions pf / queue index 0 / end
3907   Flow rule #1 validated
3908
3909   testpmd> flow create 0 ingress pattern eth / vlan tci is 321 /
3910        vlan tci is 654 / end actions pf / queue index 1 / end
3911   Flow rule #1 created
3912
3913   testpmd> flow list 0
3914   ID      Group   Prio    Attr    Rule
3915   0       0       0       i-      ETH VLAN VLAN=>VF QUEUE
3916   1       0       0       i-      ETH VLAN VLAN=>PF QUEUE
3917
3918BPF Functions
3919--------------
3920
3921The following sections show functions to load/unload eBPF based filters.
3922
3923bpf-load
3924~~~~~~~~
3925
3926Load an eBPF program as a callback for partciular RX/TX queue::
3927
3928   testpmd> bpf-load rx|tx (portid) (queueid) (load-flags) (bpf-prog-filename)
3929
3930The available load-flags are:
3931
3932* ``J``: use JIT generated native code, otherwise BPF interpreter will be used.
3933
3934* ``M``: assume input parameter is a pointer to rte_mbuf, otherwise assume it is a pointer to first segment's data.
3935
3936* ``-``: none.
3937
3938.. note::
3939
3940   You'll need clang v3.7 or above to build bpf program you'd like to load
3941
3942For example:
3943
3944.. code-block:: console
3945
3946   cd test/bpf
3947   clang -O2 -target bpf -c t1.c
3948
3949Then to load (and JIT compile) t1.o at RX queue 0, port 1::
3950
3951.. code-block:: console
3952
3953   testpmd> bpf-load rx 1 0 J ./dpdk.org/test/bpf/t1.o
3954
3955To load (not JITed) t1.o at TX queue 0, port 0::
3956
3957.. code-block:: console
3958
3959   testpmd> bpf-load tx 0 0 - ./dpdk.org/test/bpf/t1.o
3960
3961bpf-unload
3962~~~~~~~~~~
3963
3964Unload previously loaded eBPF program for partciular RX/TX queue::
3965
3966   testpmd> bpf-unload rx|tx (portid) (queueid)
3967
3968For example to unload BPF filter from TX queue 0, port 0:
3969
3970.. code-block:: console
3971
3972   testpmd> bpf-load tx 0 0 - ./dpdk.org/test/bpf/t1.o
3973