xref: /dpdk/doc/guides/testpmd_app_ug/testpmd_funcs.rst (revision dcd962fc6b4e6012bd22710bc94ac244b3aaf7d7)
1..  SPDX-License-Identifier: BSD-3-Clause
2    Copyright(c) 2010-2016 Intel Corporation.
3
4.. _testpmd_runtime:
5
6Testpmd Runtime Functions
7=========================
8
9Where the testpmd application is started in interactive mode, (``-i|--interactive``),
10it displays a prompt that can be used to start and stop forwarding,
11configure the application, display statistics (including the extended NIC
12statistics aka xstats) , set the Flow Director and other tasks::
13
14   testpmd>
15
16The testpmd prompt has some, limited, readline support.
17Common bash command-line functions such as ``Ctrl+a`` and ``Ctrl+e`` to go to the start and end of the prompt line are supported
18as well as access to the command history via the up-arrow.
19
20There is also support for tab completion.
21If you type a partial command and hit ``<TAB>`` you get a list of the available completions:
22
23.. code-block:: console
24
25   testpmd> show port <TAB>
26
27       info [Mul-choice STRING]: show|clear port info|stats|xstats|fdir|stat_qmap|dcb_tc|cap X
28       info [Mul-choice STRING]: show|clear port info|stats|xstats|fdir|stat_qmap|dcb_tc|cap all
29       stats [Mul-choice STRING]: show|clear port info|stats|xstats|fdir|stat_qmap|dcb_tc|cap X
30       stats [Mul-choice STRING]: show|clear port info|stats|xstats|fdir|stat_qmap|dcb_tc|cap all
31       ...
32
33
34.. note::
35
36   Some examples in this document are too long to fit on one line are are shown wrapped at `"\\"` for display purposes::
37
38      testpmd> set flow_ctrl rx (on|off) tx (on|off) (high_water) (low_water) \
39               (pause_time) (send_xon) (port_id)
40
41In the real ``testpmd>`` prompt these commands should be on a single line.
42
43Help Functions
44--------------
45
46The testpmd has on-line help for the functions that are available at runtime.
47These are divided into sections and can be accessed using help, help section or help all:
48
49.. code-block:: console
50
51   testpmd> help
52
53       help control    : Start and stop forwarding.
54       help display    : Displaying port, stats and config information.
55       help config     : Configuration information.
56       help ports      : Configuring ports.
57       help registers  : Reading and setting port registers.
58       help filters    : Filters configuration help.
59       help all        : All of the above sections.
60
61
62Command File Functions
63----------------------
64
65To facilitate loading large number of commands or to avoid cutting and pasting where not
66practical or possible testpmd supports alternative methods for executing commands.
67
68* If started with the ``--cmdline-file=FILENAME`` command line argument testpmd
69  will execute all CLI commands contained within the file immediately before
70  starting packet forwarding or entering interactive mode.
71
72.. code-block:: console
73
74   ./testpmd -n4 -r2 ... -- -i --cmdline-file=/home/ubuntu/flow-create-commands.txt
75   Interactive-mode selected
76   CLI commands to be read from /home/ubuntu/flow-create-commands.txt
77   Configuring Port 0 (socket 0)
78   Port 0: 7C:FE:90:CB:74:CE
79   Configuring Port 1 (socket 0)
80   Port 1: 7C:FE:90:CB:74:CA
81   Checking link statuses...
82   Port 0 Link Up - speed 10000 Mbps - full-duplex
83   Port 1 Link Up - speed 10000 Mbps - full-duplex
84   Done
85   Flow rule #0 created
86   Flow rule #1 created
87   ...
88   ...
89   Flow rule #498 created
90   Flow rule #499 created
91   Read all CLI commands from /home/ubuntu/flow-create-commands.txt
92   testpmd>
93
94
95* At run-time additional commands can be loaded in bulk by invoking the ``load FILENAME``
96  command.
97
98.. code-block:: console
99
100   testpmd> load /home/ubuntu/flow-create-commands.txt
101   Flow rule #0 created
102   Flow rule #1 created
103   ...
104   ...
105   Flow rule #498 created
106   Flow rule #499 created
107   Read all CLI commands from /home/ubuntu/flow-create-commands.txt
108   testpmd>
109
110
111In all cases output from any included command will be displayed as standard output.
112Execution will continue until the end of the file is reached regardless of
113whether any errors occur.  The end user must examine the output to determine if
114any failures occurred.
115
116
117Control Functions
118-----------------
119
120start
121~~~~~
122
123Start packet forwarding with current configuration::
124
125   testpmd> start
126
127start tx_first
128~~~~~~~~~~~~~~
129
130Start packet forwarding with current configuration after sending specified number of bursts of packets::
131
132   testpmd> start tx_first (""|burst_num)
133
134The default burst number is 1 when ``burst_num`` not presented.
135
136stop
137~~~~
138
139Stop packet forwarding, and display accumulated statistics::
140
141   testpmd> stop
142
143quit
144~~~~
145
146Quit to prompt::
147
148   testpmd> quit
149
150
151Display Functions
152-----------------
153
154The functions in the following sections are used to display information about the
155testpmd configuration or the NIC status.
156
157show port
158~~~~~~~~~
159
160Display information for a given port or all ports::
161
162   testpmd> show port (info|stats|xstats|fdir|stat_qmap|dcb_tc|cap) (port_id|all)
163
164The available information categories are:
165
166* ``info``: General port information such as MAC address.
167
168* ``stats``: RX/TX statistics.
169
170* ``xstats``: RX/TX extended NIC statistics.
171
172* ``fdir``: Flow Director information and statistics.
173
174* ``stat_qmap``: Queue statistics mapping.
175
176* ``dcb_tc``: DCB information such as TC mapping.
177
178* ``cap``: Supported offload capabilities.
179
180For example:
181
182.. code-block:: console
183
184   testpmd> show port info 0
185
186   ********************* Infos for port 0 *********************
187
188   MAC address: XX:XX:XX:XX:XX:XX
189   Connect to socket: 0
190   memory allocation on the socket: 0
191   Link status: up
192   Link speed: 40000 Mbps
193   Link duplex: full-duplex
194   Promiscuous mode: enabled
195   Allmulticast mode: disabled
196   Maximum number of MAC addresses: 64
197   Maximum number of MAC addresses of hash filtering: 0
198   VLAN offload:
199       strip on
200       filter on
201       qinq(extend) off
202   Redirection table size: 512
203   Supported flow types:
204     ipv4-frag
205     ipv4-tcp
206     ipv4-udp
207     ipv4-sctp
208     ipv4-other
209     ipv6-frag
210     ipv6-tcp
211     ipv6-udp
212     ipv6-sctp
213     ipv6-other
214     l2_payload
215     port
216     vxlan
217     geneve
218     nvgre
219
220show port rss reta
221~~~~~~~~~~~~~~~~~~
222
223Display the rss redirection table entry indicated by masks on port X::
224
225   testpmd> show port (port_id) rss reta (size) (mask0, mask1...)
226
227size is used to indicate the hardware supported reta size
228
229show port rss-hash
230~~~~~~~~~~~~~~~~~~
231
232Display the RSS hash functions and RSS hash key of a port::
233
234   testpmd> show port (port_id) rss-hash ipv4|ipv4-frag|ipv4-tcp|ipv4-udp|ipv4-sctp|ipv4-other|ipv6|ipv6-frag|ipv6-tcp|ipv6-udp|ipv6-sctp|ipv6-other|l2-payload|ipv6-ex|ipv6-tcp-ex|ipv6-udp-ex [key]
235
236clear port
237~~~~~~~~~~
238
239Clear the port statistics for a given port or for all ports::
240
241   testpmd> clear port (info|stats|xstats|fdir|stat_qmap) (port_id|all)
242
243For example::
244
245   testpmd> clear port stats all
246
247show (rxq|txq)
248~~~~~~~~~~~~~~
249
250Display information for a given port's RX/TX queue::
251
252   testpmd> show (rxq|txq) info (port_id) (queue_id)
253
254show config
255~~~~~~~~~~~
256
257Displays the configuration of the application.
258The configuration comes from the command-line, the runtime or the application defaults::
259
260   testpmd> show config (rxtx|cores|fwd|txpkts)
261
262The available information categories are:
263
264* ``rxtx``: RX/TX configuration items.
265
266* ``cores``: List of forwarding cores.
267
268* ``fwd``: Packet forwarding configuration.
269
270* ``txpkts``: Packets to TX configuration.
271
272For example:
273
274.. code-block:: console
275
276   testpmd> show config rxtx
277
278   io packet forwarding - CRC stripping disabled - packets/burst=16
279   nb forwarding cores=2 - nb forwarding ports=1
280   RX queues=1 - RX desc=128 - RX free threshold=0
281   RX threshold registers: pthresh=8 hthresh=8 wthresh=4
282   TX queues=1 - TX desc=512 - TX free threshold=0
283   TX threshold registers: pthresh=36 hthresh=0 wthresh=0
284   TX RS bit threshold=0 - TXQ flags=0x0
285
286set fwd
287~~~~~~~
288
289Set the packet forwarding mode::
290
291   testpmd> set fwd (io|mac|macswap|flowgen| \
292                     rxonly|txonly|csum|icmpecho) (""|retry)
293
294``retry`` can be specified for forwarding engines except ``rx_only``.
295
296The available information categories are:
297
298* ``io``: Forwards packets "as-is" in I/O mode.
299  This is the fastest possible forwarding operation as it does not access packets data.
300  This is the default mode.
301
302* ``mac``: Changes the source and the destination Ethernet addresses of packets before forwarding them.
303  Default application behaviour is to set source Ethernet address to that of the transmitting interface, and destination
304  address to a dummy value (set during init). The user may specify a target destination Ethernet address via the 'eth-peer' or
305  'eth-peer-configfile' command-line options. It is not currently possible to specify a specific source Ethernet address.
306
307* ``macswap``: MAC swap forwarding mode.
308  Swaps the source and the destination Ethernet addresses of packets before forwarding them.
309
310* ``flowgen``: Multi-flow generation mode.
311  Originates a number of flows (with varying destination IP addresses), and terminate receive traffic.
312
313* ``rxonly``: Receives packets but doesn't transmit them.
314
315* ``txonly``: Generates and transmits packets without receiving any.
316
317* ``csum``: Changes the checksum field with hardware or software methods depending on the offload flags on the packet.
318
319* ``icmpecho``: Receives a burst of packets, lookup for IMCP echo requests and, if any, send back ICMP echo replies.
320
321* ``ieee1588``: Demonstrate L2 IEEE1588 V2 PTP timestamping for RX and TX. Requires ``CONFIG_RTE_LIBRTE_IEEE1588=y``.
322
323* ``tm``: Traffic Management forwarding mode
324  Demonstrates the use of ethdev traffic management APIs and softnic PMD for
325  QoS traffic management. In this mode, 5-level hierarchical QoS scheduler is
326  available as an default option that can be enabled through CLI. The user can
327  also modify the default hierarchy or specify the new hierarchy through CLI for
328  implementing QoS scheduler.  Requires ``CONFIG_RTE_LIBRTE_PMD_SOFTNIC=y`` ``CONFIG_RTE_LIBRTE_SCHED=y``.
329
330Example::
331
332   testpmd> set fwd rxonly
333
334   Set rxonly packet forwarding mode
335
336
337read rxd
338~~~~~~~~
339
340Display an RX descriptor for a port RX queue::
341
342   testpmd> read rxd (port_id) (queue_id) (rxd_id)
343
344For example::
345
346   testpmd> read rxd 0 0 4
347        0x0000000B - 0x001D0180 / 0x0000000B - 0x001D0180
348
349read txd
350~~~~~~~~
351
352Display a TX descriptor for a port TX queue::
353
354   testpmd> read txd (port_id) (queue_id) (txd_id)
355
356For example::
357
358   testpmd> read txd 0 0 4
359        0x00000001 - 0x24C3C440 / 0x000F0000 - 0x2330003C
360
361ddp get list
362~~~~~~~~~~~~
363
364Get loaded dynamic device personalization (DDP) package info list::
365
366   testpmd> ddp get list (port_id)
367
368ddp get info
369~~~~~~~~~~~~
370
371Display information about dynamic device personalization (DDP) profile::
372
373   testpmd> ddp get info (profile_path)
374
375show vf stats
376~~~~~~~~~~~~~
377
378Display VF statistics::
379
380   testpmd> show vf stats (port_id) (vf_id)
381
382clear vf stats
383~~~~~~~~~~~~~~
384
385Reset VF statistics::
386
387   testpmd> clear vf stats (port_id) (vf_id)
388
389show port pctype mapping
390~~~~~~~~~~~~~~~~~~~~~~~~
391
392List all items from the pctype mapping table::
393
394   testpmd> show port (port_id) pctype mapping
395
396show rx offloading capabilities
397~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
398
399List all per queue and per port Rx offloading capabilities of a port::
400
401   testpmd> show port (port_id) rx_offload capabilities
402
403show rx offloading configuration
404~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
405
406List port level and all queue level Rx offloading configuration::
407
408   testpmd> show port (port_id) rx_offload configuration
409
410show tx offloading capabilities
411~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
412
413List all per queue and per port Tx offloading capabilities of a port::
414
415   testpmd> show port (port_id) tx_offload capabilities
416
417show tx offloading configuration
418~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
419
420List port level and all queue level Tx offloading configuration::
421
422   testpmd> show port (port_id) tx_offload configuration
423
424
425Configuration Functions
426-----------------------
427
428The testpmd application can be configured from the runtime as well as from the command-line.
429
430This section details the available configuration functions that are available.
431
432.. note::
433
434   Configuration changes only become active when forwarding is started/restarted.
435
436set default
437~~~~~~~~~~~
438
439Reset forwarding to the default configuration::
440
441   testpmd> set default
442
443set verbose
444~~~~~~~~~~~
445
446Set the debug verbosity level::
447
448   testpmd> set verbose (level)
449
450Currently the only available levels are 0 (silent except for error) and 1 (fully verbose).
451
452set log
453~~~~~~~
454
455Set the log level for a log type::
456
457	testpmd> set log global|(type) (level)
458
459Where:
460
461* ``type`` is the log name.
462
463* ``level`` is the log level.
464
465For example, to change the global log level::
466	testpmd> set log global (level)
467
468Regexes can also be used for type. To change log level of user1, user2 and user3::
469	testpmd> set log user[1-3] (level)
470
471set nbport
472~~~~~~~~~~
473
474Set the number of ports used by the application:
475
476set nbport (num)
477
478This is equivalent to the ``--nb-ports`` command-line option.
479
480set nbcore
481~~~~~~~~~~
482
483Set the number of cores used by the application::
484
485   testpmd> set nbcore (num)
486
487This is equivalent to the ``--nb-cores`` command-line option.
488
489.. note::
490
491   The number of cores used must not be greater than number of ports used multiplied by the number of queues per port.
492
493set coremask
494~~~~~~~~~~~~
495
496Set the forwarding cores hexadecimal mask::
497
498   testpmd> set coremask (mask)
499
500This is equivalent to the ``--coremask`` command-line option.
501
502.. note::
503
504   The master lcore is reserved for command line parsing only and cannot be masked on for packet forwarding.
505
506set portmask
507~~~~~~~~~~~~
508
509Set the forwarding ports hexadecimal mask::
510
511   testpmd> set portmask (mask)
512
513This is equivalent to the ``--portmask`` command-line option.
514
515set burst
516~~~~~~~~~
517
518Set number of packets per burst::
519
520   testpmd> set burst (num)
521
522This is equivalent to the ``--burst command-line`` option.
523
524When retry is enabled, the transmit delay time and number of retries can also be set::
525
526   testpmd> set burst tx delay (microseconds) retry (num)
527
528set txpkts
529~~~~~~~~~~
530
531Set the length of each segment of the TX-ONLY packets or length of packet for FLOWGEN mode::
532
533   testpmd> set txpkts (x[,y]*)
534
535Where x[,y]* represents a CSV list of values, without white space.
536
537set txsplit
538~~~~~~~~~~~
539
540Set the split policy for the TX packets, applicable for TX-ONLY and CSUM forwarding modes::
541
542   testpmd> set txsplit (off|on|rand)
543
544Where:
545
546* ``off`` disable packet copy & split for CSUM mode.
547
548* ``on`` split outgoing packet into multiple segments. Size of each segment
549  and number of segments per packet is determined by ``set txpkts`` command
550  (see above).
551
552* ``rand`` same as 'on', but number of segments per each packet is a random value between 1 and total number of segments.
553
554set corelist
555~~~~~~~~~~~~
556
557Set the list of forwarding cores::
558
559   testpmd> set corelist (x[,y]*)
560
561For example, to change the forwarding cores:
562
563.. code-block:: console
564
565   testpmd> set corelist 3,1
566   testpmd> show config fwd
567
568   io packet forwarding - ports=2 - cores=2 - streams=2 - NUMA support disabled
569   Logical Core 3 (socket 0) forwards packets on 1 streams:
570   RX P=0/Q=0 (socket 0) -> TX P=1/Q=0 (socket 0) peer=02:00:00:00:00:01
571   Logical Core 1 (socket 0) forwards packets on 1 streams:
572   RX P=1/Q=0 (socket 0) -> TX P=0/Q=0 (socket 0) peer=02:00:00:00:00:00
573
574.. note::
575
576   The cores are used in the same order as specified on the command line.
577
578set portlist
579~~~~~~~~~~~~
580
581Set the list of forwarding ports::
582
583   testpmd> set portlist (x[,y]*)
584
585For example, to change the port forwarding:
586
587.. code-block:: console
588
589   testpmd> set portlist 0,2,1,3
590   testpmd> show config fwd
591
592   io packet forwarding - ports=4 - cores=1 - streams=4
593   Logical Core 3 (socket 0) forwards packets on 4 streams:
594   RX P=0/Q=0 (socket 0) -> TX P=2/Q=0 (socket 0) peer=02:00:00:00:00:01
595   RX P=2/Q=0 (socket 0) -> TX P=0/Q=0 (socket 0) peer=02:00:00:00:00:00
596   RX P=1/Q=0 (socket 0) -> TX P=3/Q=0 (socket 0) peer=02:00:00:00:00:03
597   RX P=3/Q=0 (socket 0) -> TX P=1/Q=0 (socket 0) peer=02:00:00:00:00:02
598
599set tx loopback
600~~~~~~~~~~~~~~~
601
602Enable/disable tx loopback::
603
604   testpmd> set tx loopback (port_id) (on|off)
605
606set drop enable
607~~~~~~~~~~~~~~~
608
609set drop enable bit for all queues::
610
611   testpmd> set all queues drop (port_id) (on|off)
612
613set split drop enable (for VF)
614~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
615
616set split drop enable bit for VF from PF::
617
618   testpmd> set vf split drop (port_id) (vf_id) (on|off)
619
620set mac antispoof (for VF)
621~~~~~~~~~~~~~~~~~~~~~~~~~~
622
623Set mac antispoof for a VF from the PF::
624
625   testpmd> set vf mac antispoof  (port_id) (vf_id) (on|off)
626
627set macsec offload
628~~~~~~~~~~~~~~~~~~
629
630Enable/disable MACsec offload::
631
632   testpmd> set macsec offload (port_id) on encrypt (on|off) replay-protect (on|off)
633   testpmd> set macsec offload (port_id) off
634
635set macsec sc
636~~~~~~~~~~~~~
637
638Configure MACsec secure connection (SC)::
639
640   testpmd> set macsec sc (tx|rx) (port_id) (mac) (pi)
641
642.. note::
643
644   The pi argument is ignored for tx.
645   Check the NIC Datasheet for hardware limits.
646
647set macsec sa
648~~~~~~~~~~~~~
649
650Configure MACsec secure association (SA)::
651
652   testpmd> set macsec sa (tx|rx) (port_id) (idx) (an) (pn) (key)
653
654.. note::
655
656   The IDX value must be 0 or 1.
657   Check the NIC Datasheet for hardware limits.
658
659set broadcast mode (for VF)
660~~~~~~~~~~~~~~~~~~~~~~~~~~~
661
662Set broadcast mode for a VF from the PF::
663
664   testpmd> set vf broadcast (port_id) (vf_id) (on|off)
665
666vlan set strip
667~~~~~~~~~~~~~~
668
669Set the VLAN strip on a port::
670
671   testpmd> vlan set strip (on|off) (port_id)
672
673vlan set stripq
674~~~~~~~~~~~~~~~
675
676Set the VLAN strip for a queue on a port::
677
678   testpmd> vlan set stripq (on|off) (port_id,queue_id)
679
680vlan set stripq (for VF)
681~~~~~~~~~~~~~~~~~~~~~~~~
682
683Set VLAN strip for all queues in a pool for a VF from the PF::
684
685   testpmd> set vf vlan stripq (port_id) (vf_id) (on|off)
686
687vlan set insert (for VF)
688~~~~~~~~~~~~~~~~~~~~~~~~
689
690Set VLAN insert for a VF from the PF::
691
692   testpmd> set vf vlan insert (port_id) (vf_id) (vlan_id)
693
694vlan set tag (for VF)
695~~~~~~~~~~~~~~~~~~~~~
696
697Set VLAN tag for a VF from the PF::
698
699   testpmd> set vf vlan tag (port_id) (vf_id) (on|off)
700
701vlan set antispoof (for VF)
702~~~~~~~~~~~~~~~~~~~~~~~~~~~
703
704Set VLAN antispoof for a VF from the PF::
705
706   testpmd> set vf vlan antispoof (port_id) (vf_id) (on|off)
707
708vlan set filter
709~~~~~~~~~~~~~~~
710
711Set the VLAN filter on a port::
712
713   testpmd> vlan set filter (on|off) (port_id)
714
715vlan set qinq
716~~~~~~~~~~~~~
717
718Set the VLAN QinQ (extended queue in queue) on for a port::
719
720   testpmd> vlan set qinq (on|off) (port_id)
721
722vlan set tpid
723~~~~~~~~~~~~~
724
725Set the inner or outer VLAN TPID for packet filtering on a port::
726
727   testpmd> vlan set (inner|outer) tpid (value) (port_id)
728
729.. note::
730
731   TPID value must be a 16-bit number (value <= 65536).
732
733rx_vlan add
734~~~~~~~~~~~
735
736Add a VLAN ID, or all identifiers, to the set of VLAN identifiers filtered by port ID::
737
738   testpmd> rx_vlan add (vlan_id|all) (port_id)
739
740.. note::
741
742   VLAN filter must be set on that port. VLAN ID < 4096.
743   Depending on the NIC used, number of vlan_ids may be limited to the maximum entries
744   in VFTA table. This is important if enabling all vlan_ids.
745
746rx_vlan rm
747~~~~~~~~~~
748
749Remove a VLAN ID, or all identifiers, from the set of VLAN identifiers filtered by port ID::
750
751   testpmd> rx_vlan rm (vlan_id|all) (port_id)
752
753rx_vlan add (for VF)
754~~~~~~~~~~~~~~~~~~~~
755
756Add a VLAN ID, to the set of VLAN identifiers filtered for VF(s) for port ID::
757
758   testpmd> rx_vlan add (vlan_id) port (port_id) vf (vf_mask)
759
760rx_vlan rm (for VF)
761~~~~~~~~~~~~~~~~~~~
762
763Remove a VLAN ID, from the set of VLAN identifiers filtered for VF(s) for port ID::
764
765   testpmd> rx_vlan rm (vlan_id) port (port_id) vf (vf_mask)
766
767tunnel_filter add
768~~~~~~~~~~~~~~~~~
769
770Add a tunnel filter on a port::
771
772   testpmd> tunnel_filter add (port_id) (outer_mac) (inner_mac) (ip_addr) \
773            (inner_vlan) (vxlan|nvgre|ipingre) (imac-ivlan|imac-ivlan-tenid|\
774            imac-tenid|imac|omac-imac-tenid|oip|iip) (tenant_id) (queue_id)
775
776The available information categories are:
777
778* ``vxlan``: Set tunnel type as VXLAN.
779
780* ``nvgre``: Set tunnel type as NVGRE.
781
782* ``ipingre``: Set tunnel type as IP-in-GRE.
783
784* ``imac-ivlan``: Set filter type as Inner MAC and VLAN.
785
786* ``imac-ivlan-tenid``: Set filter type as Inner MAC, VLAN and tenant ID.
787
788* ``imac-tenid``: Set filter type as Inner MAC and tenant ID.
789
790* ``imac``: Set filter type as Inner MAC.
791
792* ``omac-imac-tenid``: Set filter type as Outer MAC, Inner MAC and tenant ID.
793
794* ``oip``: Set filter type as Outer IP.
795
796* ``iip``: Set filter type as Inner IP.
797
798Example::
799
800   testpmd> tunnel_filter add 0 68:05:CA:28:09:82 00:00:00:00:00:00 \
801            192.168.2.2 0 ipingre oip 1 1
802
803   Set an IP-in-GRE tunnel on port 0, and the filter type is Outer IP.
804
805tunnel_filter remove
806~~~~~~~~~~~~~~~~~~~~
807
808Remove a tunnel filter on a port::
809
810   testpmd> tunnel_filter rm (port_id) (outer_mac) (inner_mac) (ip_addr) \
811            (inner_vlan) (vxlan|nvgre|ipingre) (imac-ivlan|imac-ivlan-tenid|\
812            imac-tenid|imac|omac-imac-tenid|oip|iip) (tenant_id) (queue_id)
813
814rx_vxlan_port add
815~~~~~~~~~~~~~~~~~
816
817Add an UDP port for VXLAN packet filter on a port::
818
819   testpmd> rx_vxlan_port add (udp_port) (port_id)
820
821rx_vxlan_port remove
822~~~~~~~~~~~~~~~~~~~~
823
824Remove an UDP port for VXLAN packet filter on a port::
825
826   testpmd> rx_vxlan_port rm (udp_port) (port_id)
827
828tx_vlan set
829~~~~~~~~~~~
830
831Set hardware insertion of VLAN IDs in packets sent on a port::
832
833   testpmd> tx_vlan set (port_id) vlan_id[, vlan_id_outer]
834
835For example, set a single VLAN ID (5) insertion on port 0::
836
837   tx_vlan set 0 5
838
839Or, set double VLAN ID (inner: 2, outer: 3) insertion on port 1::
840
841   tx_vlan set 1 2 3
842
843
844tx_vlan set pvid
845~~~~~~~~~~~~~~~~
846
847Set port based hardware insertion of VLAN ID in packets sent on a port::
848
849   testpmd> tx_vlan set pvid (port_id) (vlan_id) (on|off)
850
851tx_vlan reset
852~~~~~~~~~~~~~
853
854Disable hardware insertion of a VLAN header in packets sent on a port::
855
856   testpmd> tx_vlan reset (port_id)
857
858csum set
859~~~~~~~~
860
861Select hardware or software calculation of the checksum when
862transmitting a packet using the ``csum`` forwarding engine::
863
864   testpmd> csum set (ip|udp|tcp|sctp|outer-ip) (hw|sw) (port_id)
865
866Where:
867
868* ``ip|udp|tcp|sctp`` always relate to  the inner layer.
869
870* ``outer-ip`` relates to the outer IP layer (only for IPv4) in the case where the packet is recognized
871  as a tunnel packet by the forwarding engine (vxlan, gre and ipip are
872  supported). See also the ``csum parse-tunnel`` command.
873
874.. note::
875
876   Check the NIC Datasheet for hardware limits.
877
878RSS queue region
879~~~~~~~~~~~~~~~~
880
881Set RSS queue region span on a port::
882
883   testpmd> set port (port_id) queue-region region_id (value) \
884		queue_start_index (value) queue_num (value)
885
886Set flowtype mapping on a RSS queue region on a port::
887
888   testpmd> set port (port_id) queue-region region_id (value) flowtype (value)
889
890where:
891
892* For the flowtype(pctype) of packet,the specific index for each type has
893  been defined in file i40e_type.h as enum i40e_filter_pctype.
894
895Set user priority mapping on a RSS queue region on a port::
896
897   testpmd> set port (port_id) queue-region UP (value) region_id (value)
898
899Flush all queue region related configuration on a port::
900
901   testpmd> set port (port_id) queue-region flush (on|off)
902
903where:
904
905* "on"is just an enable function which server for other configuration,
906  it is for all configuration about queue region from up layer,
907  at first will only keep in DPDK softwarestored in driver,
908  only after "flush on", it commit all configuration to HW.
909  "off" is just clean all configuration about queue region just now,
910  and restore all to DPDK i40e driver default config when start up.
911
912Show all queue region related configuration info on a port::
913
914   testpmd> show port (port_id) queue-region
915
916.. note::
917
918  Queue region only support on PF by now, so these command is
919  only for configuration of queue region on PF port.
920
921csum parse-tunnel
922~~~~~~~~~~~~~~~~~
923
924Define how tunneled packets should be handled by the csum forward
925engine::
926
927   testpmd> csum parse-tunnel (on|off) (tx_port_id)
928
929If enabled, the csum forward engine will try to recognize supported
930tunnel headers (vxlan, gre, ipip).
931
932If disabled, treat tunnel packets as non-tunneled packets (a inner
933header is handled as a packet payload).
934
935.. note::
936
937   The port argument is the TX port like in the ``csum set`` command.
938
939Example:
940
941Consider a packet in packet like the following::
942
943   eth_out/ipv4_out/udp_out/vxlan/eth_in/ipv4_in/tcp_in
944
945* If parse-tunnel is enabled, the ``ip|udp|tcp|sctp`` parameters of ``csum set``
946  command relate to the inner headers (here ``ipv4_in`` and ``tcp_in``), and the
947  ``outer-ip parameter`` relates to the outer headers (here ``ipv4_out``).
948
949* If parse-tunnel is disabled, the ``ip|udp|tcp|sctp`` parameters of ``csum  set``
950   command relate to the outer headers, here ``ipv4_out`` and ``udp_out``.
951
952csum show
953~~~~~~~~~
954
955Display tx checksum offload configuration::
956
957   testpmd> csum show (port_id)
958
959tso set
960~~~~~~~
961
962Enable TCP Segmentation Offload (TSO) in the ``csum`` forwarding engine::
963
964   testpmd> tso set (segsize) (port_id)
965
966.. note::
967
968   Check the NIC datasheet for hardware limits.
969
970tso show
971~~~~~~~~
972
973Display the status of TCP Segmentation Offload::
974
975   testpmd> tso show (port_id)
976
977set port - gro
978~~~~~~~~~~~~~~
979
980Enable or disable GRO in ``csum`` forwarding engine::
981
982   testpmd> set port <port_id> gro on|off
983
984If enabled, the csum forwarding engine will perform GRO on the TCP/IPv4
985packets received from the given port.
986
987If disabled, packets received from the given port won't be performed
988GRO. By default, GRO is disabled for all ports.
989
990.. note::
991
992   When enable GRO for a port, TCP/IPv4 packets received from the port
993   will be performed GRO. After GRO, all merged packets have bad
994   checksums, since the GRO library doesn't re-calculate checksums for
995   the merged packets. Therefore, if users want the merged packets to
996   have correct checksums, please select HW IP checksum calculation and
997   HW TCP checksum calculation for the port which the merged packets are
998   transmitted to.
999
1000show port - gro
1001~~~~~~~~~~~~~~~
1002
1003Display GRO configuration for a given port::
1004
1005   testpmd> show port <port_id> gro
1006
1007set gro flush
1008~~~~~~~~~~~~~
1009
1010Set the cycle to flush the GROed packets from reassembly tables::
1011
1012   testpmd> set gro flush <cycles>
1013
1014When enable GRO, the csum forwarding engine performs GRO on received
1015packets, and the GROed packets are stored in reassembly tables. Users
1016can use this command to determine when the GROed packets are flushed
1017from the reassembly tables.
1018
1019The ``cycles`` is measured in GRO operation times. The csum forwarding
1020engine flushes the GROed packets from the tables every ``cycles`` GRO
1021operations.
1022
1023By default, the value of ``cycles`` is 1, which means flush GROed packets
1024from the reassembly tables as soon as one GRO operation finishes. The value
1025of ``cycles`` should be in the range of 1 to ``GRO_MAX_FLUSH_CYCLES``.
1026
1027Please note that the large value of ``cycles`` may cause the poor TCP/IP
1028stack performance. Because the GROed packets are delayed to arrive the
1029stack, thus causing more duplicated ACKs and TCP retransmissions.
1030
1031set port - gso
1032~~~~~~~~~~~~~~
1033
1034Toggle per-port GSO support in ``csum`` forwarding engine::
1035
1036   testpmd> set port <port_id> gso on|off
1037
1038If enabled, the csum forwarding engine will perform GSO on supported IPv4
1039packets, transmitted on the given port.
1040
1041If disabled, packets transmitted on the given port will not undergo GSO.
1042By default, GSO is disabled for all ports.
1043
1044.. note::
1045
1046   When GSO is enabled on a port, supported IPv4 packets transmitted on that
1047   port undergo GSO. Afterwards, the segmented packets are represented by
1048   multi-segment mbufs; however, the csum forwarding engine doesn't calculation
1049   of checksums for GSO'd segments in SW. As a result, if users want correct
1050   checksums in GSO segments, they should enable HW checksum calculation for
1051   GSO-enabled ports.
1052
1053   For example, HW checksum calculation for VxLAN GSO'd packets may be enabled
1054   by setting the following options in the csum forwarding engine:
1055
1056   testpmd> csum set outer_ip hw <port_id>
1057
1058   testpmd> csum set ip hw <port_id>
1059
1060   testpmd> csum set tcp hw <port_id>
1061
1062   UDP GSO is the same as IP fragmentation, which treats the UDP header
1063   as the payload and does not modify it during segmentation. That is,
1064   after UDP GSO, only the first output fragment has the original UDP
1065   header. Therefore, users need to enable HW IP checksum calculation
1066   and SW UDP checksum calculation for GSO-enabled ports, if they want
1067   correct checksums for UDP/IPv4 packets.
1068
1069set gso segsz
1070~~~~~~~~~~~~~
1071
1072Set the maximum GSO segment size (measured in bytes), which includes the
1073packet header and the packet payload for GSO-enabled ports (global)::
1074
1075   testpmd> set gso segsz <length>
1076
1077show port - gso
1078~~~~~~~~~~~~~~~
1079
1080Display the status of Generic Segmentation Offload for a given port::
1081
1082   testpmd> show port <port_id> gso
1083
1084mac_addr add
1085~~~~~~~~~~~~
1086
1087Add an alternative MAC address to a port::
1088
1089   testpmd> mac_addr add (port_id) (XX:XX:XX:XX:XX:XX)
1090
1091mac_addr remove
1092~~~~~~~~~~~~~~~
1093
1094Remove a MAC address from a port::
1095
1096   testpmd> mac_addr remove (port_id) (XX:XX:XX:XX:XX:XX)
1097
1098mac_addr add (for VF)
1099~~~~~~~~~~~~~~~~~~~~~
1100
1101Add an alternative MAC address for a VF to a port::
1102
1103   testpmd> mac_add add port (port_id) vf (vf_id) (XX:XX:XX:XX:XX:XX)
1104
1105mac_addr set
1106~~~~~~~~~~~~
1107
1108Set the default MAC address for a port::
1109
1110   testpmd> mac_addr set (port_id) (XX:XX:XX:XX:XX:XX)
1111
1112mac_addr set (for VF)
1113~~~~~~~~~~~~~~~~~~~~~
1114
1115Set the MAC address for a VF from the PF::
1116
1117   testpmd> set vf mac addr (port_id) (vf_id) (XX:XX:XX:XX:XX:XX)
1118
1119set eth-peer
1120~~~~~~~~~~~~
1121
1122Set the forwarding peer address for certain port::
1123
1124   testpmd> set eth-peer (port_id) (perr_addr)
1125
1126This is equivalent to the ``--eth-peer`` command-line option.
1127
1128set port-uta
1129~~~~~~~~~~~~
1130
1131Set the unicast hash filter(s) on/off for a port::
1132
1133   testpmd> set port (port_id) uta (XX:XX:XX:XX:XX:XX|all) (on|off)
1134
1135set promisc
1136~~~~~~~~~~~
1137
1138Set the promiscuous mode on for a port or for all ports.
1139In promiscuous mode packets are not dropped if they aren't for the specified MAC address::
1140
1141   testpmd> set promisc (port_id|all) (on|off)
1142
1143set allmulti
1144~~~~~~~~~~~~
1145
1146Set the allmulti mode for a port or for all ports::
1147
1148   testpmd> set allmulti (port_id|all) (on|off)
1149
1150Same as the ifconfig (8) option. Controls how multicast packets are handled.
1151
1152set promisc (for VF)
1153~~~~~~~~~~~~~~~~~~~~
1154
1155Set the unicast promiscuous mode for a VF from PF.
1156It's supported by Intel i40e NICs now.
1157In promiscuous mode packets are not dropped if they aren't for the specified MAC address::
1158
1159   testpmd> set vf promisc (port_id) (vf_id) (on|off)
1160
1161set allmulticast (for VF)
1162~~~~~~~~~~~~~~~~~~~~~~~~~
1163
1164Set the multicast promiscuous mode for a VF from PF.
1165It's supported by Intel i40e NICs now.
1166In promiscuous mode packets are not dropped if they aren't for the specified MAC address::
1167
1168   testpmd> set vf allmulti (port_id) (vf_id) (on|off)
1169
1170set tx max bandwidth (for VF)
1171~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
1172
1173Set TX max absolute bandwidth (Mbps) for a VF from PF::
1174
1175   testpmd> set vf tx max-bandwidth (port_id) (vf_id) (max_bandwidth)
1176
1177set tc tx min bandwidth (for VF)
1178~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
1179
1180Set all TCs' TX min relative bandwidth (%) for a VF from PF::
1181
1182   testpmd> set vf tc tx min-bandwidth (port_id) (vf_id) (bw1, bw2, ...)
1183
1184set tc tx max bandwidth (for VF)
1185~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
1186
1187Set a TC's TX max absolute bandwidth (Mbps) for a VF from PF::
1188
1189   testpmd> set vf tc tx max-bandwidth (port_id) (vf_id) (tc_no) (max_bandwidth)
1190
1191set tc strict link priority mode
1192~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
1193
1194Set some TCs' strict link priority mode on a physical port::
1195
1196   testpmd> set tx strict-link-priority (port_id) (tc_bitmap)
1197
1198set tc tx min bandwidth
1199~~~~~~~~~~~~~~~~~~~~~~~
1200
1201Set all TCs' TX min relative bandwidth (%) globally for all PF and VFs::
1202
1203   testpmd> set tc tx min-bandwidth (port_id) (bw1, bw2, ...)
1204
1205set flow_ctrl rx
1206~~~~~~~~~~~~~~~~
1207
1208Set the link flow control parameter on a port::
1209
1210   testpmd> set flow_ctrl rx (on|off) tx (on|off) (high_water) (low_water) \
1211            (pause_time) (send_xon) mac_ctrl_frame_fwd (on|off) \
1212	    autoneg (on|off) (port_id)
1213
1214Where:
1215
1216* ``high_water`` (integer): High threshold value to trigger XOFF.
1217
1218* ``low_water`` (integer): Low threshold value to trigger XON.
1219
1220* ``pause_time`` (integer): Pause quota in the Pause frame.
1221
1222* ``send_xon`` (0/1): Send XON frame.
1223
1224* ``mac_ctrl_frame_fwd``: Enable receiving MAC control frames.
1225
1226* ``autoneg``: Change the auto-negotiation parameter.
1227
1228set pfc_ctrl rx
1229~~~~~~~~~~~~~~~
1230
1231Set the priority flow control parameter on a port::
1232
1233   testpmd> set pfc_ctrl rx (on|off) tx (on|off) (high_water) (low_water) \
1234            (pause_time) (priority) (port_id)
1235
1236Where:
1237
1238* ``high_water`` (integer): High threshold value.
1239
1240* ``low_water`` (integer): Low threshold value.
1241
1242* ``pause_time`` (integer): Pause quota in the Pause frame.
1243
1244* ``priority`` (0-7): VLAN User Priority.
1245
1246set stat_qmap
1247~~~~~~~~~~~~~
1248
1249Set statistics mapping (qmapping 0..15) for RX/TX queue on port::
1250
1251   testpmd> set stat_qmap (tx|rx) (port_id) (queue_id) (qmapping)
1252
1253For example, to set rx queue 2 on port 0 to mapping 5::
1254
1255   testpmd>set stat_qmap rx 0 2 5
1256
1257set xstats-hide-zero
1258~~~~~~~~~~~~~~~~~~~~
1259
1260Set the option to hide zero values for xstats display::
1261
1262	testpmd> set xstats-hide-zero on|off
1263
1264.. note::
1265
1266	By default, the zero values are displayed for xstats.
1267
1268set port - rx/tx (for VF)
1269~~~~~~~~~~~~~~~~~~~~~~~~~
1270
1271Set VF receive/transmit from a port::
1272
1273   testpmd> set port (port_id) vf (vf_id) (rx|tx) (on|off)
1274
1275set port - mac address filter (for VF)
1276~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
1277
1278Add/Remove unicast or multicast MAC addr filter for a VF::
1279
1280   testpmd> set port (port_id) vf (vf_id) (mac_addr) \
1281            (exact-mac|exact-mac-vlan|hashmac|hashmac-vlan) (on|off)
1282
1283set port - rx mode(for VF)
1284~~~~~~~~~~~~~~~~~~~~~~~~~~
1285
1286Set the VF receive mode of a port::
1287
1288   testpmd> set port (port_id) vf (vf_id) \
1289            rxmode (AUPE|ROPE|BAM|MPE) (on|off)
1290
1291The available receive modes are:
1292
1293* ``AUPE``: Accepts untagged VLAN.
1294
1295* ``ROPE``: Accepts unicast hash.
1296
1297* ``BAM``: Accepts broadcast packets.
1298
1299* ``MPE``: Accepts all multicast packets.
1300
1301set port - tx_rate (for Queue)
1302~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
1303
1304Set TX rate limitation for a queue on a port::
1305
1306   testpmd> set port (port_id) queue (queue_id) rate (rate_value)
1307
1308set port - tx_rate (for VF)
1309~~~~~~~~~~~~~~~~~~~~~~~~~~~
1310
1311Set TX rate limitation for queues in VF on a port::
1312
1313   testpmd> set port (port_id) vf (vf_id) rate (rate_value) queue_mask (queue_mask)
1314
1315set port - mirror rule
1316~~~~~~~~~~~~~~~~~~~~~~
1317
1318Set pool or vlan type mirror rule for a port::
1319
1320   testpmd> set port (port_id) mirror-rule (rule_id) \
1321            (pool-mirror-up|pool-mirror-down|vlan-mirror) \
1322            (poolmask|vlanid[,vlanid]*) dst-pool (pool_id) (on|off)
1323
1324Set link mirror rule for a port::
1325
1326   testpmd> set port (port_id) mirror-rule (rule_id) \
1327           (uplink-mirror|downlink-mirror) dst-pool (pool_id) (on|off)
1328
1329For example to enable mirror traffic with vlan 0,1 to pool 0::
1330
1331   set port 0 mirror-rule 0 vlan-mirror 0,1 dst-pool 0 on
1332
1333reset port - mirror rule
1334~~~~~~~~~~~~~~~~~~~~~~~~
1335
1336Reset a mirror rule for a port::
1337
1338   testpmd> reset port (port_id) mirror-rule (rule_id)
1339
1340set flush_rx
1341~~~~~~~~~~~~
1342
1343Set the flush on RX streams before forwarding.
1344The default is flush ``on``.
1345Mainly used with PCAP drivers to turn off the default behavior of flushing the first 512 packets on RX streams::
1346
1347   testpmd> set flush_rx off
1348
1349set bypass mode
1350~~~~~~~~~~~~~~~
1351
1352Set the bypass mode for the lowest port on bypass enabled NIC::
1353
1354   testpmd> set bypass mode (normal|bypass|isolate) (port_id)
1355
1356set bypass event
1357~~~~~~~~~~~~~~~~
1358
1359Set the event required to initiate specified bypass mode for the lowest port on a bypass enabled::
1360
1361   testpmd> set bypass event (timeout|os_on|os_off|power_on|power_off) \
1362            mode (normal|bypass|isolate) (port_id)
1363
1364Where:
1365
1366* ``timeout``: Enable bypass after watchdog timeout.
1367
1368* ``os_on``: Enable bypass when OS/board is powered on.
1369
1370* ``os_off``: Enable bypass when OS/board is powered off.
1371
1372* ``power_on``: Enable bypass when power supply is turned on.
1373
1374* ``power_off``: Enable bypass when power supply is turned off.
1375
1376
1377set bypass timeout
1378~~~~~~~~~~~~~~~~~~
1379
1380Set the bypass watchdog timeout to ``n`` seconds where 0 = instant::
1381
1382   testpmd> set bypass timeout (0|1.5|2|3|4|8|16|32)
1383
1384show bypass config
1385~~~~~~~~~~~~~~~~~~
1386
1387Show the bypass configuration for a bypass enabled NIC using the lowest port on the NIC::
1388
1389   testpmd> show bypass config (port_id)
1390
1391set link up
1392~~~~~~~~~~~
1393
1394Set link up for a port::
1395
1396   testpmd> set link-up port (port id)
1397
1398set link down
1399~~~~~~~~~~~~~
1400
1401Set link down for a port::
1402
1403   testpmd> set link-down port (port id)
1404
1405E-tag set
1406~~~~~~~~~
1407
1408Enable E-tag insertion for a VF on a port::
1409
1410   testpmd> E-tag set insertion on port-tag-id (value) port (port_id) vf (vf_id)
1411
1412Disable E-tag insertion for a VF on a port::
1413
1414   testpmd> E-tag set insertion off port (port_id) vf (vf_id)
1415
1416Enable/disable E-tag stripping on a port::
1417
1418   testpmd> E-tag set stripping (on|off) port (port_id)
1419
1420Enable/disable E-tag based forwarding on a port::
1421
1422   testpmd> E-tag set forwarding (on|off) port (port_id)
1423
1424Add an E-tag forwarding filter on a port::
1425
1426   testpmd> E-tag set filter add e-tag-id (value) dst-pool (pool_id) port (port_id)
1427
1428Delete an E-tag forwarding filter on a port::
1429   testpmd> E-tag set filter del e-tag-id (value) port (port_id)
1430
1431ddp add
1432~~~~~~~
1433
1434Load a dynamic device personalization (DDP) profile and store backup profile::
1435
1436   testpmd> ddp add (port_id) (profile_path[,backup_profile_path])
1437
1438ddp del
1439~~~~~~~
1440
1441Delete a dynamic device personalization profile and restore backup profile::
1442
1443   testpmd> ddp del (port_id) (backup_profile_path)
1444
1445ptype mapping
1446~~~~~~~~~~~~~
1447
1448List all items from the ptype mapping table::
1449
1450   testpmd> ptype mapping get (port_id) (valid_only)
1451
1452Where:
1453
1454* ``valid_only``: A flag indicates if only list valid items(=1) or all itemss(=0).
1455
1456Replace a specific or a group of software defined ptype with a new one::
1457
1458   testpmd> ptype mapping replace  (port_id) (target) (mask) (pkt_type)
1459
1460where:
1461
1462* ``target``: A specific software ptype or a mask to represent a group of software ptypes.
1463
1464* ``mask``: A flag indicate if "target" is a specific software ptype(=0) or a ptype mask(=1).
1465
1466* ``pkt_type``: The new software ptype to replace the old ones.
1467
1468Update hardware defined ptype to software defined packet type mapping table::
1469
1470   testpmd> ptype mapping update (port_id) (hw_ptype) (sw_ptype)
1471
1472where:
1473
1474* ``hw_ptype``: hardware ptype as the index of the ptype mapping table.
1475
1476* ``sw_ptype``: software ptype as the value of the ptype mapping table.
1477
1478Reset ptype mapping table::
1479
1480   testpmd> ptype mapping reset (port_id)
1481
1482config per port Rx offloading
1483~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
1484
1485Enable or disable a per port Rx offloading on all Rx queues of a port::
1486
1487   testpmd> port config (port_id) rx_offload (offloading) on|off
1488
1489* ``offloading``: can be any of these offloading capability:
1490                  vlan_strip, ipv4_cksum, udp_cksum, tcp_cksum, tcp_lro,
1491                  qinq_strip, outer_ipv4_cksum, macsec_strip,
1492                  header_split, vlan_filter, vlan_extend, jumbo_frame,
1493                  crc_strip, scatter, timestamp, security
1494
1495This command should be run when the port is stopped, or else it will fail.
1496
1497config per queue Rx offloading
1498~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
1499
1500Enable or disable a per queue Rx offloading only on a specific Rx queue::
1501
1502   testpmd> port (port_id) rxq (queue_id) rx_offload (offloading) on|off
1503
1504* ``offloading``: can be any of these offloading capability:
1505                  vlan_strip, ipv4_cksum, udp_cksum, tcp_cksum, tcp_lro,
1506                  qinq_strip, outer_ipv4_cksum, macsec_strip,
1507                  header_split, vlan_filter, vlan_extend, jumbo_frame,
1508                  crc_strip, scatter, timestamp, security
1509
1510This command should be run when the port is stopped, or else it will fail.
1511
1512config per port Tx offloading
1513~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
1514
1515Enable or disable a per port Tx offloading on all Tx queues of a port::
1516
1517   testpmd> port config (port_id) tx_offload (offloading) on|off
1518
1519* ``offloading``: can be any of these offloading capability:
1520                  vlan_insert, ipv4_cksum, udp_cksum, tcp_cksum,
1521                  sctp_cksum, tcp_tso, udp_tso, outer_ipv4_cksum,
1522                  qinq_insert, vxlan_tnl_tso, gre_tnl_tso,
1523                  ipip_tnl_tso, geneve_tnl_tso, macsec_insert,
1524                  mt_lockfree, multi_segs, fast_free, security
1525
1526This command should be run when the port is stopped, or else it will fail.
1527
1528config per queue Tx offloading
1529~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
1530
1531Enable or disable a per queue Tx offloading only on a specific Tx queue::
1532
1533   testpmd> port (port_id) txq (queue_id) tx_offload (offloading) on|off
1534
1535* ``offloading``: can be any of these offloading capability:
1536                  vlan_insert, ipv4_cksum, udp_cksum, tcp_cksum,
1537                  sctp_cksum, tcp_tso, udp_tso, outer_ipv4_cksum,
1538                  qinq_insert, vxlan_tnl_tso, gre_tnl_tso,
1539                  ipip_tnl_tso, geneve_tnl_tso, macsec_insert,
1540                  mt_lockfree, multi_segs, fast_free, security
1541
1542This command should be run when the port is stopped, or else it will fail.
1543
1544Config VXLAN Encap outer layers
1545~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
1546
1547Configure the outer layer to encapsulate a packet inside a VXLAN tunnel::
1548
1549 set vxlan ip-version (ipv4|ipv6) vni (vni) udp-src (udp-src) \
1550 udp-dst (udp-dst) ip-src (ip-src) ip-dst (ip-dst) eth-src (eth-src) \
1551 eth-dst (eth-dst)
1552
1553 set vxlan-with-vlan ip-version (ipv4|ipv6) vni (vni) udp-src (udp-src) \
1554 udp-dst (udp-dst) ip-src (ip-src) ip-dst (ip-dst) vlan-tci (vlan-tci) \
1555 eth-src (eth-src) eth-dst (eth-dst)
1556
1557Those command will set an internal configuration inside testpmd, any following
1558flow rule using the action vxlan_encap will use the last configuration set.
1559To have a different encapsulation header, one of those commands must be called
1560before the flow rule creation.
1561
1562Config NVGRE Encap outer layers
1563~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
1564
1565Configure the outer layer to encapsulate a packet inside a NVGRE tunnel::
1566
1567 set nvgre ip-version (ipv4|ipv6) tni (tni) ip-src (ip-src) ip-dst (ip-dst) \
1568        eth-src (eth-src) eth-dst (eth-dst)
1569 set nvgre-with-vlan ip-version (ipv4|ipv6) tni (tni) ip-src (ip-src) \
1570        ip-dst (ip-dst) vlan-tci (vlan-tci) eth-src (eth-src) eth-dst (eth-dst)
1571
1572Those command will set an internal configuration inside testpmd, any following
1573flow rule using the action nvgre_encap will use the last configuration set.
1574To have a different encapsulation header, one of those commands must be called
1575before the flow rule creation.
1576
1577Port Functions
1578--------------
1579
1580The following sections show functions for configuring ports.
1581
1582.. note::
1583
1584   Port configuration changes only become active when forwarding is started/restarted.
1585
1586port attach
1587~~~~~~~~~~~
1588
1589Attach a port specified by pci address or virtual device args::
1590
1591   testpmd> port attach (identifier)
1592
1593To attach a new pci device, the device should be recognized by kernel first.
1594Then it should be moved under DPDK management.
1595Finally the port can be attached to testpmd.
1596
1597For example, to move a pci device using ixgbe under DPDK management:
1598
1599.. code-block:: console
1600
1601   # Check the status of the available devices.
1602   ./usertools/dpdk-devbind.py --status
1603
1604   Network devices using DPDK-compatible driver
1605   ============================================
1606   <none>
1607
1608   Network devices using kernel driver
1609   ===================================
1610   0000:0a:00.0 '82599ES 10-Gigabit' if=eth2 drv=ixgbe unused=
1611
1612
1613   # Bind the device to igb_uio.
1614   sudo ./usertools/dpdk-devbind.py -b igb_uio 0000:0a:00.0
1615
1616
1617   # Recheck the status of the devices.
1618   ./usertools/dpdk-devbind.py --status
1619   Network devices using DPDK-compatible driver
1620   ============================================
1621   0000:0a:00.0 '82599ES 10-Gigabit' drv=igb_uio unused=
1622
1623To attach a port created by virtual device, above steps are not needed.
1624
1625For example, to attach a port whose pci address is 0000:0a:00.0.
1626
1627.. code-block:: console
1628
1629   testpmd> port attach 0000:0a:00.0
1630   Attaching a new port...
1631   EAL: PCI device 0000:0a:00.0 on NUMA socket -1
1632   EAL:   probe driver: 8086:10fb rte_ixgbe_pmd
1633   EAL:   PCI memory mapped at 0x7f83bfa00000
1634   EAL:   PCI memory mapped at 0x7f83bfa80000
1635   PMD: eth_ixgbe_dev_init(): MAC: 2, PHY: 18, SFP+: 5
1636   PMD: eth_ixgbe_dev_init(): port 0 vendorID=0x8086 deviceID=0x10fb
1637   Port 0 is attached. Now total ports is 1
1638   Done
1639
1640For example, to attach a port created by pcap PMD.
1641
1642.. code-block:: console
1643
1644   testpmd> port attach net_pcap0
1645   Attaching a new port...
1646   PMD: Initializing pmd_pcap for net_pcap0
1647   PMD: Creating pcap-backed ethdev on numa socket 0
1648   Port 0 is attached. Now total ports is 1
1649   Done
1650
1651In this case, identifier is ``net_pcap0``.
1652This identifier format is the same as ``--vdev`` format of DPDK applications.
1653
1654For example, to re-attach a bonded port which has been previously detached,
1655the mode and slave parameters must be given.
1656
1657.. code-block:: console
1658
1659   testpmd> port attach net_bond_0,mode=0,slave=1
1660   Attaching a new port...
1661   EAL: Initializing pmd_bond for net_bond_0
1662   EAL: Create bonded device net_bond_0 on port 0 in mode 0 on socket 0.
1663   Port 0 is attached. Now total ports is 1
1664   Done
1665
1666
1667port detach
1668~~~~~~~~~~~
1669
1670Detach a specific port::
1671
1672   testpmd> port detach (port_id)
1673
1674Before detaching a port, the port should be stopped and closed.
1675
1676For example, to detach a pci device port 0.
1677
1678.. code-block:: console
1679
1680   testpmd> port stop 0
1681   Stopping ports...
1682   Done
1683   testpmd> port close 0
1684   Closing ports...
1685   Done
1686
1687   testpmd> port detach 0
1688   Detaching a port...
1689   EAL: PCI device 0000:0a:00.0 on NUMA socket -1
1690   EAL:   remove driver: 8086:10fb rte_ixgbe_pmd
1691   EAL:   PCI memory unmapped at 0x7f83bfa00000
1692   EAL:   PCI memory unmapped at 0x7f83bfa80000
1693   Done
1694
1695
1696For example, to detach a virtual device port 0.
1697
1698.. code-block:: console
1699
1700   testpmd> port stop 0
1701   Stopping ports...
1702   Done
1703   testpmd> port close 0
1704   Closing ports...
1705   Done
1706
1707   testpmd> port detach 0
1708   Detaching a port...
1709   PMD: Closing pcap ethdev on numa socket 0
1710   Port 'net_pcap0' is detached. Now total ports is 0
1711   Done
1712
1713To remove a pci device completely from the system, first detach the port from testpmd.
1714Then the device should be moved under kernel management.
1715Finally the device can be removed using kernel pci hotplug functionality.
1716
1717For example, to move a pci device under kernel management:
1718
1719.. code-block:: console
1720
1721   sudo ./usertools/dpdk-devbind.py -b ixgbe 0000:0a:00.0
1722
1723   ./usertools/dpdk-devbind.py --status
1724
1725   Network devices using DPDK-compatible driver
1726   ============================================
1727   <none>
1728
1729   Network devices using kernel driver
1730   ===================================
1731   0000:0a:00.0 '82599ES 10-Gigabit' if=eth2 drv=ixgbe unused=igb_uio
1732
1733To remove a port created by a virtual device, above steps are not needed.
1734
1735port start
1736~~~~~~~~~~
1737
1738Start all ports or a specific port::
1739
1740   testpmd> port start (port_id|all)
1741
1742port stop
1743~~~~~~~~~
1744
1745Stop all ports or a specific port::
1746
1747   testpmd> port stop (port_id|all)
1748
1749port close
1750~~~~~~~~~~
1751
1752Close all ports or a specific port::
1753
1754   testpmd> port close (port_id|all)
1755
1756port config - queue ring size
1757~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
1758
1759Configure a rx/tx queue ring size::
1760
1761   testpmd> port (port_id) (rxq|txq) (queue_id) ring_size (value)
1762
1763Only take effect after command that (re-)start the port or command that setup specific queue.
1764
1765port start/stop queue
1766~~~~~~~~~~~~~~~~~~~~~
1767
1768Start/stop a rx/tx queue on a specific port::
1769
1770   testpmd> port (port_id) (rxq|txq) (queue_id) (start|stop)
1771
1772port setup queue
1773~~~~~~~~~~~~~~~~~~~~~
1774
1775Setup a rx/tx queue on a specific port::
1776
1777   testpmd> port (port_id) (rxq|txq) (queue_id) setup
1778
1779Only take effect when port is started.
1780
1781port config - speed
1782~~~~~~~~~~~~~~~~~~~
1783
1784Set the speed and duplex mode for all ports or a specific port::
1785
1786   testpmd> port config (port_id|all) speed (10|100|1000|10000|25000|40000|50000|100000|auto) \
1787            duplex (half|full|auto)
1788
1789port config - queues/descriptors
1790~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
1791
1792Set number of queues/descriptors for rxq, txq, rxd and txd::
1793
1794   testpmd> port config all (rxq|txq|rxd|txd) (value)
1795
1796This is equivalent to the ``--rxq``, ``--txq``, ``--rxd`` and ``--txd`` command-line options.
1797
1798port config - max-pkt-len
1799~~~~~~~~~~~~~~~~~~~~~~~~~
1800
1801Set the maximum packet length::
1802
1803   testpmd> port config all max-pkt-len (value)
1804
1805This is equivalent to the ``--max-pkt-len`` command-line option.
1806
1807port config - CRC Strip
1808~~~~~~~~~~~~~~~~~~~~~~~
1809
1810Set hardware CRC stripping on or off for all ports::
1811
1812   testpmd> port config all crc-strip (on|off)
1813
1814CRC stripping is on by default.
1815
1816The ``off`` option is equivalent to the ``--disable-crc-strip`` command-line option.
1817
1818port config - scatter
1819~~~~~~~~~~~~~~~~~~~~~~~
1820
1821Set RX scatter mode on or off for all ports::
1822
1823   testpmd> port config all scatter (on|off)
1824
1825RX scatter mode is off by default.
1826
1827The ``on`` option is equivalent to the ``--enable-scatter`` command-line option.
1828
1829port config - RX Checksum
1830~~~~~~~~~~~~~~~~~~~~~~~~~
1831
1832Set hardware RX checksum offload to on or off for all ports::
1833
1834   testpmd> port config all rx-cksum (on|off)
1835
1836Checksum offload is off by default.
1837
1838The ``on`` option is equivalent to the ``--enable-rx-cksum`` command-line option.
1839
1840port config - VLAN
1841~~~~~~~~~~~~~~~~~~
1842
1843Set hardware VLAN on or off for all ports::
1844
1845   testpmd> port config all hw-vlan (on|off)
1846
1847Hardware VLAN is off by default.
1848
1849The ``on`` option is equivalent to the ``--enable-hw-vlan`` command-line option.
1850
1851port config - VLAN filter
1852~~~~~~~~~~~~~~~~~~~~~~~~~
1853
1854Set hardware VLAN filter on or off for all ports::
1855
1856   testpmd> port config all hw-vlan-filter (on|off)
1857
1858Hardware VLAN filter is off by default.
1859
1860The ``on`` option is equivalent to the ``--enable-hw-vlan-filter`` command-line option.
1861
1862port config - VLAN strip
1863~~~~~~~~~~~~~~~~~~~~~~~~
1864
1865Set hardware VLAN strip on or off for all ports::
1866
1867   testpmd> port config all hw-vlan-strip (on|off)
1868
1869Hardware VLAN strip is off by default.
1870
1871The ``on`` option is equivalent to the ``--enable-hw-vlan-strip`` command-line option.
1872
1873port config - VLAN extend
1874~~~~~~~~~~~~~~~~~~~~~~~~~
1875
1876Set hardware VLAN extend on or off for all ports::
1877
1878   testpmd> port config all hw-vlan-extend (on|off)
1879
1880Hardware VLAN extend is off by default.
1881
1882The ``on`` option is equivalent to the ``--enable-hw-vlan-extend`` command-line option.
1883
1884port config - Drop Packets
1885~~~~~~~~~~~~~~~~~~~~~~~~~~
1886
1887Set packet drop for packets with no descriptors on or off for all ports::
1888
1889   testpmd> port config all drop-en (on|off)
1890
1891Packet dropping for packets with no descriptors is off by default.
1892
1893The ``on`` option is equivalent to the ``--enable-drop-en`` command-line option.
1894
1895port config - RSS
1896~~~~~~~~~~~~~~~~~
1897
1898Set the RSS (Receive Side Scaling) mode on or off::
1899
1900   testpmd> port config all rss (all|default|ip|tcp|udp|sctp|ether|port|vxlan|geneve|nvgre|none)
1901
1902RSS is on by default.
1903
1904The ``all`` option is equivalent to ip|tcp|udp|sctp|ether.
1905The ``default`` option enables all supported RSS types reported by device info.
1906The ``none`` option is equivalent to the ``--disable-rss`` command-line option.
1907
1908port config - RSS Reta
1909~~~~~~~~~~~~~~~~~~~~~~
1910
1911Set the RSS (Receive Side Scaling) redirection table::
1912
1913   testpmd> port config all rss reta (hash,queue)[,(hash,queue)]
1914
1915port config - DCB
1916~~~~~~~~~~~~~~~~~
1917
1918Set the DCB mode for an individual port::
1919
1920   testpmd> port config (port_id) dcb vt (on|off) (traffic_class) pfc (on|off)
1921
1922The traffic class should be 4 or 8.
1923
1924port config - Burst
1925~~~~~~~~~~~~~~~~~~~
1926
1927Set the number of packets per burst::
1928
1929   testpmd> port config all burst (value)
1930
1931This is equivalent to the ``--burst`` command-line option.
1932
1933port config - Threshold
1934~~~~~~~~~~~~~~~~~~~~~~~
1935
1936Set thresholds for TX/RX queues::
1937
1938   testpmd> port config all (threshold) (value)
1939
1940Where the threshold type can be:
1941
1942* ``txpt:`` Set the prefetch threshold register of the TX rings, 0 <= value <= 255.
1943
1944* ``txht:`` Set the host threshold register of the TX rings, 0 <= value <= 255.
1945
1946* ``txwt:`` Set the write-back threshold register of the TX rings, 0 <= value <= 255.
1947
1948* ``rxpt:`` Set the prefetch threshold register of the RX rings, 0 <= value <= 255.
1949
1950* ``rxht:`` Set the host threshold register of the RX rings, 0 <= value <= 255.
1951
1952* ``rxwt:`` Set the write-back threshold register of the RX rings, 0 <= value <= 255.
1953
1954* ``txfreet:`` Set the transmit free threshold of the TX rings, 0 <= value <= txd.
1955
1956* ``rxfreet:`` Set the transmit free threshold of the RX rings, 0 <= value <= rxd.
1957
1958* ``txrst:`` Set the transmit RS bit threshold of TX rings, 0 <= value <= txd.
1959
1960These threshold options are also available from the command-line.
1961
1962port config - E-tag
1963~~~~~~~~~~~~~~~~~~~
1964
1965Set the value of ether-type for E-tag::
1966
1967   testpmd> port config (port_id|all) l2-tunnel E-tag ether-type (value)
1968
1969Enable/disable the E-tag support::
1970
1971   testpmd> port config (port_id|all) l2-tunnel E-tag (enable|disable)
1972
1973port config pctype mapping
1974~~~~~~~~~~~~~~~~~~~~~~~~~~
1975
1976Reset pctype mapping table::
1977
1978   testpmd> port config (port_id) pctype mapping reset
1979
1980Update hardware defined pctype to software defined flow type mapping table::
1981
1982   testpmd> port config (port_id) pctype mapping update (pctype_id_0[,pctype_id_1]*) (flow_type_id)
1983
1984where:
1985
1986* ``pctype_id_x``: hardware pctype id as index of bit in bitmask value of the pctype mapping table.
1987
1988* ``flow_type_id``: software flow type id as the index of the pctype mapping table.
1989
1990port config input set
1991~~~~~~~~~~~~~~~~~~~~~
1992
1993Config RSS/FDIR/FDIR flexible payload input set for some pctype::
1994   testpmd> port config (port_id) pctype (pctype_id) \
1995            (hash_inset|fdir_inset|fdir_flx_inset) \
1996	    (get|set|clear) field (field_idx)
1997
1998Clear RSS/FDIR/FDIR flexible payload input set for some pctype::
1999   testpmd> port config (port_id) pctype (pctype_id) \
2000            (hash_inset|fdir_inset|fdir_flx_inset) clear all
2001
2002where:
2003
2004* ``pctype_id``: hardware packet classification types.
2005* ``field_idx``: hardware field index.
2006
2007port config udp_tunnel_port
2008~~~~~~~~~~~~~~~~~~~~~~~~~~~
2009
2010Add/remove UDP tunnel port for VXLAN/GENEVE tunneling protocols::
2011    testpmd> port config (port_id) udp_tunnel_port add|rm vxlan|geneve (udp_port)
2012
2013Link Bonding Functions
2014----------------------
2015
2016The Link Bonding functions make it possible to dynamically create and
2017manage link bonding devices from within testpmd interactive prompt.
2018
2019create bonded device
2020~~~~~~~~~~~~~~~~~~~~
2021
2022Create a new bonding device::
2023
2024   testpmd> create bonded device (mode) (socket)
2025
2026For example, to create a bonded device in mode 1 on socket 0::
2027
2028   testpmd> create bonded 1 0
2029   created new bonded device (port X)
2030
2031add bonding slave
2032~~~~~~~~~~~~~~~~~
2033
2034Adds Ethernet device to a Link Bonding device::
2035
2036   testpmd> add bonding slave (slave id) (port id)
2037
2038For example, to add Ethernet device (port 6) to a Link Bonding device (port 10)::
2039
2040   testpmd> add bonding slave 6 10
2041
2042
2043remove bonding slave
2044~~~~~~~~~~~~~~~~~~~~
2045
2046Removes an Ethernet slave device from a Link Bonding device::
2047
2048   testpmd> remove bonding slave (slave id) (port id)
2049
2050For example, to remove Ethernet slave device (port 6) to a Link Bonding device (port 10)::
2051
2052   testpmd> remove bonding slave 6 10
2053
2054set bonding mode
2055~~~~~~~~~~~~~~~~
2056
2057Set the Link Bonding mode of a Link Bonding device::
2058
2059   testpmd> set bonding mode (value) (port id)
2060
2061For example, to set the bonding mode of a Link Bonding device (port 10) to broadcast (mode 3)::
2062
2063   testpmd> set bonding mode 3 10
2064
2065set bonding primary
2066~~~~~~~~~~~~~~~~~~~
2067
2068Set an Ethernet slave device as the primary device on a Link Bonding device::
2069
2070   testpmd> set bonding primary (slave id) (port id)
2071
2072For example, to set the Ethernet slave device (port 6) as the primary port of a Link Bonding device (port 10)::
2073
2074   testpmd> set bonding primary 6 10
2075
2076set bonding mac
2077~~~~~~~~~~~~~~~
2078
2079Set the MAC address of a Link Bonding device::
2080
2081   testpmd> set bonding mac (port id) (mac)
2082
2083For example, to set the MAC address of a Link Bonding device (port 10) to 00:00:00:00:00:01::
2084
2085   testpmd> set bonding mac 10 00:00:00:00:00:01
2086
2087set bonding xmit_balance_policy
2088~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
2089
2090Set the transmission policy for a Link Bonding device when it is in Balance XOR mode::
2091
2092   testpmd> set bonding xmit_balance_policy (port_id) (l2|l23|l34)
2093
2094For example, set a Link Bonding device (port 10) to use a balance policy of layer 3+4 (IP addresses & UDP ports)::
2095
2096   testpmd> set bonding xmit_balance_policy 10 l34
2097
2098
2099set bonding mon_period
2100~~~~~~~~~~~~~~~~~~~~~~
2101
2102Set the link status monitoring polling period in milliseconds for a bonding device.
2103
2104This adds support for PMD slave devices which do not support link status interrupts.
2105When the mon_period is set to a value greater than 0 then all PMD's which do not support
2106link status ISR will be queried every polling interval to check if their link status has changed::
2107
2108   testpmd> set bonding mon_period (port_id) (value)
2109
2110For example, to set the link status monitoring polling period of bonded device (port 5) to 150ms::
2111
2112   testpmd> set bonding mon_period 5 150
2113
2114
2115set bonding lacp dedicated_queue
2116~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
2117
2118Enable dedicated tx/rx queues on bonding devices slaves to handle LACP control plane traffic
2119when in mode 4 (link-aggregration-802.3ad)::
2120
2121   testpmd> set bonding lacp dedicated_queues (port_id) (enable|disable)
2122
2123
2124set bonding agg_mode
2125~~~~~~~~~~~~~~~~~~~~
2126
2127Enable one of the specific aggregators mode when in mode 4 (link-aggregration-802.3ad)::
2128
2129   testpmd> set bonding agg_mode (port_id) (bandwidth|count|stable)
2130
2131
2132show bonding config
2133~~~~~~~~~~~~~~~~~~~
2134
2135Show the current configuration of a Link Bonding device::
2136
2137   testpmd> show bonding config (port id)
2138
2139For example,
2140to show the configuration a Link Bonding device (port 9) with 3 slave devices (1, 3, 4)
2141in balance mode with a transmission policy of layer 2+3::
2142
2143   testpmd> show bonding config 9
2144        Bonding mode: 2
2145        Balance Xmit Policy: BALANCE_XMIT_POLICY_LAYER23
2146        Slaves (3): [1 3 4]
2147        Active Slaves (3): [1 3 4]
2148        Primary: [3]
2149
2150
2151Register Functions
2152------------------
2153
2154The Register Functions can be used to read from and write to registers on the network card referenced by a port number.
2155This is mainly useful for debugging purposes.
2156Reference should be made to the appropriate datasheet for the network card for details on the register addresses
2157and fields that can be accessed.
2158
2159read reg
2160~~~~~~~~
2161
2162Display the value of a port register::
2163
2164   testpmd> read reg (port_id) (address)
2165
2166For example, to examine the Flow Director control register (FDIRCTL, 0x0000EE000) on an Intel 82599 10 GbE Controller::
2167
2168   testpmd> read reg 0 0xEE00
2169   port 0 PCI register at offset 0xEE00: 0x4A060029 (1241907241)
2170
2171read regfield
2172~~~~~~~~~~~~~
2173
2174Display a port register bit field::
2175
2176   testpmd> read regfield (port_id) (address) (bit_x) (bit_y)
2177
2178For example, reading the lowest two bits from the register in the example above::
2179
2180   testpmd> read regfield 0 0xEE00 0 1
2181   port 0 PCI register at offset 0xEE00: bits[0, 1]=0x1 (1)
2182
2183read regbit
2184~~~~~~~~~~~
2185
2186Display a single port register bit::
2187
2188   testpmd> read regbit (port_id) (address) (bit_x)
2189
2190For example, reading the lowest bit from the register in the example above::
2191
2192   testpmd> read regbit 0 0xEE00 0
2193   port 0 PCI register at offset 0xEE00: bit 0=1
2194
2195write reg
2196~~~~~~~~~
2197
2198Set the value of a port register::
2199
2200   testpmd> write reg (port_id) (address) (value)
2201
2202For example, to clear a register::
2203
2204   testpmd> write reg 0 0xEE00 0x0
2205   port 0 PCI register at offset 0xEE00: 0x00000000 (0)
2206
2207write regfield
2208~~~~~~~~~~~~~~
2209
2210Set bit field of a port register::
2211
2212   testpmd> write regfield (port_id) (address) (bit_x) (bit_y) (value)
2213
2214For example, writing to the register cleared in the example above::
2215
2216   testpmd> write regfield 0 0xEE00 0 1 2
2217   port 0 PCI register at offset 0xEE00: 0x00000002 (2)
2218
2219write regbit
2220~~~~~~~~~~~~
2221
2222Set single bit value of a port register::
2223
2224   testpmd> write regbit (port_id) (address) (bit_x) (value)
2225
2226For example, to set the high bit in the register from the example above::
2227
2228   testpmd> write regbit 0 0xEE00 31 1
2229   port 0 PCI register at offset 0xEE00: 0x8000000A (2147483658)
2230
2231Traffic Metering and Policing
2232-----------------------------
2233
2234The following section shows functions for configuring traffic metering and
2235policing on the ethernet device through the use of generic ethdev API.
2236
2237show port traffic management capability
2238~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
2239
2240Show traffic metering and policing capability of the port::
2241
2242   testpmd> show port meter cap (port_id)
2243
2244add port meter profile (srTCM rfc2967)
2245~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
2246
2247Add meter profile (srTCM rfc2697) to the ethernet device::
2248
2249   testpmd> add port meter profile srtcm_rfc2697 (port_id) (profile_id) \
2250   (cir) (cbs) (ebs)
2251
2252where:
2253
2254* ``profile_id``: ID for the meter profile.
2255* ``cir``: Committed Information Rate (CIR) (bytes/second).
2256* ``cbs``: Committed Burst Size (CBS) (bytes).
2257* ``ebs``: Excess Burst Size (EBS) (bytes).
2258
2259add port meter profile (trTCM rfc2968)
2260~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
2261
2262Add meter profile (srTCM rfc2698) to the ethernet device::
2263
2264   testpmd> add port meter profile trtcm_rfc2698 (port_id) (profile_id) \
2265   (cir) (pir) (cbs) (pbs)
2266
2267where:
2268
2269* ``profile_id``: ID for the meter profile.
2270* ``cir``: Committed information rate (bytes/second).
2271* ``pir``: Peak information rate (bytes/second).
2272* ``cbs``: Committed burst size (bytes).
2273* ``pbs``: Peak burst size (bytes).
2274
2275add port meter profile (trTCM rfc4115)
2276~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
2277
2278Add meter profile (trTCM rfc4115) to the ethernet device::
2279
2280   testpmd> add port meter profile trtcm_rfc4115 (port_id) (profile_id) \
2281   (cir) (eir) (cbs) (ebs)
2282
2283where:
2284
2285* ``profile_id``: ID for the meter profile.
2286* ``cir``: Committed information rate (bytes/second).
2287* ``eir``: Excess information rate (bytes/second).
2288* ``cbs``: Committed burst size (bytes).
2289* ``ebs``: Excess burst size (bytes).
2290
2291delete port meter profile
2292~~~~~~~~~~~~~~~~~~~~~~~~~
2293
2294Delete meter profile from the ethernet device::
2295
2296   testpmd> del port meter profile (port_id) (profile_id)
2297
2298create port meter
2299~~~~~~~~~~~~~~~~~
2300
2301Create new meter object for the ethernet device::
2302
2303   testpmd> create port meter (port_id) (mtr_id) (profile_id) \
2304   (meter_enable) (g_action) (y_action) (r_action) (stats_mask) (shared) \
2305   (use_pre_meter_color) [(dscp_tbl_entry0) (dscp_tbl_entry1)...\
2306   (dscp_tbl_entry63)]
2307
2308where:
2309
2310* ``mtr_id``: meter object ID.
2311* ``profile_id``: ID for the meter profile.
2312* ``meter_enable``: When this parameter has a non-zero value, the meter object
2313  gets enabled at the time of creation, otherwise remains disabled.
2314* ``g_action``: Policer action for the packet with green color.
2315* ``y_action``: Policer action for the packet with yellow color.
2316* ``r_action``: Policer action for the packet with red color.
2317* ``stats_mask``: Mask of statistics counter types to be enabled for the
2318  meter object.
2319* ``shared``:  When this parameter has a non-zero value, the meter object is
2320  shared by multiple flows. Otherwise, meter object is used by single flow.
2321* ``use_pre_meter_color``: When this parameter has a non-zero value, the
2322  input color for the current meter object is determined by the latest meter
2323  object in the same flow. Otherwise, the current meter object uses the
2324  *dscp_table* to determine the input color.
2325* ``dscp_tbl_entryx``: DSCP table entry x providing meter providing input
2326  color, 0 <= x <= 63.
2327
2328enable port meter
2329~~~~~~~~~~~~~~~~~
2330
2331Enable meter for the ethernet device::
2332
2333   testpmd> enable port meter (port_id) (mtr_id)
2334
2335disable port meter
2336~~~~~~~~~~~~~~~~~~
2337
2338Disable meter for the ethernet device::
2339
2340   testpmd> disable port meter (port_id) (mtr_id)
2341
2342delete port meter
2343~~~~~~~~~~~~~~~~~
2344
2345Delete meter for the ethernet device::
2346
2347   testpmd> del port meter (port_id) (mtr_id)
2348
2349Set port meter profile
2350~~~~~~~~~~~~~~~~~~~~~~
2351
2352Set meter profile for the ethernet device::
2353
2354   testpmd> set port meter profile (port_id) (mtr_id) (profile_id)
2355
2356set port meter dscp table
2357~~~~~~~~~~~~~~~~~~~~~~~~~
2358
2359Set meter dscp table for the ethernet device::
2360
2361   testpmd> set port meter dscp table (port_id) (mtr_id) [(dscp_tbl_entry0) \
2362   (dscp_tbl_entry1)...(dscp_tbl_entry63)]
2363
2364set port meter policer action
2365~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
2366
2367Set meter policer action for the ethernet device::
2368
2369   testpmd> set port meter policer action (port_id) (mtr_id) (action_mask) \
2370   (action0) [(action1) (action1)]
2371
2372where:
2373
2374* ``action_mask``: Bit mask indicating which policer actions need to be
2375  updated. One or more policer actions can be updated in a single function
2376  invocation. To update the policer action associated with color C, bit
2377  (1 << C) needs to be set in *action_mask* and element at position C
2378  in the *actions* array needs to be valid.
2379* ``actionx``: Policer action for the color x,
2380  RTE_MTR_GREEN <= x < RTE_MTR_COLORS
2381
2382set port meter stats mask
2383~~~~~~~~~~~~~~~~~~~~~~~~~
2384
2385Set meter stats mask for the ethernet device::
2386
2387   testpmd> set port meter stats mask (port_id) (mtr_id) (stats_mask)
2388
2389where:
2390
2391* ``stats_mask``: Bit mask indicating statistics counter types to be enabled.
2392
2393show port meter stats
2394~~~~~~~~~~~~~~~~~~~~~
2395
2396Show meter stats of the ethernet device::
2397
2398   testpmd> show port meter stats (port_id) (mtr_id) (clear)
2399
2400where:
2401
2402* ``clear``: Flag that indicates whether the statistics counters should
2403  be cleared (i.e. set to zero) immediately after they have been read or not.
2404
2405Traffic Management
2406------------------
2407
2408The following section shows functions for configuring traffic management on
2409on the ethernet device through the use of generic TM API.
2410
2411show port traffic management capability
2412~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
2413
2414Show traffic management capability of the port::
2415
2416   testpmd> show port tm cap (port_id)
2417
2418show port traffic management capability (hierarchy level)
2419~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
2420
2421Show traffic management hierarchy level capability of the port::
2422
2423   testpmd> show port tm level cap (port_id) (level_id)
2424
2425show port traffic management capability (hierarchy node level)
2426~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
2427
2428Show the traffic management hierarchy node capability of the port::
2429
2430   testpmd> show port tm node cap (port_id) (node_id)
2431
2432show port traffic management hierarchy node type
2433~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
2434
2435Show the port traffic management hierarchy node type::
2436
2437   testpmd> show port tm node type (port_id) (node_id)
2438
2439show port traffic management hierarchy node stats
2440~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
2441
2442Show the port traffic management hierarchy node statistics::
2443
2444   testpmd> show port tm node stats (port_id) (node_id) (clear)
2445
2446where:
2447
2448* ``clear``: When this parameter has a non-zero value, the statistics counters
2449  are cleared (i.e. set to zero) immediately after they have been read,
2450  otherwise the statistics counters are left untouched.
2451
2452Add port traffic management private shaper profile
2453~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
2454
2455Add the port traffic management private shaper profile::
2456
2457   testpmd> add port tm node shaper profile (port_id) (shaper_profile_id) \
2458   (tb_rate) (tb_size) (packet_length_adjust)
2459
2460where:
2461
2462* ``shaper_profile id``: Shaper profile ID for the new profile.
2463* ``tb_rate``: Token bucket rate (bytes per second).
2464* ``tb_size``: Token bucket size (bytes).
2465* ``packet_length_adjust``: The value (bytes) to be added to the length of
2466  each packet for the purpose of shaping. This parameter value can be used to
2467  correct the packet length with the framing overhead bytes that are consumed
2468  on the wire.
2469
2470Delete port traffic management private shaper profile
2471~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
2472
2473Delete the port traffic management private shaper::
2474
2475   testpmd> del port tm node shaper profile (port_id) (shaper_profile_id)
2476
2477where:
2478
2479* ``shaper_profile id``: Shaper profile ID that needs to be deleted.
2480
2481Add port traffic management shared shaper
2482~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
2483
2484Create the port traffic management shared shaper::
2485
2486   testpmd> add port tm node shared shaper (port_id) (shared_shaper_id) \
2487   (shaper_profile_id)
2488
2489where:
2490
2491* ``shared_shaper_id``: Shared shaper ID to be created.
2492* ``shaper_profile id``: Shaper profile ID for shared shaper.
2493
2494Set port traffic management shared shaper
2495~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
2496
2497Update the port traffic management shared shaper::
2498
2499   testpmd> set port tm node shared shaper (port_id) (shared_shaper_id) \
2500   (shaper_profile_id)
2501
2502where:
2503
2504* ``shared_shaper_id``: Shared shaper ID to be update.
2505* ``shaper_profile id``: Shaper profile ID for shared shaper.
2506
2507Delete port traffic management shared shaper
2508~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
2509
2510Delete the port traffic management shared shaper::
2511
2512   testpmd> del port tm node shared shaper (port_id) (shared_shaper_id)
2513
2514where:
2515
2516* ``shared_shaper_id``: Shared shaper ID to be deleted.
2517
2518Set port traffic management hiearchy node private shaper
2519~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
2520
2521set the port traffic management hierarchy node private shaper::
2522
2523   testpmd> set port tm node shaper profile (port_id) (node_id) \
2524   (shaper_profile_id)
2525
2526where:
2527
2528* ``shaper_profile id``: Private shaper profile ID to be enabled on the
2529  hierarchy node.
2530
2531Add port traffic management WRED profile
2532~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
2533
2534Create a new WRED profile::
2535
2536   testpmd> add port tm node wred profile (port_id) (wred_profile_id) \
2537   (color_g) (min_th_g) (max_th_g) (maxp_inv_g) (wq_log2_g) \
2538   (color_y) (min_th_y) (max_th_y) (maxp_inv_y) (wq_log2_y) \
2539   (color_r) (min_th_r) (max_th_r) (maxp_inv_r) (wq_log2_r)
2540
2541where:
2542
2543* ``wred_profile id``: Identifier for the newly create WRED profile
2544* ``color_g``: Packet color (green)
2545* ``min_th_g``: Minimum queue threshold for packet with green color
2546* ``max_th_g``: Minimum queue threshold for packet with green color
2547* ``maxp_inv_g``: Inverse of packet marking probability maximum value (maxp)
2548* ``wq_log2_g``: Negated log2 of queue weight (wq)
2549* ``color_y``: Packet color (yellow)
2550* ``min_th_y``: Minimum queue threshold for packet with yellow color
2551* ``max_th_y``: Minimum queue threshold for packet with yellow color
2552* ``maxp_inv_y``: Inverse of packet marking probability maximum value (maxp)
2553* ``wq_log2_y``: Negated log2 of queue weight (wq)
2554* ``color_r``: Packet color (red)
2555* ``min_th_r``: Minimum queue threshold for packet with yellow color
2556* ``max_th_r``: Minimum queue threshold for packet with yellow color
2557* ``maxp_inv_r``: Inverse of packet marking probability maximum value (maxp)
2558* ``wq_log2_r``: Negated log2 of queue weight (wq)
2559
2560Delete port traffic management WRED profile
2561~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
2562
2563Delete the WRED profile::
2564
2565   testpmd> del port tm node wred profile (port_id) (wred_profile_id)
2566
2567Add port traffic management hierarchy nonleaf node
2568~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
2569
2570Add nonleaf node to port traffic management hiearchy::
2571
2572   testpmd> add port tm nonleaf node (port_id) (node_id) (parent_node_id) \
2573   (priority) (weight) (level_id) (shaper_profile_id) \
2574   (n_sp_priorities) (stats_mask) (n_shared_shapers) \
2575   [(shared_shaper_0) (shared_shaper_1) ...] \
2576
2577where:
2578
2579* ``parent_node_id``: Node ID of the parent.
2580* ``priority``: Node priority (highest node priority is zero). This is used by
2581  the SP algorithm running on the parent node for scheduling this node.
2582* ``weight``: Node weight (lowest weight is one). The node weight is relative
2583  to the weight sum of all siblings that have the same priority. It is used by
2584  the WFQ algorithm running on the parent node for scheduling this node.
2585* ``level_id``: Hiearchy level of the node.
2586* ``shaper_profile_id``: Shaper profile ID of the private shaper to be used by
2587  the node.
2588* ``n_sp_priorities``: Number of strict priorities.
2589* ``stats_mask``: Mask of statistics counter types to be enabled for this node.
2590* ``n_shared_shapers``: Number of shared shapers.
2591* ``shared_shaper_id``: Shared shaper id.
2592
2593Add port traffic management hierarchy leaf node
2594~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
2595
2596Add leaf node to port traffic management hiearchy::
2597
2598   testpmd> add port tm leaf node (port_id) (node_id) (parent_node_id) \
2599   (priority) (weight) (level_id) (shaper_profile_id) \
2600   (cman_mode) (wred_profile_id) (stats_mask) (n_shared_shapers) \
2601   [(shared_shaper_id) (shared_shaper_id) ...] \
2602
2603where:
2604
2605* ``parent_node_id``: Node ID of the parent.
2606* ``priority``: Node priority (highest node priority is zero). This is used by
2607  the SP algorithm running on the parent node for scheduling this node.
2608* ``weight``: Node weight (lowest weight is one). The node weight is relative
2609  to the weight sum of all siblings that have the same priority. It is used by
2610  the WFQ algorithm running on the parent node for scheduling this node.
2611* ``level_id``: Hiearchy level of the node.
2612* ``shaper_profile_id``: Shaper profile ID of the private shaper to be used by
2613  the node.
2614* ``cman_mode``: Congestion management mode to be enabled for this node.
2615* ``wred_profile_id``: WRED profile id to be enabled for this node.
2616* ``stats_mask``: Mask of statistics counter types to be enabled for this node.
2617* ``n_shared_shapers``: Number of shared shapers.
2618* ``shared_shaper_id``: Shared shaper id.
2619
2620Delete port traffic management hierarchy node
2621~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
2622
2623Delete node from port traffic management hiearchy::
2624
2625   testpmd> del port tm node (port_id) (node_id)
2626
2627Update port traffic management hierarchy parent node
2628~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
2629
2630Update port traffic management hierarchy parent node::
2631
2632   testpmd> set port tm node parent (port_id) (node_id) (parent_node_id) \
2633   (priority) (weight)
2634
2635This function can only be called after the hierarchy commit invocation. Its
2636success depends on the port support for this operation, as advertised through
2637the port capability set. This function is valid for all nodes of the traffic
2638management hierarchy except root node.
2639
2640Suspend port traffic management hierarchy node
2641~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
2642
2643   testpmd> suspend port tm node (port_id) (node_id)
2644
2645Resume port traffic management hierarchy node
2646~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
2647
2648   testpmd> resume port tm node (port_id) (node_id)
2649
2650Commit port traffic management hierarchy
2651~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
2652
2653Commit the traffic management hierarchy on the port::
2654
2655   testpmd> port tm hierarchy commit (port_id) (clean_on_fail)
2656
2657where:
2658
2659* ``clean_on_fail``: When set to non-zero, hierarchy is cleared on function
2660  call failure. On the other hand, hierarchy is preserved when this parameter
2661  is equal to zero.
2662
2663Set port traffic management default hierarchy (tm forwarding mode)
2664~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
2665
2666set the traffic management default hierarchy on the port::
2667
2668   testpmd> set port tm hierarchy default (port_id)
2669
2670Filter Functions
2671----------------
2672
2673This section details the available filter functions that are available.
2674
2675Note these functions interface the deprecated legacy filtering framework,
2676superseded by *rte_flow*. See `Flow rules management`_.
2677
2678ethertype_filter
2679~~~~~~~~~~~~~~~~~~~~
2680
2681Add or delete a L2 Ethertype filter, which identify packets by their L2 Ethertype mainly assign them to a receive queue::
2682
2683   ethertype_filter (port_id) (add|del) (mac_addr|mac_ignr) (mac_address) \
2684                    ethertype (ether_type) (drop|fwd) queue (queue_id)
2685
2686The available information parameters are:
2687
2688* ``port_id``: The port which the Ethertype filter assigned on.
2689
2690* ``mac_addr``: Compare destination mac address.
2691
2692* ``mac_ignr``: Ignore destination mac address match.
2693
2694* ``mac_address``: Destination mac address to match.
2695
2696* ``ether_type``: The EtherType value want to match,
2697  for example 0x0806 for ARP packet. 0x0800 (IPv4) and 0x86DD (IPv6) are invalid.
2698
2699* ``queue_id``: The receive queue associated with this EtherType filter.
2700  It is meaningless when deleting or dropping.
2701
2702Example, to add/remove an ethertype filter rule::
2703
2704   testpmd> ethertype_filter 0 add mac_ignr 00:11:22:33:44:55 \
2705                             ethertype 0x0806 fwd queue 3
2706
2707   testpmd> ethertype_filter 0 del mac_ignr 00:11:22:33:44:55 \
2708                             ethertype 0x0806 fwd queue 3
2709
27102tuple_filter
2711~~~~~~~~~~~~~~~~~
2712
2713Add or delete a 2-tuple filter,
2714which identifies packets by specific protocol and destination TCP/UDP port
2715and forwards packets into one of the receive queues::
2716
2717   2tuple_filter (port_id) (add|del) dst_port (dst_port_value) \
2718                 protocol (protocol_value) mask (mask_value) \
2719                 tcp_flags (tcp_flags_value) priority (prio_value) \
2720                 queue (queue_id)
2721
2722The available information parameters are:
2723
2724* ``port_id``: The port which the 2-tuple filter assigned on.
2725
2726* ``dst_port_value``: Destination port in L4.
2727
2728* ``protocol_value``: IP L4 protocol.
2729
2730* ``mask_value``: Participates in the match or not by bit for field above, 1b means participate.
2731
2732* ``tcp_flags_value``: TCP control bits. The non-zero value is invalid, when the pro_value is not set to 0x06 (TCP).
2733
2734* ``prio_value``: Priority of this filter.
2735
2736* ``queue_id``: The receive queue associated with this 2-tuple filter.
2737
2738Example, to add/remove an 2tuple filter rule::
2739
2740   testpmd> 2tuple_filter 0 add dst_port 32 protocol 0x06 mask 0x03 \
2741                          tcp_flags 0x02 priority 3 queue 3
2742
2743   testpmd> 2tuple_filter 0 del dst_port 32 protocol 0x06 mask 0x03 \
2744                          tcp_flags 0x02 priority 3 queue 3
2745
27465tuple_filter
2747~~~~~~~~~~~~~~~~~
2748
2749Add or delete a 5-tuple filter,
2750which consists of a 5-tuple (protocol, source and destination IP addresses, source and destination TCP/UDP/SCTP port)
2751and routes packets into one of the receive queues::
2752
2753   5tuple_filter (port_id) (add|del) dst_ip (dst_address) src_ip \
2754                 (src_address) dst_port (dst_port_value) \
2755                 src_port (src_port_value) protocol (protocol_value) \
2756                 mask (mask_value) tcp_flags (tcp_flags_value) \
2757                 priority (prio_value) queue (queue_id)
2758
2759The available information parameters are:
2760
2761* ``port_id``: The port which the 5-tuple filter assigned on.
2762
2763* ``dst_address``: Destination IP address.
2764
2765* ``src_address``: Source IP address.
2766
2767* ``dst_port_value``: TCP/UDP destination port.
2768
2769* ``src_port_value``: TCP/UDP source port.
2770
2771* ``protocol_value``: L4 protocol.
2772
2773* ``mask_value``: Participates in the match or not by bit for field above, 1b means participate
2774
2775* ``tcp_flags_value``: TCP control bits. The non-zero value is invalid, when the protocol_value is not set to 0x06 (TCP).
2776
2777* ``prio_value``: The priority of this filter.
2778
2779* ``queue_id``: The receive queue associated with this 5-tuple filter.
2780
2781Example, to add/remove an 5tuple filter rule::
2782
2783   testpmd> 5tuple_filter 0 add dst_ip 2.2.2.5 src_ip 2.2.2.4 \
2784            dst_port 64 src_port 32 protocol 0x06 mask 0x1F \
2785            flags 0x0 priority 3 queue 3
2786
2787   testpmd> 5tuple_filter 0 del dst_ip 2.2.2.5 src_ip 2.2.2.4 \
2788            dst_port 64 src_port 32 protocol 0x06 mask 0x1F \
2789            flags 0x0 priority 3 queue 3
2790
2791syn_filter
2792~~~~~~~~~~
2793
2794Using the  SYN filter, TCP packets whose *SYN* flag is set can be forwarded to a separate queue::
2795
2796   syn_filter (port_id) (add|del) priority (high|low) queue (queue_id)
2797
2798The available information parameters are:
2799
2800* ``port_id``: The port which the SYN filter assigned on.
2801
2802* ``high``: This SYN filter has higher priority than other filters.
2803
2804* ``low``: This SYN filter has lower priority than other filters.
2805
2806* ``queue_id``: The receive queue associated with this SYN filter
2807
2808Example::
2809
2810   testpmd> syn_filter 0 add priority high queue 3
2811
2812flex_filter
2813~~~~~~~~~~~
2814
2815With flex filter, packets can be recognized by any arbitrary pattern within the first 128 bytes of the packet
2816and routed into one of the receive queues::
2817
2818   flex_filter (port_id) (add|del) len (len_value) bytes (bytes_value) \
2819               mask (mask_value) priority (prio_value) queue (queue_id)
2820
2821The available information parameters are:
2822
2823* ``port_id``: The port which the Flex filter is assigned on.
2824
2825* ``len_value``: Filter length in bytes, no greater than 128.
2826
2827* ``bytes_value``: A string in hexadecimal, means the value the flex filter needs to match.
2828
2829* ``mask_value``: A string in hexadecimal, bit 1 means corresponding byte participates in the match.
2830
2831* ``prio_value``: The priority of this filter.
2832
2833* ``queue_id``: The receive queue associated with this Flex filter.
2834
2835Example::
2836
2837   testpmd> flex_filter 0 add len 16 bytes 0x00000000000000000000000008060000 \
2838                          mask 000C priority 3 queue 3
2839
2840   testpmd> flex_filter 0 del len 16 bytes 0x00000000000000000000000008060000 \
2841                          mask 000C priority 3 queue 3
2842
2843
2844.. _testpmd_flow_director:
2845
2846flow_director_filter
2847~~~~~~~~~~~~~~~~~~~~
2848
2849The Flow Director works in receive mode to identify specific flows or sets of flows and route them to specific queues.
2850
2851Four types of filtering are supported which are referred to as Perfect Match, Signature, Perfect-mac-vlan and
2852Perfect-tunnel filters, the match mode is set by the ``--pkt-filter-mode`` command-line parameter:
2853
2854* Perfect match filters.
2855  The hardware checks a match between the masked fields of the received packets and the programmed filters.
2856  The masked fields are for IP flow.
2857
2858* Signature filters.
2859  The hardware checks a match between a hash-based signature of the masked fields of the received packet.
2860
2861* Perfect-mac-vlan match filters.
2862  The hardware checks a match between the masked fields of the received packets and the programmed filters.
2863  The masked fields are for MAC VLAN flow.
2864
2865* Perfect-tunnel match filters.
2866  The hardware checks a match between the masked fields of the received packets and the programmed filters.
2867  The masked fields are for tunnel flow.
2868
2869* Perfect-raw-flow-type match filters.
2870  The hardware checks a match between the masked fields of the received packets and pre-loaded raw (template) packet.
2871  The masked fields are specified by input sets.
2872
2873The Flow Director filters can match the different fields for different type of packet: flow type, specific input set
2874per flow type and the flexible payload.
2875
2876The Flow Director can also mask out parts of all of these fields so that filters
2877are only applied to certain fields or parts of the fields.
2878
2879Note that for raw flow type mode the source and destination fields in the
2880raw packet buffer need to be presented in a reversed order with respect
2881to the expected received packets.
2882For example: IP source and destination addresses or TCP/UDP/SCTP
2883source and destination ports
2884
2885Different NICs may have different capabilities, command show port fdir (port_id) can be used to acquire the information.
2886
2887# Commands to add flow director filters of different flow types::
2888
2889   flow_director_filter (port_id) mode IP (add|del|update) \
2890                        flow (ipv4-other|ipv4-frag|ipv6-other|ipv6-frag) \
2891                        src (src_ip_address) dst (dst_ip_address) \
2892                        tos (tos_value) proto (proto_value) ttl (ttl_value) \
2893                        vlan (vlan_value) flexbytes (flexbytes_value) \
2894                        (drop|fwd) pf|vf(vf_id) queue (queue_id) \
2895                        fd_id (fd_id_value)
2896
2897   flow_director_filter (port_id) mode IP (add|del|update) \
2898                        flow (ipv4-tcp|ipv4-udp|ipv6-tcp|ipv6-udp) \
2899                        src (src_ip_address) (src_port) \
2900                        dst (dst_ip_address) (dst_port) \
2901                        tos (tos_value) ttl (ttl_value) \
2902                        vlan (vlan_value) flexbytes (flexbytes_value) \
2903                        (drop|fwd) queue pf|vf(vf_id) (queue_id) \
2904                        fd_id (fd_id_value)
2905
2906   flow_director_filter (port_id) mode IP (add|del|update) \
2907                        flow (ipv4-sctp|ipv6-sctp) \
2908                        src (src_ip_address) (src_port) \
2909                        dst (dst_ip_address) (dst_port) \
2910                        tos (tos_value) ttl (ttl_value) \
2911                        tag (verification_tag) vlan (vlan_value) \
2912                        flexbytes (flexbytes_value) (drop|fwd) \
2913                        pf|vf(vf_id) queue (queue_id) fd_id (fd_id_value)
2914
2915   flow_director_filter (port_id) mode IP (add|del|update) flow l2_payload \
2916                        ether (ethertype) flexbytes (flexbytes_value) \
2917                        (drop|fwd) pf|vf(vf_id) queue (queue_id)
2918                        fd_id (fd_id_value)
2919
2920   flow_director_filter (port_id) mode MAC-VLAN (add|del|update) \
2921                        mac (mac_address) vlan (vlan_value) \
2922                        flexbytes (flexbytes_value) (drop|fwd) \
2923                        queue (queue_id) fd_id (fd_id_value)
2924
2925   flow_director_filter (port_id) mode Tunnel (add|del|update) \
2926                        mac (mac_address) vlan (vlan_value) \
2927                        tunnel (NVGRE|VxLAN) tunnel-id (tunnel_id_value) \
2928                        flexbytes (flexbytes_value) (drop|fwd) \
2929                        queue (queue_id) fd_id (fd_id_value)
2930
2931   flow_director_filter (port_id) mode raw (add|del|update) flow (flow_id) \
2932                        (drop|fwd) queue (queue_id) fd_id (fd_id_value) \
2933                        packet (packet file name)
2934
2935For example, to add an ipv4-udp flow type filter::
2936
2937   testpmd> flow_director_filter 0 mode IP add flow ipv4-udp src 2.2.2.3 32 \
2938            dst 2.2.2.5 33 tos 2 ttl 40 vlan 0x1 flexbytes (0x88,0x48) \
2939            fwd pf queue 1 fd_id 1
2940
2941For example, add an ipv4-other flow type filter::
2942
2943   testpmd> flow_director_filter 0 mode IP add flow ipv4-other src 2.2.2.3 \
2944             dst 2.2.2.5 tos 2 proto 20 ttl 40 vlan 0x1 \
2945             flexbytes (0x88,0x48) fwd pf queue 1 fd_id 1
2946
2947flush_flow_director
2948~~~~~~~~~~~~~~~~~~~
2949
2950Flush all flow director filters on a device::
2951
2952   testpmd> flush_flow_director (port_id)
2953
2954Example, to flush all flow director filter on port 0::
2955
2956   testpmd> flush_flow_director 0
2957
2958flow_director_mask
2959~~~~~~~~~~~~~~~~~~
2960
2961Set flow director's input masks::
2962
2963   flow_director_mask (port_id) mode IP vlan (vlan_value) \
2964                      src_mask (ipv4_src) (ipv6_src) (src_port) \
2965                      dst_mask (ipv4_dst) (ipv6_dst) (dst_port)
2966
2967   flow_director_mask (port_id) mode MAC-VLAN vlan (vlan_value)
2968
2969   flow_director_mask (port_id) mode Tunnel vlan (vlan_value) \
2970                      mac (mac_value) tunnel-type (tunnel_type_value) \
2971                      tunnel-id (tunnel_id_value)
2972
2973Example, to set flow director mask on port 0::
2974
2975   testpmd> flow_director_mask 0 mode IP vlan 0xefff \
2976            src_mask 255.255.255.255 \
2977                FFFF:FFFF:FFFF:FFFF:FFFF:FFFF:FFFF:FFFF 0xFFFF \
2978            dst_mask 255.255.255.255 \
2979                FFFF:FFFF:FFFF:FFFF:FFFF:FFFF:FFFF:FFFF 0xFFFF
2980
2981flow_director_flex_mask
2982~~~~~~~~~~~~~~~~~~~~~~~
2983
2984set masks of flow director's flexible payload based on certain flow type::
2985
2986   testpmd> flow_director_flex_mask (port_id) \
2987            flow (none|ipv4-other|ipv4-frag|ipv4-tcp|ipv4-udp|ipv4-sctp| \
2988                  ipv6-other|ipv6-frag|ipv6-tcp|ipv6-udp|ipv6-sctp| \
2989                  l2_payload|all) (mask)
2990
2991Example, to set flow director's flex mask for all flow type on port 0::
2992
2993   testpmd> flow_director_flex_mask 0 flow all \
2994            (0xff,0xff,0,0,0,0,0,0,0,0,0,0,0,0,0,0)
2995
2996
2997flow_director_flex_payload
2998~~~~~~~~~~~~~~~~~~~~~~~~~~
2999
3000Configure flexible payload selection::
3001
3002   flow_director_flex_payload (port_id) (raw|l2|l3|l4) (config)
3003
3004For example, to select the first 16 bytes from the offset 4 (bytes) of packet's payload as flexible payload::
3005
3006   testpmd> flow_director_flex_payload 0 l4 \
3007            (4,5,6,7,8,9,10,11,12,13,14,15,16,17,18,19)
3008
3009get_sym_hash_ena_per_port
3010~~~~~~~~~~~~~~~~~~~~~~~~~
3011
3012Get symmetric hash enable configuration per port::
3013
3014   get_sym_hash_ena_per_port (port_id)
3015
3016For example, to get symmetric hash enable configuration of port 1::
3017
3018   testpmd> get_sym_hash_ena_per_port 1
3019
3020set_sym_hash_ena_per_port
3021~~~~~~~~~~~~~~~~~~~~~~~~~
3022
3023Set symmetric hash enable configuration per port to enable or disable::
3024
3025   set_sym_hash_ena_per_port (port_id) (enable|disable)
3026
3027For example, to set symmetric hash enable configuration of port 1 to enable::
3028
3029   testpmd> set_sym_hash_ena_per_port 1 enable
3030
3031get_hash_global_config
3032~~~~~~~~~~~~~~~~~~~~~~
3033
3034Get the global configurations of hash filters::
3035
3036   get_hash_global_config (port_id)
3037
3038For example, to get the global configurations of hash filters of port 1::
3039
3040   testpmd> get_hash_global_config 1
3041
3042set_hash_global_config
3043~~~~~~~~~~~~~~~~~~~~~~
3044
3045Set the global configurations of hash filters::
3046
3047   set_hash_global_config (port_id) (toeplitz|simple_xor|default) \
3048   (ipv4|ipv4-frag|ipv4-tcp|ipv4-udp|ipv4-sctp|ipv4-other|ipv6|ipv6-frag| \
3049   ipv6-tcp|ipv6-udp|ipv6-sctp|ipv6-other|l2_payload|<flow_id>) \
3050   (enable|disable)
3051
3052For example, to enable simple_xor for flow type of ipv6 on port 2::
3053
3054   testpmd> set_hash_global_config 2 simple_xor ipv6 enable
3055
3056set_hash_input_set
3057~~~~~~~~~~~~~~~~~~
3058
3059Set the input set for hash::
3060
3061   set_hash_input_set (port_id) (ipv4-frag|ipv4-tcp|ipv4-udp|ipv4-sctp| \
3062   ipv4-other|ipv6-frag|ipv6-tcp|ipv6-udp|ipv6-sctp|ipv6-other| \
3063   l2_payload|<flow_id>) (ovlan|ivlan|src-ipv4|dst-ipv4|src-ipv6|dst-ipv6| \
3064   ipv4-tos|ipv4-proto|ipv6-tc|ipv6-next-header|udp-src-port|udp-dst-port| \
3065   tcp-src-port|tcp-dst-port|sctp-src-port|sctp-dst-port|sctp-veri-tag| \
3066   udp-key|gre-key|fld-1st|fld-2nd|fld-3rd|fld-4th|fld-5th|fld-6th|fld-7th| \
3067   fld-8th|none) (select|add)
3068
3069For example, to add source IP to hash input set for flow type of ipv4-udp on port 0::
3070
3071   testpmd> set_hash_input_set 0 ipv4-udp src-ipv4 add
3072
3073set_fdir_input_set
3074~~~~~~~~~~~~~~~~~~
3075
3076The Flow Director filters can match the different fields for different type of packet, i.e. specific input set
3077on per flow type and the flexible payload. This command can be used to change input set for each flow type.
3078
3079Set the input set for flow director::
3080
3081   set_fdir_input_set (port_id) (ipv4-frag|ipv4-tcp|ipv4-udp|ipv4-sctp| \
3082   ipv4-other|ipv6|ipv6-frag|ipv6-tcp|ipv6-udp|ipv6-sctp|ipv6-other| \
3083   l2_payload|<flow_id>) (ivlan|ethertype|src-ipv4|dst-ipv4|src-ipv6|dst-ipv6| \
3084   ipv4-tos|ipv4-proto|ipv4-ttl|ipv6-tc|ipv6-next-header|ipv6-hop-limits| \
3085   tudp-src-port|udp-dst-port|cp-src-port|tcp-dst-port|sctp-src-port| \
3086   sctp-dst-port|sctp-veri-tag|none) (select|add)
3087
3088For example to add source IP to FD input set for flow type of ipv4-udp on port 0::
3089
3090   testpmd> set_fdir_input_set 0 ipv4-udp src-ipv4 add
3091
3092global_config
3093~~~~~~~~~~~~~
3094
3095Set different GRE key length for input set::
3096
3097   global_config (port_id) gre-key-len (number in bytes)
3098
3099For example to set GRE key length for input set to 4 bytes on port 0::
3100
3101   testpmd> global_config 0 gre-key-len 4
3102
3103
3104.. _testpmd_rte_flow:
3105
3106Flow rules management
3107---------------------
3108
3109Control of the generic flow API (*rte_flow*) is fully exposed through the
3110``flow`` command (validation, creation, destruction, queries and operation
3111modes).
3112
3113Considering *rte_flow* overlaps with all `Filter Functions`_, using both
3114features simultaneously may cause undefined side-effects and is therefore
3115not recommended.
3116
3117``flow`` syntax
3118~~~~~~~~~~~~~~~
3119
3120Because the ``flow`` command uses dynamic tokens to handle the large number
3121of possible flow rules combinations, its behavior differs slightly from
3122other commands, in particular:
3123
3124- Pressing *?* or the *<tab>* key displays contextual help for the current
3125  token, not that of the entire command.
3126
3127- Optional and repeated parameters are supported (provided they are listed
3128  in the contextual help).
3129
3130The first parameter stands for the operation mode. Possible operations and
3131their general syntax are described below. They are covered in detail in the
3132following sections.
3133
3134- Check whether a flow rule can be created::
3135
3136   flow validate {port_id}
3137       [group {group_id}] [priority {level}] [ingress] [egress] [transfer]
3138       pattern {item} [/ {item} [...]] / end
3139       actions {action} [/ {action} [...]] / end
3140
3141- Create a flow rule::
3142
3143   flow create {port_id}
3144       [group {group_id}] [priority {level}] [ingress] [egress] [transfer]
3145       pattern {item} [/ {item} [...]] / end
3146       actions {action} [/ {action} [...]] / end
3147
3148- Destroy specific flow rules::
3149
3150   flow destroy {port_id} rule {rule_id} [...]
3151
3152- Destroy all flow rules::
3153
3154   flow flush {port_id}
3155
3156- Query an existing flow rule::
3157
3158   flow query {port_id} {rule_id} {action}
3159
3160- List existing flow rules sorted by priority, filtered by group
3161  identifiers::
3162
3163   flow list {port_id} [group {group_id}] [...]
3164
3165- Restrict ingress traffic to the defined flow rules::
3166
3167   flow isolate {port_id} {boolean}
3168
3169Validating flow rules
3170~~~~~~~~~~~~~~~~~~~~~
3171
3172``flow validate`` reports whether a flow rule would be accepted by the
3173underlying device in its current state but stops short of creating it. It is
3174bound to ``rte_flow_validate()``::
3175
3176   flow validate {port_id}
3177      [group {group_id}] [priority {level}] [ingress] [egress] [transfer]
3178      pattern {item} [/ {item} [...]] / end
3179      actions {action} [/ {action} [...]] / end
3180
3181If successful, it will show::
3182
3183   Flow rule validated
3184
3185Otherwise it will show an error message of the form::
3186
3187   Caught error type [...] ([...]): [...]
3188
3189This command uses the same parameters as ``flow create``, their format is
3190described in `Creating flow rules`_.
3191
3192Check whether redirecting any Ethernet packet received on port 0 to RX queue
3193index 6 is supported::
3194
3195   testpmd> flow validate 0 ingress pattern eth / end
3196      actions queue index 6 / end
3197   Flow rule validated
3198   testpmd>
3199
3200Port 0 does not support TCPv6 rules::
3201
3202   testpmd> flow validate 0 ingress pattern eth / ipv6 / tcp / end
3203      actions drop / end
3204   Caught error type 9 (specific pattern item): Invalid argument
3205   testpmd>
3206
3207Creating flow rules
3208~~~~~~~~~~~~~~~~~~~
3209
3210``flow create`` validates and creates the specified flow rule. It is bound
3211to ``rte_flow_create()``::
3212
3213   flow create {port_id}
3214      [group {group_id}] [priority {level}] [ingress] [egress] [transfer]
3215      pattern {item} [/ {item} [...]] / end
3216      actions {action} [/ {action} [...]] / end
3217
3218If successful, it will return a flow rule ID usable with other commands::
3219
3220   Flow rule #[...] created
3221
3222Otherwise it will show an error message of the form::
3223
3224   Caught error type [...] ([...]): [...]
3225
3226Parameters describe in the following order:
3227
3228- Attributes (*group*, *priority*, *ingress*, *egress*, *transfer* tokens).
3229- A matching pattern, starting with the *pattern* token and terminated by an
3230  *end* pattern item.
3231- Actions, starting with the *actions* token and terminated by an *end*
3232  action.
3233
3234These translate directly to *rte_flow* objects provided as-is to the
3235underlying functions.
3236
3237The shortest valid definition only comprises mandatory tokens::
3238
3239   testpmd> flow create 0 pattern end actions end
3240
3241Note that PMDs may refuse rules that essentially do nothing such as this
3242one.
3243
3244**All unspecified object values are automatically initialized to 0.**
3245
3246Attributes
3247^^^^^^^^^^
3248
3249These tokens affect flow rule attributes (``struct rte_flow_attr``) and are
3250specified before the ``pattern`` token.
3251
3252- ``group {group id}``: priority group.
3253- ``priority {level}``: priority level within group.
3254- ``ingress``: rule applies to ingress traffic.
3255- ``egress``: rule applies to egress traffic.
3256- ``transfer``: apply rule directly to endpoints found in pattern.
3257
3258Each instance of an attribute specified several times overrides the previous
3259value as shown below (group 4 is used)::
3260
3261   testpmd> flow create 0 group 42 group 24 group 4 [...]
3262
3263Note that once enabled, ``ingress`` and ``egress`` cannot be disabled.
3264
3265While not specifying a direction is an error, some rules may allow both
3266simultaneously.
3267
3268Most rules affect RX therefore contain the ``ingress`` token::
3269
3270   testpmd> flow create 0 ingress pattern [...]
3271
3272Matching pattern
3273^^^^^^^^^^^^^^^^
3274
3275A matching pattern starts after the ``pattern`` token. It is made of pattern
3276items and is terminated by a mandatory ``end`` item.
3277
3278Items are named after their type (*RTE_FLOW_ITEM_TYPE_* from ``enum
3279rte_flow_item_type``).
3280
3281The ``/`` token is used as a separator between pattern items as shown
3282below::
3283
3284   testpmd> flow create 0 ingress pattern eth / ipv4 / udp / end [...]
3285
3286Note that protocol items like these must be stacked from lowest to highest
3287layer to make sense. For instance, the following rule is either invalid or
3288unlikely to match any packet::
3289
3290   testpmd> flow create 0 ingress pattern eth / udp / ipv4 / end [...]
3291
3292More information on these restrictions can be found in the *rte_flow*
3293documentation.
3294
3295Several items support additional specification structures, for example
3296``ipv4`` allows specifying source and destination addresses as follows::
3297
3298   testpmd> flow create 0 ingress pattern eth / ipv4 src is 10.1.1.1
3299      dst is 10.2.0.0 / end [...]
3300
3301This rule matches all IPv4 traffic with the specified properties.
3302
3303In this example, ``src`` and ``dst`` are field names of the underlying
3304``struct rte_flow_item_ipv4`` object. All item properties can be specified
3305in a similar fashion.
3306
3307The ``is`` token means that the subsequent value must be matched exactly,
3308and assigns ``spec`` and ``mask`` fields in ``struct rte_flow_item``
3309accordingly. Possible assignment tokens are:
3310
3311- ``is``: match value perfectly (with full bit-mask).
3312- ``spec``: match value according to configured bit-mask.
3313- ``last``: specify upper bound to establish a range.
3314- ``mask``: specify bit-mask with relevant bits set to one.
3315- ``prefix``: generate bit-mask from a prefix length.
3316
3317These yield identical results::
3318
3319   ipv4 src is 10.1.1.1
3320
3321::
3322
3323   ipv4 src spec 10.1.1.1 src mask 255.255.255.255
3324
3325::
3326
3327   ipv4 src spec 10.1.1.1 src prefix 32
3328
3329::
3330
3331   ipv4 src is 10.1.1.1 src last 10.1.1.1 # range with a single value
3332
3333::
3334
3335   ipv4 src is 10.1.1.1 src last 0 # 0 disables range
3336
3337Inclusive ranges can be defined with ``last``::
3338
3339   ipv4 src is 10.1.1.1 src last 10.2.3.4 # 10.1.1.1 to 10.2.3.4
3340
3341Note that ``mask`` affects both ``spec`` and ``last``::
3342
3343   ipv4 src is 10.1.1.1 src last 10.2.3.4 src mask 255.255.0.0
3344      # matches 10.1.0.0 to 10.2.255.255
3345
3346Properties can be modified multiple times::
3347
3348   ipv4 src is 10.1.1.1 src is 10.1.2.3 src is 10.2.3.4 # matches 10.2.3.4
3349
3350::
3351
3352   ipv4 src is 10.1.1.1 src prefix 24 src prefix 16 # matches 10.1.0.0/16
3353
3354Pattern items
3355^^^^^^^^^^^^^
3356
3357This section lists supported pattern items and their attributes, if any.
3358
3359- ``end``: end list of pattern items.
3360
3361- ``void``: no-op pattern item.
3362
3363- ``invert``: perform actions when pattern does not match.
3364
3365- ``any``: match any protocol for the current layer.
3366
3367  - ``num {unsigned}``: number of layers covered.
3368
3369- ``pf``: match traffic from/to the physical function.
3370
3371- ``vf``: match traffic from/to a virtual function ID.
3372
3373  - ``id {unsigned}``: VF ID.
3374
3375- ``phy_port``: match traffic from/to a specific physical port.
3376
3377  - ``index {unsigned}``: physical port index.
3378
3379- ``port_id``: match traffic from/to a given DPDK port ID.
3380
3381  - ``id {unsigned}``: DPDK port ID.
3382
3383- ``mark``: match value set in previously matched flow rule using the mark action.
3384
3385  - ``id {unsigned}``: arbitrary integer value.
3386
3387- ``raw``: match an arbitrary byte string.
3388
3389  - ``relative {boolean}``: look for pattern after the previous item.
3390  - ``search {boolean}``: search pattern from offset (see also limit).
3391  - ``offset {integer}``: absolute or relative offset for pattern.
3392  - ``limit {unsigned}``: search area limit for start of pattern.
3393  - ``pattern {string}``: byte string to look for.
3394
3395- ``eth``: match Ethernet header.
3396
3397  - ``dst {MAC-48}``: destination MAC.
3398  - ``src {MAC-48}``: source MAC.
3399  - ``type {unsigned}``: EtherType or TPID.
3400
3401- ``vlan``: match 802.1Q/ad VLAN tag.
3402
3403  - ``tci {unsigned}``: tag control information.
3404  - ``pcp {unsigned}``: priority code point.
3405  - ``dei {unsigned}``: drop eligible indicator.
3406  - ``vid {unsigned}``: VLAN identifier.
3407  - ``inner_type {unsigned}``: inner EtherType or TPID.
3408
3409- ``ipv4``: match IPv4 header.
3410
3411  - ``tos {unsigned}``: type of service.
3412  - ``ttl {unsigned}``: time to live.
3413  - ``proto {unsigned}``: next protocol ID.
3414  - ``src {ipv4 address}``: source address.
3415  - ``dst {ipv4 address}``: destination address.
3416
3417- ``ipv6``: match IPv6 header.
3418
3419  - ``tc {unsigned}``: traffic class.
3420  - ``flow {unsigned}``: flow label.
3421  - ``proto {unsigned}``: protocol (next header).
3422  - ``hop {unsigned}``: hop limit.
3423  - ``src {ipv6 address}``: source address.
3424  - ``dst {ipv6 address}``: destination address.
3425
3426- ``icmp``: match ICMP header.
3427
3428  - ``type {unsigned}``: ICMP packet type.
3429  - ``code {unsigned}``: ICMP packet code.
3430
3431- ``udp``: match UDP header.
3432
3433  - ``src {unsigned}``: UDP source port.
3434  - ``dst {unsigned}``: UDP destination port.
3435
3436- ``tcp``: match TCP header.
3437
3438  - ``src {unsigned}``: TCP source port.
3439  - ``dst {unsigned}``: TCP destination port.
3440
3441- ``sctp``: match SCTP header.
3442
3443  - ``src {unsigned}``: SCTP source port.
3444  - ``dst {unsigned}``: SCTP destination port.
3445  - ``tag {unsigned}``: validation tag.
3446  - ``cksum {unsigned}``: checksum.
3447
3448- ``vxlan``: match VXLAN header.
3449
3450  - ``vni {unsigned}``: VXLAN identifier.
3451
3452- ``e_tag``: match IEEE 802.1BR E-Tag header.
3453
3454  - ``grp_ecid_b {unsigned}``: GRP and E-CID base.
3455
3456- ``nvgre``: match NVGRE header.
3457
3458  - ``tni {unsigned}``: virtual subnet ID.
3459
3460- ``mpls``: match MPLS header.
3461
3462  - ``label {unsigned}``: MPLS label.
3463
3464- ``gre``: match GRE header.
3465
3466  - ``protocol {unsigned}``: protocol type.
3467
3468- ``fuzzy``: fuzzy pattern match, expect faster than default.
3469
3470  - ``thresh {unsigned}``: accuracy threshold.
3471
3472- ``gtp``, ``gtpc``, ``gtpu``: match GTPv1 header.
3473
3474  - ``teid {unsigned}``: tunnel endpoint identifier.
3475
3476- ``geneve``: match GENEVE header.
3477
3478  - ``vni {unsigned}``: virtual network identifier.
3479  - ``protocol {unsigned}``: protocol type.
3480
3481- ``vxlan-gpe``: match VXLAN-GPE header.
3482
3483  - ``vni {unsigned}``: VXLAN-GPE identifier.
3484
3485- ``arp_eth_ipv4``: match ARP header for Ethernet/IPv4.
3486
3487  - ``sha {MAC-48}``: sender hardware address.
3488  - ``spa {ipv4 address}``: sender IPv4 address.
3489  - ``tha {MAC-48}``: target hardware address.
3490  - ``tpa {ipv4 address}``: target IPv4 address.
3491
3492- ``ipv6_ext``: match presence of any IPv6 extension header.
3493
3494  - ``next_hdr {unsigned}``: next header.
3495
3496- ``icmp6``: match any ICMPv6 header.
3497
3498  - ``type {unsigned}``: ICMPv6 type.
3499  - ``code {unsigned}``: ICMPv6 code.
3500
3501- ``icmp6_nd_ns``: match ICMPv6 neighbor discovery solicitation.
3502
3503  - ``target_addr {ipv6 address}``: target address.
3504
3505- ``icmp6_nd_na``: match ICMPv6 neighbor discovery advertisement.
3506
3507  - ``target_addr {ipv6 address}``: target address.
3508
3509- ``icmp6_nd_opt``: match presence of any ICMPv6 neighbor discovery option.
3510
3511  - ``type {unsigned}``: ND option type.
3512
3513- ``icmp6_nd_opt_sla_eth``: match ICMPv6 neighbor discovery source Ethernet
3514  link-layer address option.
3515
3516  - ``sla {MAC-48}``: source Ethernet LLA.
3517
3518- ``icmp6_nd_opt_sla_eth``: match ICMPv6 neighbor discovery target Ethernet
3519  link-layer address option.
3520
3521  - ``tla {MAC-48}``: target Ethernet LLA.
3522
3523Actions list
3524^^^^^^^^^^^^
3525
3526A list of actions starts after the ``actions`` token in the same fashion as
3527`Matching pattern`_; actions are separated by ``/`` tokens and the list is
3528terminated by a mandatory ``end`` action.
3529
3530Actions are named after their type (*RTE_FLOW_ACTION_TYPE_* from ``enum
3531rte_flow_action_type``).
3532
3533Dropping all incoming UDPv4 packets can be expressed as follows::
3534
3535   testpmd> flow create 0 ingress pattern eth / ipv4 / udp / end
3536      actions drop / end
3537
3538Several actions have configurable properties which must be specified when
3539there is no valid default value. For example, ``queue`` requires a target
3540queue index.
3541
3542This rule redirects incoming UDPv4 traffic to queue index 6::
3543
3544   testpmd> flow create 0 ingress pattern eth / ipv4 / udp / end
3545      actions queue index 6 / end
3546
3547While this one could be rejected by PMDs (unspecified queue index)::
3548
3549   testpmd> flow create 0 ingress pattern eth / ipv4 / udp / end
3550      actions queue / end
3551
3552As defined by *rte_flow*, the list is not ordered, all actions of a given
3553rule are performed simultaneously. These are equivalent::
3554
3555   queue index 6 / void / mark id 42 / end
3556
3557::
3558
3559   void / mark id 42 / queue index 6 / end
3560
3561All actions in a list should have different types, otherwise only the last
3562action of a given type is taken into account::
3563
3564   queue index 4 / queue index 5 / queue index 6 / end # will use queue 6
3565
3566::
3567
3568   drop / drop / drop / end # drop is performed only once
3569
3570::
3571
3572   mark id 42 / queue index 3 / mark id 24 / end # mark will be 24
3573
3574Considering they are performed simultaneously, opposite and overlapping
3575actions can sometimes be combined when the end result is unambiguous::
3576
3577   drop / queue index 6 / end # drop has no effect
3578
3579::
3580
3581   queue index 6 / rss queues 6 7 8 / end # queue has no effect
3582
3583::
3584
3585   drop / passthru / end # drop has no effect
3586
3587Note that PMDs may still refuse such combinations.
3588
3589Actions
3590^^^^^^^
3591
3592This section lists supported actions and their attributes, if any.
3593
3594- ``end``: end list of actions.
3595
3596- ``void``: no-op action.
3597
3598- ``passthru``: let subsequent rule process matched packets.
3599
3600- ``jump``: redirect traffic to group on device.
3601
3602  - ``group {unsigned}``: group to redirect to.
3603
3604- ``mark``: attach 32 bit value to packets.
3605
3606  - ``id {unsigned}``: 32 bit value to return with packets.
3607
3608- ``flag``: flag packets.
3609
3610- ``queue``: assign packets to a given queue index.
3611
3612  - ``index {unsigned}``: queue index to use.
3613
3614- ``drop``: drop packets (note: passthru has priority).
3615
3616- ``count``: enable counters for this rule.
3617
3618- ``rss``: spread packets among several queues.
3619
3620  - ``func {hash function}``: RSS hash function to apply, allowed tokens are
3621    the same as `set_hash_global_config`_.
3622
3623  - ``level {unsigned}``: encapsulation level for ``types``.
3624
3625  - ``types [{RSS hash type} [...]] end``: specific RSS hash types, allowed
3626    tokens are the same as `set_hash_input_set`_, except that an empty list
3627    does not disable RSS but instead requests unspecified "best-effort"
3628    settings.
3629
3630  - ``key {string}``: RSS hash key, overrides ``key_len``.
3631
3632  - ``key_len {unsigned}``: RSS hash key length in bytes, can be used in
3633    conjunction with ``key`` to pad or truncate it.
3634
3635  - ``queues [{unsigned} [...]] end``: queue indices to use.
3636
3637- ``pf``: direct traffic to physical function.
3638
3639- ``vf``: direct traffic to a virtual function ID.
3640
3641  - ``original {boolean}``: use original VF ID if possible.
3642  - ``id {unsigned}``: VF ID.
3643
3644- ``phy_port``: direct packets to physical port index.
3645
3646  - ``original {boolean}``: use original port index if possible.
3647  - ``index {unsigned}``: physical port index.
3648
3649- ``port_id``: direct matching traffic to a given DPDK port ID.
3650
3651  - ``original {boolean}``: use original DPDK port ID if possible.
3652  - ``id {unsigned}``: DPDK port ID.
3653
3654- ``of_set_mpls_ttl``: OpenFlow's ``OFPAT_SET_MPLS_TTL``.
3655
3656  - ``mpls_ttl``: MPLS TTL.
3657
3658- ``of_dec_mpls_ttl``: OpenFlow's ``OFPAT_DEC_MPLS_TTL``.
3659
3660- ``of_set_nw_ttl``: OpenFlow's ``OFPAT_SET_NW_TTL``.
3661
3662  - ``nw_ttl``: IP TTL.
3663
3664- ``of_dec_nw_ttl``: OpenFlow's ``OFPAT_DEC_NW_TTL``.
3665
3666- ``of_copy_ttl_out``: OpenFlow's ``OFPAT_COPY_TTL_OUT``.
3667
3668- ``of_copy_ttl_in``: OpenFlow's ``OFPAT_COPY_TTL_IN``.
3669
3670- ``of_pop_vlan``: OpenFlow's ``OFPAT_POP_VLAN``.
3671
3672- ``of_push_vlan``: OpenFlow's ``OFPAT_PUSH_VLAN``.
3673
3674  - ``ethertype``: Ethertype.
3675
3676- ``of_set_vlan_vid``: OpenFlow's ``OFPAT_SET_VLAN_VID``.
3677
3678  - ``vlan_vid``: VLAN id.
3679
3680- ``of_set_vlan_pcp``: OpenFlow's ``OFPAT_SET_VLAN_PCP``.
3681
3682  - ``vlan_pcp``: VLAN priority.
3683
3684- ``of_pop_mpls``: OpenFlow's ``OFPAT_POP_MPLS``.
3685
3686  - ``ethertype``: Ethertype.
3687
3688- ``of_push_mpls``: OpenFlow's ``OFPAT_PUSH_MPLS``.
3689
3690  - ``ethertype``: Ethertype.
3691
3692- ``vxlan_encap``: Performs a VXLAN encapsulation, outer layer configuration
3693  is done through `Config VXLAN Encap outer layers`_.
3694
3695- ``vxlan_decap``: Performs a decapsulation action by stripping all headers of
3696  the VXLAN tunnel network overlay from the matched flow.
3697
3698- ``nvgre_encap``: Performs a NVGRE encapsulation, outer layer configuration
3699  is done through `Config NVGRE Encap outer layers`_.
3700
3701- ``nvgre_decap``: Performs a decapsulation action by stripping all headers of
3702  the NVGRE tunnel network overlay from the matched flow.
3703
3704Destroying flow rules
3705~~~~~~~~~~~~~~~~~~~~~
3706
3707``flow destroy`` destroys one or more rules from their rule ID (as returned
3708by ``flow create``), this command calls ``rte_flow_destroy()`` as many
3709times as necessary::
3710
3711   flow destroy {port_id} rule {rule_id} [...]
3712
3713If successful, it will show::
3714
3715   Flow rule #[...] destroyed
3716
3717It does not report anything for rule IDs that do not exist. The usual error
3718message is shown when a rule cannot be destroyed::
3719
3720   Caught error type [...] ([...]): [...]
3721
3722``flow flush`` destroys all rules on a device and does not take extra
3723arguments. It is bound to ``rte_flow_flush()``::
3724
3725   flow flush {port_id}
3726
3727Any errors are reported as above.
3728
3729Creating several rules and destroying them::
3730
3731   testpmd> flow create 0 ingress pattern eth / ipv6 / end
3732      actions queue index 2 / end
3733   Flow rule #0 created
3734   testpmd> flow create 0 ingress pattern eth / ipv4 / end
3735      actions queue index 3 / end
3736   Flow rule #1 created
3737   testpmd> flow destroy 0 rule 0 rule 1
3738   Flow rule #1 destroyed
3739   Flow rule #0 destroyed
3740   testpmd>
3741
3742The same result can be achieved using ``flow flush``::
3743
3744   testpmd> flow create 0 ingress pattern eth / ipv6 / end
3745      actions queue index 2 / end
3746   Flow rule #0 created
3747   testpmd> flow create 0 ingress pattern eth / ipv4 / end
3748      actions queue index 3 / end
3749   Flow rule #1 created
3750   testpmd> flow flush 0
3751   testpmd>
3752
3753Non-existent rule IDs are ignored::
3754
3755   testpmd> flow create 0 ingress pattern eth / ipv6 / end
3756      actions queue index 2 / end
3757   Flow rule #0 created
3758   testpmd> flow create 0 ingress pattern eth / ipv4 / end
3759      actions queue index 3 / end
3760   Flow rule #1 created
3761   testpmd> flow destroy 0 rule 42 rule 10 rule 2
3762   testpmd>
3763   testpmd> flow destroy 0 rule 0
3764   Flow rule #0 destroyed
3765   testpmd>
3766
3767Querying flow rules
3768~~~~~~~~~~~~~~~~~~~
3769
3770``flow query`` queries a specific action of a flow rule having that
3771ability. Such actions collect information that can be reported using this
3772command. It is bound to ``rte_flow_query()``::
3773
3774   flow query {port_id} {rule_id} {action}
3775
3776If successful, it will display either the retrieved data for known actions
3777or the following message::
3778
3779   Cannot display result for action type [...] ([...])
3780
3781Otherwise, it will complain either that the rule does not exist or that some
3782error occurred::
3783
3784   Flow rule #[...] not found
3785
3786::
3787
3788   Caught error type [...] ([...]): [...]
3789
3790Currently only the ``count`` action is supported. This action reports the
3791number of packets that hit the flow rule and the total number of bytes. Its
3792output has the following format::
3793
3794   count:
3795    hits_set: [...] # whether "hits" contains a valid value
3796    bytes_set: [...] # whether "bytes" contains a valid value
3797    hits: [...] # number of packets
3798    bytes: [...] # number of bytes
3799
3800Querying counters for TCPv6 packets redirected to queue 6::
3801
3802   testpmd> flow create 0 ingress pattern eth / ipv6 / tcp / end
3803      actions queue index 6 / count / end
3804   Flow rule #4 created
3805   testpmd> flow query 0 4 count
3806   count:
3807    hits_set: 1
3808    bytes_set: 0
3809    hits: 386446
3810    bytes: 0
3811   testpmd>
3812
3813Listing flow rules
3814~~~~~~~~~~~~~~~~~~
3815
3816``flow list`` lists existing flow rules sorted by priority and optionally
3817filtered by group identifiers::
3818
3819   flow list {port_id} [group {group_id}] [...]
3820
3821This command only fails with the following message if the device does not
3822exist::
3823
3824   Invalid port [...]
3825
3826Output consists of a header line followed by a short description of each
3827flow rule, one per line. There is no output at all when no flow rules are
3828configured on the device::
3829
3830   ID      Group   Prio    Attr    Rule
3831   [...]   [...]   [...]   [...]   [...]
3832
3833``Attr`` column flags:
3834
3835- ``i`` for ``ingress``.
3836- ``e`` for ``egress``.
3837
3838Creating several flow rules and listing them::
3839
3840   testpmd> flow create 0 ingress pattern eth / ipv4 / end
3841      actions queue index 6 / end
3842   Flow rule #0 created
3843   testpmd> flow create 0 ingress pattern eth / ipv6 / end
3844      actions queue index 2 / end
3845   Flow rule #1 created
3846   testpmd> flow create 0 priority 5 ingress pattern eth / ipv4 / udp / end
3847      actions rss queues 6 7 8 end / end
3848   Flow rule #2 created
3849   testpmd> flow list 0
3850   ID      Group   Prio    Attr    Rule
3851   0       0       0       i-      ETH IPV4 => QUEUE
3852   1       0       0       i-      ETH IPV6 => QUEUE
3853   2       0       5       i-      ETH IPV4 UDP => RSS
3854   testpmd>
3855
3856Rules are sorted by priority (i.e. group ID first, then priority level)::
3857
3858   testpmd> flow list 1
3859   ID      Group   Prio    Attr    Rule
3860   0       0       0       i-      ETH => COUNT
3861   6       0       500     i-      ETH IPV6 TCP => DROP COUNT
3862   5       0       1000    i-      ETH IPV6 ICMP => QUEUE
3863   1       24      0       i-      ETH IPV4 UDP => QUEUE
3864   4       24      10      i-      ETH IPV4 TCP => DROP
3865   3       24      20      i-      ETH IPV4 => DROP
3866   2       24      42      i-      ETH IPV4 UDP => QUEUE
3867   7       63      0       i-      ETH IPV6 UDP VXLAN => MARK QUEUE
3868   testpmd>
3869
3870Output can be limited to specific groups::
3871
3872   testpmd> flow list 1 group 0 group 63
3873   ID      Group   Prio    Attr    Rule
3874   0       0       0       i-      ETH => COUNT
3875   6       0       500     i-      ETH IPV6 TCP => DROP COUNT
3876   5       0       1000    i-      ETH IPV6 ICMP => QUEUE
3877   7       63      0       i-      ETH IPV6 UDP VXLAN => MARK QUEUE
3878   testpmd>
3879
3880Toggling isolated mode
3881~~~~~~~~~~~~~~~~~~~~~~
3882
3883``flow isolate`` can be used to tell the underlying PMD that ingress traffic
3884must only be injected from the defined flow rules; that no default traffic
3885is expected outside those rules and the driver is free to assign more
3886resources to handle them. It is bound to ``rte_flow_isolate()``::
3887
3888 flow isolate {port_id} {boolean}
3889
3890If successful, enabling or disabling isolated mode shows either::
3891
3892 Ingress traffic on port [...]
3893    is now restricted to the defined flow rules
3894
3895Or::
3896
3897 Ingress traffic on port [...]
3898    is not restricted anymore to the defined flow rules
3899
3900Otherwise, in case of error::
3901
3902   Caught error type [...] ([...]): [...]
3903
3904Mainly due to its side effects, PMDs supporting this mode may not have the
3905ability to toggle it more than once without reinitializing affected ports
3906first (e.g. by exiting testpmd).
3907
3908Enabling isolated mode::
3909
3910 testpmd> flow isolate 0 true
3911 Ingress traffic on port 0 is now restricted to the defined flow rules
3912 testpmd>
3913
3914Disabling isolated mode::
3915
3916 testpmd> flow isolate 0 false
3917 Ingress traffic on port 0 is not restricted anymore to the defined flow rules
3918 testpmd>
3919
3920Sample QinQ flow rules
3921~~~~~~~~~~~~~~~~~~~~~~
3922
3923Before creating QinQ rule(s) the following commands should be issued to enable QinQ::
3924
3925   testpmd> port stop 0
3926   testpmd> vlan set qinq on 0
3927
3928The above command sets the inner and outer TPID's to 0x8100.
3929
3930To change the TPID's the following commands should be used::
3931
3932   testpmd> vlan set outer tpid 0xa100 0
3933   testpmd> vlan set inner tpid 0x9100 0
3934   testpmd> port start 0
3935
3936Validate and create a QinQ rule on port 0 to steer traffic to a VF queue in a VM.
3937
3938::
3939
3940   testpmd> flow validate 0 ingress pattern eth / vlan tci is 123 /
3941       vlan tci is 456 / end actions vf id 1 / queue index 0 / end
3942   Flow rule #0 validated
3943
3944   testpmd> flow create 0 ingress pattern eth / vlan tci is 4 /
3945       vlan tci is 456 / end actions vf id 123 / queue index 0 / end
3946   Flow rule #0 created
3947
3948   testpmd> flow list 0
3949   ID      Group   Prio    Attr    Rule
3950   0       0       0       i-      ETH VLAN VLAN=>VF QUEUE
3951
3952Validate and create a QinQ rule on port 0 to steer traffic to a queue on the host.
3953
3954::
3955
3956   testpmd> flow validate 0 ingress pattern eth / vlan tci is 321 /
3957        vlan tci is 654 / end actions pf / queue index 0 / end
3958   Flow rule #1 validated
3959
3960   testpmd> flow create 0 ingress pattern eth / vlan tci is 321 /
3961        vlan tci is 654 / end actions pf / queue index 1 / end
3962   Flow rule #1 created
3963
3964   testpmd> flow list 0
3965   ID      Group   Prio    Attr    Rule
3966   0       0       0       i-      ETH VLAN VLAN=>VF QUEUE
3967   1       0       0       i-      ETH VLAN VLAN=>PF QUEUE
3968
3969Sample VXLAN encapsulation rule
3970~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
3971
3972VXLAN encapsulation outer layer has default value pre-configured in testpmd
3973source code, those can be changed by using the following commands
3974
3975IPv4 VXLAN outer header::
3976
3977 testpmd> set vxlan ip-version ipv4 vni 4 udp-src 4 udp-dst 4 ip-src 127.0.0.1
3978        ip-dst 128.0.0.1 eth-src 11:11:11:11:11:11 eth-dst 22:22:22:22:22:22
3979 testpmd> flow create 0 ingress pattern end actions vxlan_encap /
3980        queue index 0 / end
3981
3982 testpmd> set vxlan-with-vlan ip-version ipv4 vni 4 udp-src 4 udp-dst 4 ip-src
3983         127.0.0.1 ip-dst 128.0.0.1 vlan-tci 34 eth-src 11:11:11:11:11:11
3984         eth-dst 22:22:22:22:22:22
3985 testpmd> flow create 0 ingress pattern end actions vxlan_encap /
3986         queue index 0 / end
3987
3988IPv6 VXLAN outer header::
3989
3990 testpmd> set vxlan ip-version ipv6 vni 4 udp-src 4 udp-dst 4 ip-src ::1
3991        ip-dst ::2222 eth-src 11:11:11:11:11:11 eth-dst 22:22:22:22:22:22
3992 testpmd> flow create 0 ingress pattern end actions vxlan_encap /
3993         queue index 0 / end
3994
3995 testpmd> set vxlan-with-vlan ip-version ipv6 vni 4 udp-src 4 udp-dst 4
3996         ip-src ::1 ip-dst ::2222 vlan-tci 34 eth-src 11:11:11:11:11:11
3997         eth-dst 22:22:22:22:22:22
3998 testpmd> flow create 0 ingress pattern end actions vxlan_encap /
3999         queue index 0 / end
4000
4001Sample NVGRE encapsulation rule
4002~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
4003
4004NVGRE encapsulation outer layer has default value pre-configured in testpmd
4005source code, those can be changed by using the following commands
4006
4007IPv4 NVGRE outer header::
4008
4009 testpmd> set nvgre ip-version ipv4 tni 4 ip-src 127.0.0.1 ip-dst 128.0.0.1
4010        eth-src 11:11:11:11:11:11 eth-dst 22:22:22:22:22:22
4011 testpmd> flow create 0 ingress pattern end actions nvgre_encap /
4012        queue index 0 / end
4013
4014 testpmd> set nvgre-with-vlan ip-version ipv4 tni 4 ip-src 127.0.0.1
4015         ip-dst 128.0.0.1 vlan-tci 34 eth-src 11:11:11:11:11:11
4016         eth-dst 22:22:22:22:22:22
4017 testpmd> flow create 0 ingress pattern end actions nvgre_encap /
4018         queue index 0 / end
4019
4020IPv6 NVGRE outer header::
4021
4022 testpmd> set nvgre ip-version ipv6 tni 4 ip-src ::1 ip-dst ::2222
4023        eth-src 11:11:11:11:11:11 eth-dst 22:22:22:22:22:22
4024 testpmd> flow create 0 ingress pattern end actions nvgre_encap /
4025        queue index 0 / end
4026
4027 testpmd> set nvgre-with-vlan ip-version ipv6 tni 4 ip-src ::1 ip-dst ::2222
4028        vlan-tci 34 eth-src 11:11:11:11:11:11 eth-dst 22:22:22:22:22:22
4029 testpmd> flow create 0 ingress pattern end actions nvgre_encap /
4030        queue index 0 / end
4031
4032BPF Functions
4033--------------
4034
4035The following sections show functions to load/unload eBPF based filters.
4036
4037bpf-load
4038~~~~~~~~
4039
4040Load an eBPF program as a callback for partciular RX/TX queue::
4041
4042   testpmd> bpf-load rx|tx (portid) (queueid) (load-flags) (bpf-prog-filename)
4043
4044The available load-flags are:
4045
4046* ``J``: use JIT generated native code, otherwise BPF interpreter will be used.
4047
4048* ``M``: assume input parameter is a pointer to rte_mbuf, otherwise assume it is a pointer to first segment's data.
4049
4050* ``-``: none.
4051
4052.. note::
4053
4054   You'll need clang v3.7 or above to build bpf program you'd like to load
4055
4056For example:
4057
4058.. code-block:: console
4059
4060   cd test/bpf
4061   clang -O2 -target bpf -c t1.c
4062
4063Then to load (and JIT compile) t1.o at RX queue 0, port 1::
4064
4065.. code-block:: console
4066
4067   testpmd> bpf-load rx 1 0 J ./dpdk.org/test/bpf/t1.o
4068
4069To load (not JITed) t1.o at TX queue 0, port 0::
4070
4071.. code-block:: console
4072
4073   testpmd> bpf-load tx 0 0 - ./dpdk.org/test/bpf/t1.o
4074
4075bpf-unload
4076~~~~~~~~~~
4077
4078Unload previously loaded eBPF program for partciular RX/TX queue::
4079
4080   testpmd> bpf-unload rx|tx (portid) (queueid)
4081
4082For example to unload BPF filter from TX queue 0, port 0:
4083
4084.. code-block:: console
4085
4086   testpmd> bpf-load tx 0 0 - ./dpdk.org/test/bpf/t1.o
4087