xref: /dpdk/doc/guides/testpmd_app_ug/testpmd_funcs.rst (revision 0bd797e9fa6c5be46f91e25263f5b4c1b9d78cf1)
1..  BSD LICENSE
2    Copyright(c) 2010-2016 Intel Corporation. All rights reserved.
3    All rights reserved.
4
5    Redistribution and use in source and binary forms, with or without
6    modification, are permitted provided that the following conditions
7    are met:
8
9    * Redistributions of source code must retain the above copyright
10    notice, this list of conditions and the following disclaimer.
11    * Redistributions in binary form must reproduce the above copyright
12    notice, this list of conditions and the following disclaimer in
13    the documentation and/or other materials provided with the
14    distribution.
15    * Neither the name of Intel Corporation nor the names of its
16    contributors may be used to endorse or promote products derived
17    from this software without specific prior written permission.
18
19    THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS
20    "AS IS" AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT
21    LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR
22    A PARTICULAR PURPOSE ARE DISCLAIMED. IN NO EVENT SHALL THE COPYRIGHT
23    OWNER OR CONTRIBUTORS BE LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL,
24    SPECIAL, EXEMPLARY, OR CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT
25    LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES; LOSS OF USE,
26    DATA, OR PROFITS; OR BUSINESS INTERRUPTION) HOWEVER CAUSED AND ON ANY
27    THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT LIABILITY, OR TORT
28    (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY OUT OF THE USE
29    OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE.
30
31.. _testpmd_runtime:
32
33Testpmd Runtime Functions
34=========================
35
36Where the testpmd application is started in interactive mode, (``-i|--interactive``),
37it displays a prompt that can be used to start and stop forwarding,
38configure the application, display statistics (including the extended NIC
39statistics aka xstats) , set the Flow Director and other tasks::
40
41   testpmd>
42
43The testpmd prompt has some, limited, readline support.
44Common bash command-line functions such as ``Ctrl+a`` and ``Ctrl+e`` to go to the start and end of the prompt line are supported
45as well as access to the command history via the up-arrow.
46
47There is also support for tab completion.
48If you type a partial command and hit ``<TAB>`` you get a list of the available completions:
49
50.. code-block:: console
51
52   testpmd> show port <TAB>
53
54       info [Mul-choice STRING]: show|clear port info|stats|xstats|fdir|stat_qmap|dcb_tc|cap X
55       info [Mul-choice STRING]: show|clear port info|stats|xstats|fdir|stat_qmap|dcb_tc|cap all
56       stats [Mul-choice STRING]: show|clear port info|stats|xstats|fdir|stat_qmap|dcb_tc|cap X
57       stats [Mul-choice STRING]: show|clear port info|stats|xstats|fdir|stat_qmap|dcb_tc|cap all
58       ...
59
60
61.. note::
62
63   Some examples in this document are too long to fit on one line are are shown wrapped at `"\\"` for display purposes::
64
65      testpmd> set flow_ctrl rx (on|off) tx (on|off) (high_water) (low_water) \
66               (pause_time) (send_xon) (port_id)
67
68In the real ``testpmd>`` prompt these commands should be on a single line.
69
70Help Functions
71--------------
72
73The testpmd has on-line help for the functions that are available at runtime.
74These are divided into sections and can be accessed using help, help section or help all:
75
76.. code-block:: console
77
78   testpmd> help
79
80       help control    : Start and stop forwarding.
81       help display    : Displaying port, stats and config information.
82       help config     : Configuration information.
83       help ports      : Configuring ports.
84       help registers  : Reading and setting port registers.
85       help filters    : Filters configuration help.
86       help all        : All of the above sections.
87
88
89Command File Functions
90----------------------
91
92To facilitate loading large number of commands or to avoid cutting and pasting where not
93practical or possible testpmd supports alternative methods for executing commands.
94
95* If started with the ``--cmdline-file=FILENAME`` command line argument testpmd
96  will execute all CLI commands contained within the file immediately before
97  starting packet forwarding or entering interactive mode.
98
99.. code-block:: console
100
101   ./testpmd -n4 -r2 ... -- -i --cmdline-file=/home/ubuntu/flow-create-commands.txt
102   Interactive-mode selected
103   CLI commands to be read from /home/ubuntu/flow-create-commands.txt
104   Configuring Port 0 (socket 0)
105   Port 0: 7C:FE:90:CB:74:CE
106   Configuring Port 1 (socket 0)
107   Port 1: 7C:FE:90:CB:74:CA
108   Checking link statuses...
109   Port 0 Link Up - speed 10000 Mbps - full-duplex
110   Port 1 Link Up - speed 10000 Mbps - full-duplex
111   Done
112   Flow rule #0 created
113   Flow rule #1 created
114   ...
115   ...
116   Flow rule #498 created
117   Flow rule #499 created
118   Read all CLI commands from /home/ubuntu/flow-create-commands.txt
119   testpmd>
120
121
122* At run-time additional commands can be loaded in bulk by invoking the ``load FILENAME``
123  command.
124
125.. code-block:: console
126
127   testpmd> load /home/ubuntu/flow-create-commands.txt
128   Flow rule #0 created
129   Flow rule #1 created
130   ...
131   ...
132   Flow rule #498 created
133   Flow rule #499 created
134   Read all CLI commands from /home/ubuntu/flow-create-commands.txt
135   testpmd>
136
137
138In all cases output from any included command will be displayed as standard output.
139Execution will continue until the end of the file is reached regardless of
140whether any errors occur.  The end user must examine the output to determine if
141any failures occurred.
142
143
144Control Functions
145-----------------
146
147start
148~~~~~
149
150Start packet forwarding with current configuration::
151
152   testpmd> start
153
154start tx_first
155~~~~~~~~~~~~~~
156
157Start packet forwarding with current configuration after sending specified number of bursts of packets::
158
159   testpmd> start tx_first (""|burst_num)
160
161The default burst number is 1 when ``burst_num`` not presented.
162
163stop
164~~~~
165
166Stop packet forwarding, and display accumulated statistics::
167
168   testpmd> stop
169
170quit
171~~~~
172
173Quit to prompt::
174
175   testpmd> quit
176
177
178Display Functions
179-----------------
180
181The functions in the following sections are used to display information about the
182testpmd configuration or the NIC status.
183
184show port
185~~~~~~~~~
186
187Display information for a given port or all ports::
188
189   testpmd> show port (info|stats|xstats|fdir|stat_qmap|dcb_tc|cap) (port_id|all)
190
191The available information categories are:
192
193* ``info``: General port information such as MAC address.
194
195* ``stats``: RX/TX statistics.
196
197* ``xstats``: RX/TX extended NIC statistics.
198
199* ``fdir``: Flow Director information and statistics.
200
201* ``stat_qmap``: Queue statistics mapping.
202
203* ``dcb_tc``: DCB information such as TC mapping.
204
205* ``cap``: Supported offload capabilities.
206
207For example:
208
209.. code-block:: console
210
211   testpmd> show port info 0
212
213   ********************* Infos for port 0 *********************
214
215   MAC address: XX:XX:XX:XX:XX:XX
216   Connect to socket: 0
217   memory allocation on the socket: 0
218   Link status: up
219   Link speed: 40000 Mbps
220   Link duplex: full-duplex
221   Promiscuous mode: enabled
222   Allmulticast mode: disabled
223   Maximum number of MAC addresses: 64
224   Maximum number of MAC addresses of hash filtering: 0
225   VLAN offload:
226       strip on
227       filter on
228       qinq(extend) off
229   Redirection table size: 512
230   Supported flow types:
231     ipv4-frag
232     ipv4-tcp
233     ipv4-udp
234     ipv4-sctp
235     ipv4-other
236     ipv6-frag
237     ipv6-tcp
238     ipv6-udp
239     ipv6-sctp
240     ipv6-other
241     l2_payload
242     port
243     vxlan
244     geneve
245     nvgre
246
247show port rss reta
248~~~~~~~~~~~~~~~~~~
249
250Display the rss redirection table entry indicated by masks on port X::
251
252   testpmd> show port (port_id) rss reta (size) (mask0, mask1...)
253
254size is used to indicate the hardware supported reta size
255
256show port rss-hash
257~~~~~~~~~~~~~~~~~~
258
259Display the RSS hash functions and RSS hash key of a port::
260
261   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]
262
263clear port
264~~~~~~~~~~
265
266Clear the port statistics for a given port or for all ports::
267
268   testpmd> clear port (info|stats|xstats|fdir|stat_qmap) (port_id|all)
269
270For example::
271
272   testpmd> clear port stats all
273
274show (rxq|txq)
275~~~~~~~~~~~~~~
276
277Display information for a given port's RX/TX queue::
278
279   testpmd> show (rxq|txq) info (port_id) (queue_id)
280
281show config
282~~~~~~~~~~~
283
284Displays the configuration of the application.
285The configuration comes from the command-line, the runtime or the application defaults::
286
287   testpmd> show config (rxtx|cores|fwd|txpkts)
288
289The available information categories are:
290
291* ``rxtx``: RX/TX configuration items.
292
293* ``cores``: List of forwarding cores.
294
295* ``fwd``: Packet forwarding configuration.
296
297* ``txpkts``: Packets to TX configuration.
298
299For example:
300
301.. code-block:: console
302
303   testpmd> show config rxtx
304
305   io packet forwarding - CRC stripping disabled - packets/burst=16
306   nb forwarding cores=2 - nb forwarding ports=1
307   RX queues=1 - RX desc=128 - RX free threshold=0
308   RX threshold registers: pthresh=8 hthresh=8 wthresh=4
309   TX queues=1 - TX desc=512 - TX free threshold=0
310   TX threshold registers: pthresh=36 hthresh=0 wthresh=0
311   TX RS bit threshold=0 - TXQ flags=0x0
312
313set fwd
314~~~~~~~
315
316Set the packet forwarding mode::
317
318   testpmd> set fwd (io|mac|macswap|flowgen| \
319                     rxonly|txonly|csum|icmpecho) (""|retry)
320
321``retry`` can be specified for forwarding engines except ``rx_only``.
322
323The available information categories are:
324
325* ``io``: Forwards packets "as-is" in I/O mode.
326  This is the fastest possible forwarding operation as it does not access packets data.
327  This is the default mode.
328
329* ``mac``: Changes the source and the destination Ethernet addresses of packets before forwarding them.
330  Default application behaviour is to set source Ethernet address to that of the transmitting interface, and destination
331  address to a dummy value (set during init). The user may specify a target destination Ethernet address via the 'eth-peer' or
332  'eth-peer-configfile' command-line options. It is not currently possible to specify a specific source Ethernet address.
333
334* ``macswap``: MAC swap forwarding mode.
335  Swaps the source and the destination Ethernet addresses of packets before forwarding them.
336
337* ``flowgen``: Multi-flow generation mode.
338  Originates a number of flows (with varying destination IP addresses), and terminate receive traffic.
339
340* ``rxonly``: Receives packets but doesn't transmit them.
341
342* ``txonly``: Generates and transmits packets without receiving any.
343
344* ``csum``: Changes the checksum field with hardware or software methods depending on the offload flags on the packet.
345
346* ``icmpecho``: Receives a burst of packets, lookup for IMCP echo requests and, if any, send back ICMP echo replies.
347
348* ``ieee1588``: Demonstrate L2 IEEE1588 V2 PTP timestamping for RX and TX. Requires ``CONFIG_RTE_LIBRTE_IEEE1588=y``.
349
350* ``tm``: Traffic Management forwarding mode
351  Demonstrates the use of ethdev traffic management APIs and softnic PMD for
352  QoS traffic management. In this mode, 5-level hierarchical QoS scheduler is
353  available as an default option that can be enabled through CLI. The user can
354  also modify the default hierarchy or specify the new hierarchy through CLI for
355  implementing QoS scheduler.  Requires ``CONFIG_RTE_LIBRTE_PMD_SOFTNIC=y`` ``CONFIG_RTE_LIBRTE_SCHED=y``.
356
357Example::
358
359   testpmd> set fwd rxonly
360
361   Set rxonly packet forwarding mode
362
363
364read rxd
365~~~~~~~~
366
367Display an RX descriptor for a port RX queue::
368
369   testpmd> read rxd (port_id) (queue_id) (rxd_id)
370
371For example::
372
373   testpmd> read rxd 0 0 4
374        0x0000000B - 0x001D0180 / 0x0000000B - 0x001D0180
375
376read txd
377~~~~~~~~
378
379Display a TX descriptor for a port TX queue::
380
381   testpmd> read txd (port_id) (queue_id) (txd_id)
382
383For example::
384
385   testpmd> read txd 0 0 4
386        0x00000001 - 0x24C3C440 / 0x000F0000 - 0x2330003C
387
388ddp get list
389~~~~~~~~~~~~
390
391Get loaded dynamic device personalization (DDP) package info list::
392
393   testpmd> ddp get list (port_id)
394
395ddp get info
396~~~~~~~~~~~~
397
398Display information about dynamic device personalization (DDP) profile::
399
400   testpmd> ddp get info (profile_path)
401
402show vf stats
403~~~~~~~~~~~~~
404
405Display VF statistics::
406
407   testpmd> show vf stats (port_id) (vf_id)
408
409clear vf stats
410~~~~~~~~~~~~~~
411
412Reset VF statistics::
413
414   testpmd> clear vf stats (port_id) (vf_id)
415
416show port pctype mapping
417~~~~~~~~~~~~~~~~~~~~~~~~
418
419List all items from the pctype mapping table::
420
421   testpmd> show port (port_id) pctype mapping
422
423
424Configuration Functions
425-----------------------
426
427The testpmd application can be configured from the runtime as well as from the command-line.
428
429This section details the available configuration functions that are available.
430
431.. note::
432
433   Configuration changes only become active when forwarding is started/restarted.
434
435set default
436~~~~~~~~~~~
437
438Reset forwarding to the default configuration::
439
440   testpmd> set default
441
442set verbose
443~~~~~~~~~~~
444
445Set the debug verbosity level::
446
447   testpmd> set verbose (level)
448
449Currently the only available levels are 0 (silent except for error) and 1 (fully verbose).
450
451set log
452~~~~~~~
453
454Set the log level for a log type::
455
456	testpmd> set log global|(type) (level)
457
458Where:
459
460* ``type`` is the log name.
461
462* ``level`` is the log level.
463
464For example, to change the global log level::
465	testpmd> set log global (level)
466
467Regexes can also be used for type. To change log level of user1, user2 and user3::
468	testpmd> set log user[1-3] (level)
469
470set nbport
471~~~~~~~~~~
472
473Set the number of ports used by the application:
474
475set nbport (num)
476
477This is equivalent to the ``--nb-ports`` command-line option.
478
479set nbcore
480~~~~~~~~~~
481
482Set the number of cores used by the application::
483
484   testpmd> set nbcore (num)
485
486This is equivalent to the ``--nb-cores`` command-line option.
487
488.. note::
489
490   The number of cores used must not be greater than number of ports used multiplied by the number of queues per port.
491
492set coremask
493~~~~~~~~~~~~
494
495Set the forwarding cores hexadecimal mask::
496
497   testpmd> set coremask (mask)
498
499This is equivalent to the ``--coremask`` command-line option.
500
501.. note::
502
503   The master lcore is reserved for command line parsing only and cannot be masked on for packet forwarding.
504
505set portmask
506~~~~~~~~~~~~
507
508Set the forwarding ports hexadecimal mask::
509
510   testpmd> set portmask (mask)
511
512This is equivalent to the ``--portmask`` command-line option.
513
514set burst
515~~~~~~~~~
516
517Set number of packets per burst::
518
519   testpmd> set burst (num)
520
521This is equivalent to the ``--burst command-line`` option.
522
523When retry is enabled, the transmit delay time and number of retries can also be set::
524
525   testpmd> set burst tx delay (microseconds) retry (num)
526
527set txpkts
528~~~~~~~~~~
529
530Set the length of each segment of the TX-ONLY packets or length of packet for FLOWGEN mode::
531
532   testpmd> set txpkts (x[,y]*)
533
534Where x[,y]* represents a CSV list of values, without white space.
535
536set txsplit
537~~~~~~~~~~~
538
539Set the split policy for the TX packets, applicable for TX-ONLY and CSUM forwarding modes::
540
541   testpmd> set txsplit (off|on|rand)
542
543Where:
544
545* ``off`` disable packet copy & split for CSUM mode.
546
547* ``on`` split outgoing packet into multiple segments. Size of each segment
548  and number of segments per packet is determined by ``set txpkts`` command
549  (see above).
550
551* ``rand`` same as 'on', but number of segments per each packet is a random value between 1 and total number of segments.
552
553set corelist
554~~~~~~~~~~~~
555
556Set the list of forwarding cores::
557
558   testpmd> set corelist (x[,y]*)
559
560For example, to change the forwarding cores:
561
562.. code-block:: console
563
564   testpmd> set corelist 3,1
565   testpmd> show config fwd
566
567   io packet forwarding - ports=2 - cores=2 - streams=2 - NUMA support disabled
568   Logical Core 3 (socket 0) forwards packets on 1 streams:
569   RX P=0/Q=0 (socket 0) -> TX P=1/Q=0 (socket 0) peer=02:00:00:00:00:01
570   Logical Core 1 (socket 0) forwards packets on 1 streams:
571   RX P=1/Q=0 (socket 0) -> TX P=0/Q=0 (socket 0) peer=02:00:00:00:00:00
572
573.. note::
574
575   The cores are used in the same order as specified on the command line.
576
577set portlist
578~~~~~~~~~~~~
579
580Set the list of forwarding ports::
581
582   testpmd> set portlist (x[,y]*)
583
584For example, to change the port forwarding:
585
586.. code-block:: console
587
588   testpmd> set portlist 0,2,1,3
589   testpmd> show config fwd
590
591   io packet forwarding - ports=4 - cores=1 - streams=4
592   Logical Core 3 (socket 0) forwards packets on 4 streams:
593   RX P=0/Q=0 (socket 0) -> TX P=2/Q=0 (socket 0) peer=02:00:00:00:00:01
594   RX P=2/Q=0 (socket 0) -> TX P=0/Q=0 (socket 0) peer=02:00:00:00:00:00
595   RX P=1/Q=0 (socket 0) -> TX P=3/Q=0 (socket 0) peer=02:00:00:00:00:03
596   RX P=3/Q=0 (socket 0) -> TX P=1/Q=0 (socket 0) peer=02:00:00:00:00:02
597
598set tx loopback
599~~~~~~~~~~~~~~~
600
601Enable/disable tx loopback::
602
603   testpmd> set tx loopback (port_id) (on|off)
604
605set drop enable
606~~~~~~~~~~~~~~~
607
608set drop enable bit for all queues::
609
610   testpmd> set all queues drop (port_id) (on|off)
611
612set split drop enable (for VF)
613~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
614
615set split drop enable bit for VF from PF::
616
617   testpmd> set vf split drop (port_id) (vf_id) (on|off)
618
619set mac antispoof (for VF)
620~~~~~~~~~~~~~~~~~~~~~~~~~~
621
622Set mac antispoof for a VF from the PF::
623
624   testpmd> set vf mac antispoof  (port_id) (vf_id) (on|off)
625
626set macsec offload
627~~~~~~~~~~~~~~~~~~
628
629Enable/disable MACsec offload::
630
631   testpmd> set macsec offload (port_id) on encrypt (on|off) replay-protect (on|off)
632   testpmd> set macsec offload (port_id) off
633
634set macsec sc
635~~~~~~~~~~~~~
636
637Configure MACsec secure connection (SC)::
638
639   testpmd> set macsec sc (tx|rx) (port_id) (mac) (pi)
640
641.. note::
642
643   The pi argument is ignored for tx.
644   Check the NIC Datasheet for hardware limits.
645
646set macsec sa
647~~~~~~~~~~~~~
648
649Configure MACsec secure association (SA)::
650
651   testpmd> set macsec sa (tx|rx) (port_id) (idx) (an) (pn) (key)
652
653.. note::
654
655   The IDX value must be 0 or 1.
656   Check the NIC Datasheet for hardware limits.
657
658set broadcast mode (for VF)
659~~~~~~~~~~~~~~~~~~~~~~~~~~~
660
661Set broadcast mode for a VF from the PF::
662
663   testpmd> set vf broadcast (port_id) (vf_id) (on|off)
664
665vlan set strip
666~~~~~~~~~~~~~~
667
668Set the VLAN strip on a port::
669
670   testpmd> vlan set strip (on|off) (port_id)
671
672vlan set stripq
673~~~~~~~~~~~~~~~
674
675Set the VLAN strip for a queue on a port::
676
677   testpmd> vlan set stripq (on|off) (port_id,queue_id)
678
679vlan set stripq (for VF)
680~~~~~~~~~~~~~~~~~~~~~~~~
681
682Set VLAN strip for all queues in a pool for a VF from the PF::
683
684   testpmd> set vf vlan stripq (port_id) (vf_id) (on|off)
685
686vlan set insert (for VF)
687~~~~~~~~~~~~~~~~~~~~~~~~
688
689Set VLAN insert for a VF from the PF::
690
691   testpmd> set vf vlan insert (port_id) (vf_id) (vlan_id)
692
693vlan set tag (for VF)
694~~~~~~~~~~~~~~~~~~~~~
695
696Set VLAN tag for a VF from the PF::
697
698   testpmd> set vf vlan tag (port_id) (vf_id) (on|off)
699
700vlan set antispoof (for VF)
701~~~~~~~~~~~~~~~~~~~~~~~~~~~
702
703Set VLAN antispoof for a VF from the PF::
704
705   testpmd> set vf vlan antispoof (port_id) (vf_id) (on|off)
706
707vlan set filter
708~~~~~~~~~~~~~~~
709
710Set the VLAN filter on a port::
711
712   testpmd> vlan set filter (on|off) (port_id)
713
714vlan set qinq
715~~~~~~~~~~~~~
716
717Set the VLAN QinQ (extended queue in queue) on for a port::
718
719   testpmd> vlan set qinq (on|off) (port_id)
720
721vlan set tpid
722~~~~~~~~~~~~~
723
724Set the inner or outer VLAN TPID for packet filtering on a port::
725
726   testpmd> vlan set (inner|outer) tpid (value) (port_id)
727
728.. note::
729
730   TPID value must be a 16-bit number (value <= 65536).
731
732rx_vlan add
733~~~~~~~~~~~
734
735Add a VLAN ID, or all identifiers, to the set of VLAN identifiers filtered by port ID::
736
737   testpmd> rx_vlan add (vlan_id|all) (port_id)
738
739.. note::
740
741   VLAN filter must be set on that port. VLAN ID < 4096.
742   Depending on the NIC used, number of vlan_ids may be limited to the maximum entries
743   in VFTA table. This is important if enabling all vlan_ids.
744
745rx_vlan rm
746~~~~~~~~~~
747
748Remove a VLAN ID, or all identifiers, from the set of VLAN identifiers filtered by port ID::
749
750   testpmd> rx_vlan rm (vlan_id|all) (port_id)
751
752rx_vlan add (for VF)
753~~~~~~~~~~~~~~~~~~~~
754
755Add a VLAN ID, to the set of VLAN identifiers filtered for VF(s) for port ID::
756
757   testpmd> rx_vlan add (vlan_id) port (port_id) vf (vf_mask)
758
759rx_vlan rm (for VF)
760~~~~~~~~~~~~~~~~~~~
761
762Remove a VLAN ID, from the set of VLAN identifiers filtered for VF(s) for port ID::
763
764   testpmd> rx_vlan rm (vlan_id) port (port_id) vf (vf_mask)
765
766tunnel_filter add
767~~~~~~~~~~~~~~~~~
768
769Add a tunnel filter on a port::
770
771   testpmd> tunnel_filter add (port_id) (outer_mac) (inner_mac) (ip_addr) \
772            (inner_vlan) (vxlan|nvgre|ipingre) (imac-ivlan|imac-ivlan-tenid|\
773            imac-tenid|imac|omac-imac-tenid|oip|iip) (tenant_id) (queue_id)
774
775The available information categories are:
776
777* ``vxlan``: Set tunnel type as VXLAN.
778
779* ``nvgre``: Set tunnel type as NVGRE.
780
781* ``ipingre``: Set tunnel type as IP-in-GRE.
782
783* ``imac-ivlan``: Set filter type as Inner MAC and VLAN.
784
785* ``imac-ivlan-tenid``: Set filter type as Inner MAC, VLAN and tenant ID.
786
787* ``imac-tenid``: Set filter type as Inner MAC and tenant ID.
788
789* ``imac``: Set filter type as Inner MAC.
790
791* ``omac-imac-tenid``: Set filter type as Outer MAC, Inner MAC and tenant ID.
792
793* ``oip``: Set filter type as Outer IP.
794
795* ``iip``: Set filter type as Inner IP.
796
797Example::
798
799   testpmd> tunnel_filter add 0 68:05:CA:28:09:82 00:00:00:00:00:00 \
800            192.168.2.2 0 ipingre oip 1 1
801
802   Set an IP-in-GRE tunnel on port 0, and the filter type is Outer IP.
803
804tunnel_filter remove
805~~~~~~~~~~~~~~~~~~~~
806
807Remove a tunnel filter on a port::
808
809   testpmd> tunnel_filter rm (port_id) (outer_mac) (inner_mac) (ip_addr) \
810            (inner_vlan) (vxlan|nvgre|ipingre) (imac-ivlan|imac-ivlan-tenid|\
811            imac-tenid|imac|omac-imac-tenid|oip|iip) (tenant_id) (queue_id)
812
813rx_vxlan_port add
814~~~~~~~~~~~~~~~~~
815
816Add an UDP port for VXLAN packet filter on a port::
817
818   testpmd> rx_vxlan_port add (udp_port) (port_id)
819
820rx_vxlan_port remove
821~~~~~~~~~~~~~~~~~~~~
822
823Remove an UDP port for VXLAN packet filter on a port::
824
825   testpmd> rx_vxlan_port rm (udp_port) (port_id)
826
827tx_vlan set
828~~~~~~~~~~~
829
830Set hardware insertion of VLAN IDs in packets sent on a port::
831
832   testpmd> tx_vlan set (port_id) vlan_id[, vlan_id_outer]
833
834For example, set a single VLAN ID (5) insertion on port 0::
835
836   tx_vlan set 0 5
837
838Or, set double VLAN ID (inner: 2, outer: 3) insertion on port 1::
839
840   tx_vlan set 1 2 3
841
842
843tx_vlan set pvid
844~~~~~~~~~~~~~~~~
845
846Set port based hardware insertion of VLAN ID in packets sent on a port::
847
848   testpmd> tx_vlan set pvid (port_id) (vlan_id) (on|off)
849
850tx_vlan reset
851~~~~~~~~~~~~~
852
853Disable hardware insertion of a VLAN header in packets sent on a port::
854
855   testpmd> tx_vlan reset (port_id)
856
857csum set
858~~~~~~~~
859
860Select hardware or software calculation of the checksum when
861transmitting a packet using the ``csum`` forwarding engine::
862
863   testpmd> csum set (ip|udp|tcp|sctp|outer-ip) (hw|sw) (port_id)
864
865Where:
866
867* ``ip|udp|tcp|sctp`` always relate to  the inner layer.
868
869* ``outer-ip`` relates to the outer IP layer (only for IPv4) in the case where the packet is recognized
870  as a tunnel packet by the forwarding engine (vxlan, gre and ipip are
871  supported). See also the ``csum parse-tunnel`` command.
872
873.. note::
874
875   Check the NIC Datasheet for hardware limits.
876
877RSS queue region
878~~~~~~~~~~~~~~~~
879
880Set RSS queue region span on a port::
881
882   testpmd> set port (port_id) queue-region region_id (value) \
883		queue_start_index (value) queue_num (value)
884
885Set flowtype mapping on a RSS queue region on a port::
886
887   testpmd> set port (port_id) queue-region region_id (value) flowtype (value)
888
889where:
890
891* For the flowtype(pctype) of packet,the specific index for each type has
892  been defined in file i40e_type.h as enum i40e_filter_pctype.
893
894Set user priority mapping on a RSS queue region on a port::
895
896   testpmd> set port (port_id) queue-region UP (value) region_id (value)
897
898Flush all queue region related configuration on a port::
899
900   testpmd> set port (port_id) queue-region flush (on|off)
901
902where:
903
904* "on"is just an enable function which server for other configuration,
905  it is for all configuration about queue region from up layer,
906  at first will only keep in DPDK softwarestored in driver,
907  only after "flush on", it commit all configuration to HW.
908  "off" is just clean all configuration about queue region just now,
909  and restore all to DPDK i40e driver default config when start up.
910
911Show all queue region related configuration info on a port::
912
913   testpmd> show port (port_id) queue-region
914
915.. note::
916
917  Queue region only support on PF by now, so these command is
918  only for configuration of queue region on PF port.
919
920csum parse-tunnel
921~~~~~~~~~~~~~~~~~
922
923Define how tunneled packets should be handled by the csum forward
924engine::
925
926   testpmd> csum parse-tunnel (on|off) (tx_port_id)
927
928If enabled, the csum forward engine will try to recognize supported
929tunnel headers (vxlan, gre, ipip).
930
931If disabled, treat tunnel packets as non-tunneled packets (a inner
932header is handled as a packet payload).
933
934.. note::
935
936   The port argument is the TX port like in the ``csum set`` command.
937
938Example:
939
940Consider a packet in packet like the following::
941
942   eth_out/ipv4_out/udp_out/vxlan/eth_in/ipv4_in/tcp_in
943
944* If parse-tunnel is enabled, the ``ip|udp|tcp|sctp`` parameters of ``csum set``
945  command relate to the inner headers (here ``ipv4_in`` and ``tcp_in``), and the
946  ``outer-ip parameter`` relates to the outer headers (here ``ipv4_out``).
947
948* If parse-tunnel is disabled, the ``ip|udp|tcp|sctp`` parameters of ``csum  set``
949   command relate to the outer headers, here ``ipv4_out`` and ``udp_out``.
950
951csum show
952~~~~~~~~~
953
954Display tx checksum offload configuration::
955
956   testpmd> csum show (port_id)
957
958tso set
959~~~~~~~
960
961Enable TCP Segmentation Offload (TSO) in the ``csum`` forwarding engine::
962
963   testpmd> tso set (segsize) (port_id)
964
965.. note::
966
967   Check the NIC datasheet for hardware limits.
968
969tso show
970~~~~~~~~
971
972Display the status of TCP Segmentation Offload::
973
974   testpmd> tso show (port_id)
975
976set port - gro
977~~~~~~~~~~~~~~
978
979Enable or disable GRO in ``csum`` forwarding engine::
980
981   testpmd> set port <port_id> gro on|off
982
983If enabled, the csum forwarding engine will perform GRO on the TCP/IPv4
984packets received from the given port.
985
986If disabled, packets received from the given port won't be performed
987GRO. By default, GRO is disabled for all ports.
988
989.. note::
990
991   When enable GRO for a port, TCP/IPv4 packets received from the port
992   will be performed GRO. After GRO, all merged packets have bad
993   checksums, since the GRO library doesn't re-calculate checksums for
994   the merged packets. Therefore, if users want the merged packets to
995   have correct checksums, please select HW IP checksum calculation and
996   HW TCP checksum calculation for the port which the merged packets are
997   transmitted to.
998
999show port - gro
1000~~~~~~~~~~~~~~~
1001
1002Display GRO configuration for a given port::
1003
1004   testpmd> show port <port_id> gro
1005
1006set gro flush
1007~~~~~~~~~~~~~
1008
1009Set the cycle to flush the GROed packets from reassembly tables::
1010
1011   testpmd> set gro flush <cycles>
1012
1013When enable GRO, the csum forwarding engine performs GRO on received
1014packets, and the GROed packets are stored in reassembly tables. Users
1015can use this command to determine when the GROed packets are flushed
1016from the reassembly tables.
1017
1018The ``cycles`` is measured in GRO operation times. The csum forwarding
1019engine flushes the GROed packets from the tables every ``cycles`` GRO
1020operations.
1021
1022By default, the value of ``cycles`` is 1, which means flush GROed packets
1023from the reassembly tables as soon as one GRO operation finishes. The value
1024of ``cycles`` should be in the range of 1 to ``GRO_MAX_FLUSH_CYCLES``.
1025
1026Please note that the large value of ``cycles`` may cause the poor TCP/IP
1027stack performance. Because the GROed packets are delayed to arrive the
1028stack, thus causing more duplicated ACKs and TCP retransmissions.
1029
1030set port - gso
1031~~~~~~~~~~~~~~
1032
1033Toggle per-port GSO support in ``csum`` forwarding engine::
1034
1035   testpmd> set port <port_id> gso on|off
1036
1037If enabled, the csum forwarding engine will perform GSO on supported IPv4
1038packets, transmitted on the given port.
1039
1040If disabled, packets transmitted on the given port will not undergo GSO.
1041By default, GSO is disabled for all ports.
1042
1043.. note::
1044
1045   When GSO is enabled on a port, supported IPv4 packets transmitted on that
1046   port undergo GSO. Afterwards, the segmented packets are represented by
1047   multi-segment mbufs; however, the csum forwarding engine doesn't calculation
1048   of checksums for GSO'd segments in SW. As a result, if users want correct
1049   checksums in GSO segments, they should enable HW checksum calculation for
1050   GSO-enabled ports.
1051
1052   For example, HW checksum calculation for VxLAN GSO'd packets may be enabled
1053   by setting the following options in the csum forwarding engine:
1054
1055   testpmd> csum set outer_ip hw <port_id>
1056
1057   testpmd> csum set ip hw <port_id>
1058
1059   testpmd> csum set tcp hw <port_id>
1060
1061set gso segsz
1062~~~~~~~~~~~~~
1063
1064Set the maximum GSO segment size (measured in bytes), which includes the
1065packet header and the packet payload for GSO-enabled ports (global)::
1066
1067   testpmd> set gso segsz <length>
1068
1069show port - gso
1070~~~~~~~~~~~~~~~
1071
1072Display the status of Generic Segmentation Offload for a given port::
1073
1074   testpmd> show port <port_id> gso
1075
1076mac_addr add
1077~~~~~~~~~~~~
1078
1079Add an alternative MAC address to a port::
1080
1081   testpmd> mac_addr add (port_id) (XX:XX:XX:XX:XX:XX)
1082
1083mac_addr remove
1084~~~~~~~~~~~~~~~
1085
1086Remove a MAC address from a port::
1087
1088   testpmd> mac_addr remove (port_id) (XX:XX:XX:XX:XX:XX)
1089
1090mac_addr add (for VF)
1091~~~~~~~~~~~~~~~~~~~~~
1092
1093Add an alternative MAC address for a VF to a port::
1094
1095   testpmd> mac_add add port (port_id) vf (vf_id) (XX:XX:XX:XX:XX:XX)
1096
1097mac_addr set
1098~~~~~~~~~~~~
1099
1100Set the default MAC address for a port::
1101
1102   testpmd> mac_addr set (port_id) (XX:XX:XX:XX:XX:XX)
1103
1104mac_addr set (for VF)
1105~~~~~~~~~~~~~~~~~~~~~
1106
1107Set the MAC address for a VF from the PF::
1108
1109   testpmd> set vf mac addr (port_id) (vf_id) (XX:XX:XX:XX:XX:XX)
1110
1111set eth-peer
1112~~~~~~~~~~~~
1113
1114Set the forwarding peer address for certain port::
1115
1116   testpmd> set eth-peer (port_id) (perr_addr)
1117
1118This is equivalent to the ``--eth-peer`` command-line option.
1119
1120set port-uta
1121~~~~~~~~~~~~
1122
1123Set the unicast hash filter(s) on/off for a port::
1124
1125   testpmd> set port (port_id) uta (XX:XX:XX:XX:XX:XX|all) (on|off)
1126
1127set promisc
1128~~~~~~~~~~~
1129
1130Set the promiscuous mode on for a port or for all ports.
1131In promiscuous mode packets are not dropped if they aren't for the specified MAC address::
1132
1133   testpmd> set promisc (port_id|all) (on|off)
1134
1135set allmulti
1136~~~~~~~~~~~~
1137
1138Set the allmulti mode for a port or for all ports::
1139
1140   testpmd> set allmulti (port_id|all) (on|off)
1141
1142Same as the ifconfig (8) option. Controls how multicast packets are handled.
1143
1144set promisc (for VF)
1145~~~~~~~~~~~~~~~~~~~~
1146
1147Set the unicast promiscuous mode for a VF from PF.
1148It's supported by Intel i40e NICs now.
1149In promiscuous mode packets are not dropped if they aren't for the specified MAC address::
1150
1151   testpmd> set vf promisc (port_id) (vf_id) (on|off)
1152
1153set allmulticast (for VF)
1154~~~~~~~~~~~~~~~~~~~~~~~~~
1155
1156Set the multicast promiscuous mode for a VF from PF.
1157It's supported by Intel i40e NICs now.
1158In promiscuous mode packets are not dropped if they aren't for the specified MAC address::
1159
1160   testpmd> set vf allmulti (port_id) (vf_id) (on|off)
1161
1162set tx max bandwidth (for VF)
1163~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
1164
1165Set TX max absolute bandwidth (Mbps) for a VF from PF::
1166
1167   testpmd> set vf tx max-bandwidth (port_id) (vf_id) (max_bandwidth)
1168
1169set tc tx min bandwidth (for VF)
1170~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
1171
1172Set all TCs' TX min relative bandwidth (%) for a VF from PF::
1173
1174   testpmd> set vf tc tx min-bandwidth (port_id) (vf_id) (bw1, bw2, ...)
1175
1176set tc tx max bandwidth (for VF)
1177~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
1178
1179Set a TC's TX max absolute bandwidth (Mbps) for a VF from PF::
1180
1181   testpmd> set vf tc tx max-bandwidth (port_id) (vf_id) (tc_no) (max_bandwidth)
1182
1183set tc strict link priority mode
1184~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
1185
1186Set some TCs' strict link priority mode on a physical port::
1187
1188   testpmd> set tx strict-link-priority (port_id) (tc_bitmap)
1189
1190set tc tx min bandwidth
1191~~~~~~~~~~~~~~~~~~~~~~~
1192
1193Set all TCs' TX min relative bandwidth (%) globally for all PF and VFs::
1194
1195   testpmd> set tc tx min-bandwidth (port_id) (bw1, bw2, ...)
1196
1197set flow_ctrl rx
1198~~~~~~~~~~~~~~~~
1199
1200Set the link flow control parameter on a port::
1201
1202   testpmd> set flow_ctrl rx (on|off) tx (on|off) (high_water) (low_water) \
1203            (pause_time) (send_xon) mac_ctrl_frame_fwd (on|off) \
1204	    autoneg (on|off) (port_id)
1205
1206Where:
1207
1208* ``high_water`` (integer): High threshold value to trigger XOFF.
1209
1210* ``low_water`` (integer): Low threshold value to trigger XON.
1211
1212* ``pause_time`` (integer): Pause quota in the Pause frame.
1213
1214* ``send_xon`` (0/1): Send XON frame.
1215
1216* ``mac_ctrl_frame_fwd``: Enable receiving MAC control frames.
1217
1218* ``autoneg``: Change the auto-negotiation parameter.
1219
1220set pfc_ctrl rx
1221~~~~~~~~~~~~~~~
1222
1223Set the priority flow control parameter on a port::
1224
1225   testpmd> set pfc_ctrl rx (on|off) tx (on|off) (high_water) (low_water) \
1226            (pause_time) (priority) (port_id)
1227
1228Where:
1229
1230* ``high_water`` (integer): High threshold value.
1231
1232* ``low_water`` (integer): Low threshold value.
1233
1234* ``pause_time`` (integer): Pause quota in the Pause frame.
1235
1236* ``priority`` (0-7): VLAN User Priority.
1237
1238set stat_qmap
1239~~~~~~~~~~~~~
1240
1241Set statistics mapping (qmapping 0..15) for RX/TX queue on port::
1242
1243   testpmd> set stat_qmap (tx|rx) (port_id) (queue_id) (qmapping)
1244
1245For example, to set rx queue 2 on port 0 to mapping 5::
1246
1247   testpmd>set stat_qmap rx 0 2 5
1248
1249set xstats-hide-zero
1250~~~~~~~~~~~~~~~~~~~~
1251
1252Set the option to hide zero values for xstats display::
1253
1254	testpmd> set xstats-hide-zero on|off
1255
1256.. note::
1257
1258	By default, the zero values are displayed for xstats.
1259
1260set port - rx/tx (for VF)
1261~~~~~~~~~~~~~~~~~~~~~~~~~
1262
1263Set VF receive/transmit from a port::
1264
1265   testpmd> set port (port_id) vf (vf_id) (rx|tx) (on|off)
1266
1267set port - mac address filter (for VF)
1268~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
1269
1270Add/Remove unicast or multicast MAC addr filter for a VF::
1271
1272   testpmd> set port (port_id) vf (vf_id) (mac_addr) \
1273            (exact-mac|exact-mac-vlan|hashmac|hashmac-vlan) (on|off)
1274
1275set port - rx mode(for VF)
1276~~~~~~~~~~~~~~~~~~~~~~~~~~
1277
1278Set the VF receive mode of a port::
1279
1280   testpmd> set port (port_id) vf (vf_id) \
1281            rxmode (AUPE|ROPE|BAM|MPE) (on|off)
1282
1283The available receive modes are:
1284
1285* ``AUPE``: Accepts untagged VLAN.
1286
1287* ``ROPE``: Accepts unicast hash.
1288
1289* ``BAM``: Accepts broadcast packets.
1290
1291* ``MPE``: Accepts all multicast packets.
1292
1293set port - tx_rate (for Queue)
1294~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
1295
1296Set TX rate limitation for a queue on a port::
1297
1298   testpmd> set port (port_id) queue (queue_id) rate (rate_value)
1299
1300set port - tx_rate (for VF)
1301~~~~~~~~~~~~~~~~~~~~~~~~~~~
1302
1303Set TX rate limitation for queues in VF on a port::
1304
1305   testpmd> set port (port_id) vf (vf_id) rate (rate_value) queue_mask (queue_mask)
1306
1307set port - mirror rule
1308~~~~~~~~~~~~~~~~~~~~~~
1309
1310Set pool or vlan type mirror rule for a port::
1311
1312   testpmd> set port (port_id) mirror-rule (rule_id) \
1313            (pool-mirror-up|pool-mirror-down|vlan-mirror) \
1314            (poolmask|vlanid[,vlanid]*) dst-pool (pool_id) (on|off)
1315
1316Set link mirror rule for a port::
1317
1318   testpmd> set port (port_id) mirror-rule (rule_id) \
1319           (uplink-mirror|downlink-mirror) dst-pool (pool_id) (on|off)
1320
1321For example to enable mirror traffic with vlan 0,1 to pool 0::
1322
1323   set port 0 mirror-rule 0 vlan-mirror 0,1 dst-pool 0 on
1324
1325reset port - mirror rule
1326~~~~~~~~~~~~~~~~~~~~~~~~
1327
1328Reset a mirror rule for a port::
1329
1330   testpmd> reset port (port_id) mirror-rule (rule_id)
1331
1332set flush_rx
1333~~~~~~~~~~~~
1334
1335Set the flush on RX streams before forwarding.
1336The default is flush ``on``.
1337Mainly used with PCAP drivers to turn off the default behavior of flushing the first 512 packets on RX streams::
1338
1339   testpmd> set flush_rx off
1340
1341set bypass mode
1342~~~~~~~~~~~~~~~
1343
1344Set the bypass mode for the lowest port on bypass enabled NIC::
1345
1346   testpmd> set bypass mode (normal|bypass|isolate) (port_id)
1347
1348set bypass event
1349~~~~~~~~~~~~~~~~
1350
1351Set the event required to initiate specified bypass mode for the lowest port on a bypass enabled::
1352
1353   testpmd> set bypass event (timeout|os_on|os_off|power_on|power_off) \
1354            mode (normal|bypass|isolate) (port_id)
1355
1356Where:
1357
1358* ``timeout``: Enable bypass after watchdog timeout.
1359
1360* ``os_on``: Enable bypass when OS/board is powered on.
1361
1362* ``os_off``: Enable bypass when OS/board is powered off.
1363
1364* ``power_on``: Enable bypass when power supply is turned on.
1365
1366* ``power_off``: Enable bypass when power supply is turned off.
1367
1368
1369set bypass timeout
1370~~~~~~~~~~~~~~~~~~
1371
1372Set the bypass watchdog timeout to ``n`` seconds where 0 = instant::
1373
1374   testpmd> set bypass timeout (0|1.5|2|3|4|8|16|32)
1375
1376show bypass config
1377~~~~~~~~~~~~~~~~~~
1378
1379Show the bypass configuration for a bypass enabled NIC using the lowest port on the NIC::
1380
1381   testpmd> show bypass config (port_id)
1382
1383set link up
1384~~~~~~~~~~~
1385
1386Set link up for a port::
1387
1388   testpmd> set link-up port (port id)
1389
1390set link down
1391~~~~~~~~~~~~~
1392
1393Set link down for a port::
1394
1395   testpmd> set link-down port (port id)
1396
1397E-tag set
1398~~~~~~~~~
1399
1400Enable E-tag insertion for a VF on a port::
1401
1402   testpmd> E-tag set insertion on port-tag-id (value) port (port_id) vf (vf_id)
1403
1404Disable E-tag insertion for a VF on a port::
1405
1406   testpmd> E-tag set insertion off port (port_id) vf (vf_id)
1407
1408Enable/disable E-tag stripping on a port::
1409
1410   testpmd> E-tag set stripping (on|off) port (port_id)
1411
1412Enable/disable E-tag based forwarding on a port::
1413
1414   testpmd> E-tag set forwarding (on|off) port (port_id)
1415
1416Add an E-tag forwarding filter on a port::
1417
1418   testpmd> E-tag set filter add e-tag-id (value) dst-pool (pool_id) port (port_id)
1419
1420Delete an E-tag forwarding filter on a port::
1421   testpmd> E-tag set filter del e-tag-id (value) port (port_id)
1422
1423ddp add
1424~~~~~~~
1425
1426Load a dynamic device personalization (DDP) package::
1427
1428   testpmd> ddp add (port_id) (package_path[,output_path])
1429
1430ddp del
1431~~~~~~~
1432
1433Delete a dynamic device personalization package::
1434
1435   testpmd> ddp del (port_id) (package_path)
1436
1437ptype mapping
1438~~~~~~~~~~~~~
1439
1440List all items from the ptype mapping table::
1441
1442   testpmd> ptype mapping get (port_id) (valid_only)
1443
1444Where:
1445
1446* ``valid_only``: A flag indicates if only list valid items(=1) or all itemss(=0).
1447
1448Replace a specific or a group of software defined ptype with a new one::
1449
1450   testpmd> ptype mapping replace  (port_id) (target) (mask) (pkt_type)
1451
1452where:
1453
1454* ``target``: A specific software ptype or a mask to represent a group of software ptypes.
1455
1456* ``mask``: A flag indicate if "target" is a specific software ptype(=0) or a ptype mask(=1).
1457
1458* ``pkt_type``: The new software ptype to replace the old ones.
1459
1460Update hardware defined ptype to software defined packet type mapping table::
1461
1462   testpmd> ptype mapping update (port_id) (hw_ptype) (sw_ptype)
1463
1464where:
1465
1466* ``hw_ptype``: hardware ptype as the index of the ptype mapping table.
1467
1468* ``sw_ptype``: software ptype as the value of the ptype mapping table.
1469
1470Reset ptype mapping table::
1471
1472   testpmd> ptype mapping reset (port_id)
1473
1474Port Functions
1475--------------
1476
1477The following sections show functions for configuring ports.
1478
1479.. note::
1480
1481   Port configuration changes only become active when forwarding is started/restarted.
1482
1483port attach
1484~~~~~~~~~~~
1485
1486Attach a port specified by pci address or virtual device args::
1487
1488   testpmd> port attach (identifier)
1489
1490To attach a new pci device, the device should be recognized by kernel first.
1491Then it should be moved under DPDK management.
1492Finally the port can be attached to testpmd.
1493
1494For example, to move a pci device using ixgbe under DPDK management:
1495
1496.. code-block:: console
1497
1498   # Check the status of the available devices.
1499   ./usertools/dpdk-devbind.py --status
1500
1501   Network devices using DPDK-compatible driver
1502   ============================================
1503   <none>
1504
1505   Network devices using kernel driver
1506   ===================================
1507   0000:0a:00.0 '82599ES 10-Gigabit' if=eth2 drv=ixgbe unused=
1508
1509
1510   # Bind the device to igb_uio.
1511   sudo ./usertools/dpdk-devbind.py -b igb_uio 0000:0a:00.0
1512
1513
1514   # Recheck the status of the devices.
1515   ./usertools/dpdk-devbind.py --status
1516   Network devices using DPDK-compatible driver
1517   ============================================
1518   0000:0a:00.0 '82599ES 10-Gigabit' drv=igb_uio unused=
1519
1520To attach a port created by virtual device, above steps are not needed.
1521
1522For example, to attach a port whose pci address is 0000:0a:00.0.
1523
1524.. code-block:: console
1525
1526   testpmd> port attach 0000:0a:00.0
1527   Attaching a new port...
1528   EAL: PCI device 0000:0a:00.0 on NUMA socket -1
1529   EAL:   probe driver: 8086:10fb rte_ixgbe_pmd
1530   EAL:   PCI memory mapped at 0x7f83bfa00000
1531   EAL:   PCI memory mapped at 0x7f83bfa80000
1532   PMD: eth_ixgbe_dev_init(): MAC: 2, PHY: 18, SFP+: 5
1533   PMD: eth_ixgbe_dev_init(): port 0 vendorID=0x8086 deviceID=0x10fb
1534   Port 0 is attached. Now total ports is 1
1535   Done
1536
1537For example, to attach a port created by pcap PMD.
1538
1539.. code-block:: console
1540
1541   testpmd> port attach net_pcap0
1542   Attaching a new port...
1543   PMD: Initializing pmd_pcap for net_pcap0
1544   PMD: Creating pcap-backed ethdev on numa socket 0
1545   Port 0 is attached. Now total ports is 1
1546   Done
1547
1548In this case, identifier is ``net_pcap0``.
1549This identifier format is the same as ``--vdev`` format of DPDK applications.
1550
1551For example, to re-attach a bonded port which has been previously detached,
1552the mode and slave parameters must be given.
1553
1554.. code-block:: console
1555
1556   testpmd> port attach net_bond_0,mode=0,slave=1
1557   Attaching a new port...
1558   EAL: Initializing pmd_bond for net_bond_0
1559   EAL: Create bonded device net_bond_0 on port 0 in mode 0 on socket 0.
1560   Port 0 is attached. Now total ports is 1
1561   Done
1562
1563
1564port detach
1565~~~~~~~~~~~
1566
1567Detach a specific port::
1568
1569   testpmd> port detach (port_id)
1570
1571Before detaching a port, the port should be stopped and closed.
1572
1573For example, to detach a pci device port 0.
1574
1575.. code-block:: console
1576
1577   testpmd> port stop 0
1578   Stopping ports...
1579   Done
1580   testpmd> port close 0
1581   Closing ports...
1582   Done
1583
1584   testpmd> port detach 0
1585   Detaching a port...
1586   EAL: PCI device 0000:0a:00.0 on NUMA socket -1
1587   EAL:   remove driver: 8086:10fb rte_ixgbe_pmd
1588   EAL:   PCI memory unmapped at 0x7f83bfa00000
1589   EAL:   PCI memory unmapped at 0x7f83bfa80000
1590   Done
1591
1592
1593For example, to detach a virtual device port 0.
1594
1595.. code-block:: console
1596
1597   testpmd> port stop 0
1598   Stopping ports...
1599   Done
1600   testpmd> port close 0
1601   Closing ports...
1602   Done
1603
1604   testpmd> port detach 0
1605   Detaching a port...
1606   PMD: Closing pcap ethdev on numa socket 0
1607   Port 'net_pcap0' is detached. Now total ports is 0
1608   Done
1609
1610To remove a pci device completely from the system, first detach the port from testpmd.
1611Then the device should be moved under kernel management.
1612Finally the device can be removed using kernel pci hotplug functionality.
1613
1614For example, to move a pci device under kernel management:
1615
1616.. code-block:: console
1617
1618   sudo ./usertools/dpdk-devbind.py -b ixgbe 0000:0a:00.0
1619
1620   ./usertools/dpdk-devbind.py --status
1621
1622   Network devices using DPDK-compatible driver
1623   ============================================
1624   <none>
1625
1626   Network devices using kernel driver
1627   ===================================
1628   0000:0a:00.0 '82599ES 10-Gigabit' if=eth2 drv=ixgbe unused=igb_uio
1629
1630To remove a port created by a virtual device, above steps are not needed.
1631
1632port start
1633~~~~~~~~~~
1634
1635Start all ports or a specific port::
1636
1637   testpmd> port start (port_id|all)
1638
1639port stop
1640~~~~~~~~~
1641
1642Stop all ports or a specific port::
1643
1644   testpmd> port stop (port_id|all)
1645
1646port close
1647~~~~~~~~~~
1648
1649Close all ports or a specific port::
1650
1651   testpmd> port close (port_id|all)
1652
1653port start/stop queue
1654~~~~~~~~~~~~~~~~~~~~~
1655
1656Start/stop a rx/tx queue on a specific port::
1657
1658   testpmd> port (port_id) (rxq|txq) (queue_id) (start|stop)
1659
1660Only take effect when port is started.
1661
1662port config - speed
1663~~~~~~~~~~~~~~~~~~~
1664
1665Set the speed and duplex mode for all ports or a specific port::
1666
1667   testpmd> port config (port_id|all) speed (10|100|1000|10000|25000|40000|50000|100000|auto) \
1668            duplex (half|full|auto)
1669
1670port config - queues/descriptors
1671~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
1672
1673Set number of queues/descriptors for rxq, txq, rxd and txd::
1674
1675   testpmd> port config all (rxq|txq|rxd|txd) (value)
1676
1677This is equivalent to the ``--rxq``, ``--txq``, ``--rxd`` and ``--txd`` command-line options.
1678
1679port config - max-pkt-len
1680~~~~~~~~~~~~~~~~~~~~~~~~~
1681
1682Set the maximum packet length::
1683
1684   testpmd> port config all max-pkt-len (value)
1685
1686This is equivalent to the ``--max-pkt-len`` command-line option.
1687
1688port config - CRC Strip
1689~~~~~~~~~~~~~~~~~~~~~~~
1690
1691Set hardware CRC stripping on or off for all ports::
1692
1693   testpmd> port config all crc-strip (on|off)
1694
1695CRC stripping is on by default.
1696
1697The ``off`` option is equivalent to the ``--disable-crc-strip`` command-line option.
1698
1699port config - scatter
1700~~~~~~~~~~~~~~~~~~~~~~~
1701
1702Set RX scatter mode on or off for all ports::
1703
1704   testpmd> port config all scatter (on|off)
1705
1706RX scatter mode is off by default.
1707
1708The ``on`` option is equivalent to the ``--enable-scatter`` command-line option.
1709
1710port config - RX Checksum
1711~~~~~~~~~~~~~~~~~~~~~~~~~
1712
1713Set hardware RX checksum offload to on or off for all ports::
1714
1715   testpmd> port config all rx-cksum (on|off)
1716
1717Checksum offload is off by default.
1718
1719The ``on`` option is equivalent to the ``--enable-rx-cksum`` command-line option.
1720
1721port config - VLAN
1722~~~~~~~~~~~~~~~~~~
1723
1724Set hardware VLAN on or off for all ports::
1725
1726   testpmd> port config all hw-vlan (on|off)
1727
1728Hardware VLAN is off by default.
1729
1730The ``on`` option is equivalent to the ``--enable-hw-vlan`` command-line option.
1731
1732port config - VLAN filter
1733~~~~~~~~~~~~~~~~~~~~~~~~~
1734
1735Set hardware VLAN filter on or off for all ports::
1736
1737   testpmd> port config all hw-vlan-filter (on|off)
1738
1739Hardware VLAN filter is off by default.
1740
1741The ``on`` option is equivalent to the ``--enable-hw-vlan-filter`` command-line option.
1742
1743port config - VLAN strip
1744~~~~~~~~~~~~~~~~~~~~~~~~
1745
1746Set hardware VLAN strip on or off for all ports::
1747
1748   testpmd> port config all hw-vlan-strip (on|off)
1749
1750Hardware VLAN strip is off by default.
1751
1752The ``on`` option is equivalent to the ``--enable-hw-vlan-strip`` command-line option.
1753
1754port config - VLAN extend
1755~~~~~~~~~~~~~~~~~~~~~~~~~
1756
1757Set hardware VLAN extend on or off for all ports::
1758
1759   testpmd> port config all hw-vlan-extend (on|off)
1760
1761Hardware VLAN extend is off by default.
1762
1763The ``on`` option is equivalent to the ``--enable-hw-vlan-extend`` command-line option.
1764
1765port config - Drop Packets
1766~~~~~~~~~~~~~~~~~~~~~~~~~~
1767
1768Set packet drop for packets with no descriptors on or off for all ports::
1769
1770   testpmd> port config all drop-en (on|off)
1771
1772Packet dropping for packets with no descriptors is off by default.
1773
1774The ``on`` option is equivalent to the ``--enable-drop-en`` command-line option.
1775
1776port config - RSS
1777~~~~~~~~~~~~~~~~~
1778
1779Set the RSS (Receive Side Scaling) mode on or off::
1780
1781   testpmd> port config all rss (all|ip|tcp|udp|sctp|ether|port|vxlan|geneve|nvgre|none)
1782
1783RSS is on by default.
1784
1785The ``none`` option is equivalent to the ``--disable-rss`` command-line option.
1786
1787port config - RSS Reta
1788~~~~~~~~~~~~~~~~~~~~~~
1789
1790Set the RSS (Receive Side Scaling) redirection table::
1791
1792   testpmd> port config all rss reta (hash,queue)[,(hash,queue)]
1793
1794port config - DCB
1795~~~~~~~~~~~~~~~~~
1796
1797Set the DCB mode for an individual port::
1798
1799   testpmd> port config (port_id) dcb vt (on|off) (traffic_class) pfc (on|off)
1800
1801The traffic class should be 4 or 8.
1802
1803port config - Burst
1804~~~~~~~~~~~~~~~~~~~
1805
1806Set the number of packets per burst::
1807
1808   testpmd> port config all burst (value)
1809
1810This is equivalent to the ``--burst`` command-line option.
1811
1812port config - Threshold
1813~~~~~~~~~~~~~~~~~~~~~~~
1814
1815Set thresholds for TX/RX queues::
1816
1817   testpmd> port config all (threshold) (value)
1818
1819Where the threshold type can be:
1820
1821* ``txpt:`` Set the prefetch threshold register of the TX rings, 0 <= value <= 255.
1822
1823* ``txht:`` Set the host threshold register of the TX rings, 0 <= value <= 255.
1824
1825* ``txwt:`` Set the write-back threshold register of the TX rings, 0 <= value <= 255.
1826
1827* ``rxpt:`` Set the prefetch threshold register of the RX rings, 0 <= value <= 255.
1828
1829* ``rxht:`` Set the host threshold register of the RX rings, 0 <= value <= 255.
1830
1831* ``rxwt:`` Set the write-back threshold register of the RX rings, 0 <= value <= 255.
1832
1833* ``txfreet:`` Set the transmit free threshold of the TX rings, 0 <= value <= txd.
1834
1835* ``rxfreet:`` Set the transmit free threshold of the RX rings, 0 <= value <= rxd.
1836
1837* ``txrst:`` Set the transmit RS bit threshold of TX rings, 0 <= value <= txd.
1838
1839These threshold options are also available from the command-line.
1840
1841port config - E-tag
1842~~~~~~~~~~~~~~~~~~~
1843
1844Set the value of ether-type for E-tag::
1845
1846   testpmd> port config (port_id|all) l2-tunnel E-tag ether-type (value)
1847
1848Enable/disable the E-tag support::
1849
1850   testpmd> port config (port_id|all) l2-tunnel E-tag (enable|disable)
1851
1852port config pctype mapping
1853~~~~~~~~~~~~~~~~~~~~~~~~~~
1854
1855Reset pctype mapping table::
1856
1857   testpmd> port config (port_id) pctype mapping reset
1858
1859Update hardware defined pctype to software defined flow type mapping table::
1860
1861   testpmd> port config (port_id) pctype mapping update (pctype_id_0[,pctype_id_1]*) (flow_type_id)
1862
1863where:
1864
1865* ``pctype_id_x``: hardware pctype id as index of bit in bitmask value of the pctype mapping table.
1866
1867* ``flow_type_id``: software flow type id as the index of the pctype mapping table.
1868
1869port config input set
1870~~~~~~~~~~~~~~~~~~~~~
1871
1872Config RSS/FDIR/FDIR flexible payload input set for some pctype::
1873   testpmd> port config (port_id) pctype (pctype_id) \
1874            (hash_inset|fdir_inset|fdir_flx_inset) \
1875	    (get|set|clear) field (field_idx)
1876
1877Clear RSS/FDIR/FDIR flexible payload input set for some pctype::
1878   testpmd> port config (port_id) pctype (pctype_id) \
1879            (hash_inset|fdir_inset|fdir_flx_inset) clear all
1880
1881where:
1882
1883* ``pctype_id``: hardware packet classification types.
1884* ``field_idx``: hardware field index.
1885
1886Link Bonding Functions
1887----------------------
1888
1889The Link Bonding functions make it possible to dynamically create and
1890manage link bonding devices from within testpmd interactive prompt.
1891
1892create bonded device
1893~~~~~~~~~~~~~~~~~~~~
1894
1895Create a new bonding device::
1896
1897   testpmd> create bonded device (mode) (socket)
1898
1899For example, to create a bonded device in mode 1 on socket 0::
1900
1901   testpmd> create bonded 1 0
1902   created new bonded device (port X)
1903
1904add bonding slave
1905~~~~~~~~~~~~~~~~~
1906
1907Adds Ethernet device to a Link Bonding device::
1908
1909   testpmd> add bonding slave (slave id) (port id)
1910
1911For example, to add Ethernet device (port 6) to a Link Bonding device (port 10)::
1912
1913   testpmd> add bonding slave 6 10
1914
1915
1916remove bonding slave
1917~~~~~~~~~~~~~~~~~~~~
1918
1919Removes an Ethernet slave device from a Link Bonding device::
1920
1921   testpmd> remove bonding slave (slave id) (port id)
1922
1923For example, to remove Ethernet slave device (port 6) to a Link Bonding device (port 10)::
1924
1925   testpmd> remove bonding slave 6 10
1926
1927set bonding mode
1928~~~~~~~~~~~~~~~~
1929
1930Set the Link Bonding mode of a Link Bonding device::
1931
1932   testpmd> set bonding mode (value) (port id)
1933
1934For example, to set the bonding mode of a Link Bonding device (port 10) to broadcast (mode 3)::
1935
1936   testpmd> set bonding mode 3 10
1937
1938set bonding primary
1939~~~~~~~~~~~~~~~~~~~
1940
1941Set an Ethernet slave device as the primary device on a Link Bonding device::
1942
1943   testpmd> set bonding primary (slave id) (port id)
1944
1945For example, to set the Ethernet slave device (port 6) as the primary port of a Link Bonding device (port 10)::
1946
1947   testpmd> set bonding primary 6 10
1948
1949set bonding mac
1950~~~~~~~~~~~~~~~
1951
1952Set the MAC address of a Link Bonding device::
1953
1954   testpmd> set bonding mac (port id) (mac)
1955
1956For example, to set the MAC address of a Link Bonding device (port 10) to 00:00:00:00:00:01::
1957
1958   testpmd> set bonding mac 10 00:00:00:00:00:01
1959
1960set bonding xmit_balance_policy
1961~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
1962
1963Set the transmission policy for a Link Bonding device when it is in Balance XOR mode::
1964
1965   testpmd> set bonding xmit_balance_policy (port_id) (l2|l23|l34)
1966
1967For example, set a Link Bonding device (port 10) to use a balance policy of layer 3+4 (IP addresses & UDP ports)::
1968
1969   testpmd> set bonding xmit_balance_policy 10 l34
1970
1971
1972set bonding mon_period
1973~~~~~~~~~~~~~~~~~~~~~~
1974
1975Set the link status monitoring polling period in milliseconds for a bonding device.
1976
1977This adds support for PMD slave devices which do not support link status interrupts.
1978When the mon_period is set to a value greater than 0 then all PMD's which do not support
1979link status ISR will be queried every polling interval to check if their link status has changed::
1980
1981   testpmd> set bonding mon_period (port_id) (value)
1982
1983For example, to set the link status monitoring polling period of bonded device (port 5) to 150ms::
1984
1985   testpmd> set bonding mon_period 5 150
1986
1987
1988set bonding lacp dedicated_queue
1989~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
1990
1991Enable dedicated tx/rx queues on bonding devices slaves to handle LACP control plane traffic
1992when in mode 4 (link-aggregration-802.3ad)::
1993
1994   testpmd> set bonding lacp dedicated_queues (port_id) (enable|disable)
1995
1996
1997set bonding agg_mode
1998~~~~~~~~~~~~~~~~~~~~
1999
2000Enable one of the specific aggregators mode when in mode 4 (link-aggregration-802.3ad)::
2001
2002   testpmd> set bonding agg_mode (port_id) (bandwidth|count|stable)
2003
2004
2005show bonding config
2006~~~~~~~~~~~~~~~~~~~
2007
2008Show the current configuration of a Link Bonding device::
2009
2010   testpmd> show bonding config (port id)
2011
2012For example,
2013to show the configuration a Link Bonding device (port 9) with 3 slave devices (1, 3, 4)
2014in balance mode with a transmission policy of layer 2+3::
2015
2016   testpmd> show bonding config 9
2017        Bonding mode: 2
2018        Balance Xmit Policy: BALANCE_XMIT_POLICY_LAYER23
2019        Slaves (3): [1 3 4]
2020        Active Slaves (3): [1 3 4]
2021        Primary: [3]
2022
2023
2024Register Functions
2025------------------
2026
2027The Register Functions can be used to read from and write to registers on the network card referenced by a port number.
2028This is mainly useful for debugging purposes.
2029Reference should be made to the appropriate datasheet for the network card for details on the register addresses
2030and fields that can be accessed.
2031
2032read reg
2033~~~~~~~~
2034
2035Display the value of a port register::
2036
2037   testpmd> read reg (port_id) (address)
2038
2039For example, to examine the Flow Director control register (FDIRCTL, 0x0000EE000) on an Intel 82599 10 GbE Controller::
2040
2041   testpmd> read reg 0 0xEE00
2042   port 0 PCI register at offset 0xEE00: 0x4A060029 (1241907241)
2043
2044read regfield
2045~~~~~~~~~~~~~
2046
2047Display a port register bit field::
2048
2049   testpmd> read regfield (port_id) (address) (bit_x) (bit_y)
2050
2051For example, reading the lowest two bits from the register in the example above::
2052
2053   testpmd> read regfield 0 0xEE00 0 1
2054   port 0 PCI register at offset 0xEE00: bits[0, 1]=0x1 (1)
2055
2056read regbit
2057~~~~~~~~~~~
2058
2059Display a single port register bit::
2060
2061   testpmd> read regbit (port_id) (address) (bit_x)
2062
2063For example, reading the lowest bit from the register in the example above::
2064
2065   testpmd> read regbit 0 0xEE00 0
2066   port 0 PCI register at offset 0xEE00: bit 0=1
2067
2068write reg
2069~~~~~~~~~
2070
2071Set the value of a port register::
2072
2073   testpmd> write reg (port_id) (address) (value)
2074
2075For example, to clear a register::
2076
2077   testpmd> write reg 0 0xEE00 0x0
2078   port 0 PCI register at offset 0xEE00: 0x00000000 (0)
2079
2080write regfield
2081~~~~~~~~~~~~~~
2082
2083Set bit field of a port register::
2084
2085   testpmd> write regfield (port_id) (address) (bit_x) (bit_y) (value)
2086
2087For example, writing to the register cleared in the example above::
2088
2089   testpmd> write regfield 0 0xEE00 0 1 2
2090   port 0 PCI register at offset 0xEE00: 0x00000002 (2)
2091
2092write regbit
2093~~~~~~~~~~~~
2094
2095Set single bit value of a port register::
2096
2097   testpmd> write regbit (port_id) (address) (bit_x) (value)
2098
2099For example, to set the high bit in the register from the example above::
2100
2101   testpmd> write regbit 0 0xEE00 31 1
2102   port 0 PCI register at offset 0xEE00: 0x8000000A (2147483658)
2103
2104Traffic Metering and Policing
2105-----------------------------
2106
2107The following section shows functions for configuring traffic metering and
2108policing on the ethernet device through the use of generic ethdev API.
2109
2110show port traffic management capability
2111~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
2112
2113Show traffic metering and policing capability of the port::
2114
2115   testpmd> show port meter cap (port_id)
2116
2117add port meter profile (srTCM rfc2967)
2118~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
2119
2120Add meter profile (srTCM rfc2697) to the ethernet device::
2121
2122   testpmd> add port meter profile srtcm_rfc2697 (port_id) (profile_id) \
2123   (cir) (cbs) (ebs)
2124
2125where:
2126
2127* ``profile_id``: ID for the meter profile.
2128* ``cir``: Committed Information Rate (CIR) (bytes/second).
2129* ``cbs``: Committed Burst Size (CBS) (bytes).
2130* ``ebs``: Excess Burst Size (EBS) (bytes).
2131
2132add port meter profile (trTCM rfc2968)
2133~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
2134
2135Add meter profile (srTCM rfc2698) to the ethernet device::
2136
2137   testpmd> add port meter profile trtcm_rfc2698 (port_id) (profile_id) \
2138   (cir) (pir) (cbs) (pbs)
2139
2140where:
2141
2142* ``profile_id``: ID for the meter profile.
2143* ``cir``: Committed information rate (bytes/second).
2144* ``pir``: Peak information rate (bytes/second).
2145* ``cbs``: Committed burst size (bytes).
2146* ``pbs``: Peak burst size (bytes).
2147
2148add port meter profile (trTCM rfc4115)
2149~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
2150
2151Add meter profile (trTCM rfc4115) to the ethernet device::
2152
2153   testpmd> add port meter profile trtcm_rfc4115 (port_id) (profile_id) \
2154   (cir) (eir) (cbs) (ebs)
2155
2156where:
2157
2158* ``profile_id``: ID for the meter profile.
2159* ``cir``: Committed information rate (bytes/second).
2160* ``eir``: Excess information rate (bytes/second).
2161* ``cbs``: Committed burst size (bytes).
2162* ``ebs``: Excess burst size (bytes).
2163
2164delete port meter profile
2165~~~~~~~~~~~~~~~~~~~~~~~~~
2166
2167Delete meter profile from the ethernet device::
2168
2169   testpmd> del port meter profile (port_id) (profile_id)
2170
2171create port meter
2172~~~~~~~~~~~~~~~~~
2173
2174Create new meter object for the ethernet device::
2175
2176   testpmd> create port meter (port_id) (mtr_id) (profile_id) \
2177   (meter_enable) (g_action) (y_action) (r_action) (stats_mask) (shared) \
2178   (use_pre_meter_color) [(dscp_tbl_entry0) (dscp_tbl_entry1)...\
2179   (dscp_tbl_entry63)]
2180
2181where:
2182
2183* ``mtr_id``: meter object ID.
2184* ``profile_id``: ID for the meter profile.
2185* ``meter_enable``: When this parameter has a non-zero value, the meter object
2186  gets enabled at the time of creation, otherwise remains disabled.
2187* ``g_action``: Policer action for the packet with green color.
2188* ``y_action``: Policer action for the packet with yellow color.
2189* ``r_action``: Policer action for the packet with red color.
2190* ``stats_mask``: Mask of statistics counter types to be enabled for the
2191  meter object.
2192* ``shared``:  When this parameter has a non-zero value, the meter object is
2193  shared by multiple flows. Otherwise, meter object is used by single flow.
2194* ``use_pre_meter_color``: When this parameter has a non-zero value, the
2195  input color for the current meter object is determined by the latest meter
2196  object in the same flow. Otherwise, the current meter object uses the
2197  *dscp_table* to determine the input color.
2198* ``dscp_tbl_entryx``: DSCP table entry x providing meter providing input
2199  color, 0 <= x <= 63.
2200
2201enable port meter
2202~~~~~~~~~~~~~~~~~
2203
2204Enable meter for the ethernet device::
2205
2206   testpmd> enable port meter (port_id) (mtr_id)
2207
2208disable port meter
2209~~~~~~~~~~~~~~~~~~
2210
2211Disable meter for the ethernet device::
2212
2213   testpmd> disable port meter (port_id) (mtr_id)
2214
2215delete port meter
2216~~~~~~~~~~~~~~~~~
2217
2218Delete meter for the ethernet device::
2219
2220   testpmd> del port meter (port_id) (mtr_id)
2221
2222Set port meter profile
2223~~~~~~~~~~~~~~~~~~~~~~
2224
2225Set meter profile for the ethernet device::
2226
2227   testpmd> set port meter profile (port_id) (mtr_id) (profile_id)
2228
2229set port meter dscp table
2230~~~~~~~~~~~~~~~~~~~~~~~~~
2231
2232Set meter dscp table for the ethernet device::
2233
2234   testpmd> set port meter dscp table (port_id) (mtr_id) [(dscp_tbl_entry0) \
2235   (dscp_tbl_entry1)...(dscp_tbl_entry63)]
2236
2237set port meter policer action
2238~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
2239
2240Set meter policer action for the ethernet device::
2241
2242   testpmd> set port meter policer action (port_id) (mtr_id) (action_mask) \
2243   (action0) [(action1) (action1)]
2244
2245where:
2246
2247* ``action_mask``: Bit mask indicating which policer actions need to be
2248  updated. One or more policer actions can be updated in a single function
2249  invocation. To update the policer action associated with color C, bit
2250  (1 << C) needs to be set in *action_mask* and element at position C
2251  in the *actions* array needs to be valid.
2252* ``actionx``: Policer action for the color x,
2253  RTE_MTR_GREEN <= x < RTE_MTR_COLORS
2254
2255set port meter stats mask
2256~~~~~~~~~~~~~~~~~~~~~~~~~
2257
2258Set meter stats mask for the ethernet device::
2259
2260   testpmd> set port meter stats mask (port_id) (mtr_id) (stats_mask)
2261
2262where:
2263
2264* ``stats_mask``: Bit mask indicating statistics counter types to be enabled.
2265
2266show port meter stats
2267~~~~~~~~~~~~~~~~~~~~~
2268
2269Show meter stats of the ethernet device::
2270
2271   testpmd> show port meter stats (port_id) (mtr_id) (clear)
2272
2273where:
2274
2275* ``clear``: Flag that indicates whether the statistics counters should
2276  be cleared (i.e. set to zero) immediately after they have been read or not.
2277
2278Traffic Management
2279------------------
2280
2281The following section shows functions for configuring traffic management on
2282on the ethernet device through the use of generic TM API.
2283
2284show port traffic management capability
2285~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
2286
2287Show traffic management capability of the port::
2288
2289   testpmd> show port tm cap (port_id)
2290
2291show port traffic management capability (hierarchy level)
2292~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
2293
2294Show traffic management hierarchy level capability of the port::
2295
2296   testpmd> show port tm level cap (port_id) (level_id)
2297
2298show port traffic management capability (hierarchy node level)
2299~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
2300
2301Show the traffic management hierarchy node capability of the port::
2302
2303   testpmd> show port tm node cap (port_id) (node_id)
2304
2305show port traffic management hierarchy node type
2306~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
2307
2308Show the port traffic management hierarchy node type::
2309
2310   testpmd> show port tm node type (port_id) (node_id)
2311
2312show port traffic management hierarchy node stats
2313~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
2314
2315Show the port traffic management hierarchy node statistics::
2316
2317   testpmd> show port tm node stats (port_id) (node_id) (clear)
2318
2319where:
2320
2321* ``clear``: When this parameter has a non-zero value, the statistics counters
2322  are cleared (i.e. set to zero) immediately after they have been read,
2323  otherwise the statistics counters are left untouched.
2324
2325Add port traffic management private shaper profile
2326~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
2327
2328Add the port traffic management private shaper profile::
2329
2330   testpmd> add port tm node shaper profile (port_id) (shaper_profile_id) \
2331   (tb_rate) (tb_size) (packet_length_adjust)
2332
2333where:
2334
2335* ``shaper_profile id``: Shaper profile ID for the new profile.
2336* ``tb_rate``: Token bucket rate (bytes per second).
2337* ``tb_size``: Token bucket size (bytes).
2338* ``packet_length_adjust``: The value (bytes) to be added to the length of
2339  each packet for the purpose of shaping. This parameter value can be used to
2340  correct the packet length with the framing overhead bytes that are consumed
2341  on the wire.
2342
2343Delete port traffic management private shaper profile
2344~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
2345
2346Delete the port traffic management private shaper::
2347
2348   testpmd> del port tm node shaper profile (port_id) (shaper_profile_id)
2349
2350where:
2351
2352* ``shaper_profile id``: Shaper profile ID that needs to be deleted.
2353
2354Add port traffic management shared shaper
2355~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
2356
2357Create the port traffic management shared shaper::
2358
2359   testpmd> add port tm node shared shaper (port_id) (shared_shaper_id) \
2360   (shaper_profile_id)
2361
2362where:
2363
2364* ``shared_shaper_id``: Shared shaper ID to be created.
2365* ``shaper_profile id``: Shaper profile ID for shared shaper.
2366
2367Set port traffic management shared shaper
2368~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
2369
2370Update the port traffic management shared shaper::
2371
2372   testpmd> set port tm node shared shaper (port_id) (shared_shaper_id) \
2373   (shaper_profile_id)
2374
2375where:
2376
2377* ``shared_shaper_id``: Shared shaper ID to be update.
2378* ``shaper_profile id``: Shaper profile ID for shared shaper.
2379
2380Delete port traffic management shared shaper
2381~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
2382
2383Delete the port traffic management shared shaper::
2384
2385   testpmd> del port tm node shared shaper (port_id) (shared_shaper_id)
2386
2387where:
2388
2389* ``shared_shaper_id``: Shared shaper ID to be deleted.
2390
2391Set port traffic management hiearchy node private shaper
2392~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
2393
2394set the port traffic management hierarchy node private shaper::
2395
2396   testpmd> set port tm node shaper profile (port_id) (node_id) \
2397   (shaper_profile_id)
2398
2399where:
2400
2401* ``shaper_profile id``: Private shaper profile ID to be enabled on the
2402  hierarchy node.
2403
2404Add port traffic management WRED profile
2405~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
2406
2407Create a new WRED profile::
2408
2409   testpmd> add port tm node wred profile (port_id) (wred_profile_id) \
2410   (color_g) (min_th_g) (max_th_g) (maxp_inv_g) (wq_log2_g) \
2411   (color_y) (min_th_y) (max_th_y) (maxp_inv_y) (wq_log2_y) \
2412   (color_r) (min_th_r) (max_th_r) (maxp_inv_r) (wq_log2_r)
2413
2414where:
2415
2416* ``wred_profile id``: Identifier for the newly create WRED profile
2417* ``color_g``: Packet color (green)
2418* ``min_th_g``: Minimum queue threshold for packet with green color
2419* ``max_th_g``: Minimum queue threshold for packet with green color
2420* ``maxp_inv_g``: Inverse of packet marking probability maximum value (maxp)
2421* ``wq_log2_g``: Negated log2 of queue weight (wq)
2422* ``color_y``: Packet color (yellow)
2423* ``min_th_y``: Minimum queue threshold for packet with yellow color
2424* ``max_th_y``: Minimum queue threshold for packet with yellow color
2425* ``maxp_inv_y``: Inverse of packet marking probability maximum value (maxp)
2426* ``wq_log2_y``: Negated log2 of queue weight (wq)
2427* ``color_r``: Packet color (red)
2428* ``min_th_r``: Minimum queue threshold for packet with yellow color
2429* ``max_th_r``: Minimum queue threshold for packet with yellow color
2430* ``maxp_inv_r``: Inverse of packet marking probability maximum value (maxp)
2431* ``wq_log2_r``: Negated log2 of queue weight (wq)
2432
2433Delete port traffic management WRED profile
2434~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
2435
2436Delete the WRED profile::
2437
2438   testpmd> del port tm node wred profile (port_id) (wred_profile_id)
2439
2440Add port traffic management hierarchy nonleaf node
2441~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
2442
2443Add nonleaf node to port traffic management hiearchy::
2444
2445   testpmd> add port tm nonleaf node (port_id) (node_id) (parent_node_id) \
2446   (priority) (weight) (level_id) (shaper_profile_id) \
2447   (n_sp_priorities) (stats_mask) (n_shared_shapers) \
2448   [(shared_shaper_0) (shared_shaper_1) ...] \
2449
2450where:
2451
2452* ``parent_node_id``: Node ID of the parent.
2453* ``priority``: Node priority (highest node priority is zero). This is used by
2454  the SP algorithm running on the parent node for scheduling this node.
2455* ``weight``: Node weight (lowest weight is one). The node weight is relative
2456  to the weight sum of all siblings that have the same priority. It is used by
2457  the WFQ algorithm running on the parent node for scheduling this node.
2458* ``level_id``: Hiearchy level of the node.
2459* ``shaper_profile_id``: Shaper profile ID of the private shaper to be used by
2460  the node.
2461* ``n_sp_priorities``: Number of strict priorities.
2462* ``stats_mask``: Mask of statistics counter types to be enabled for this node.
2463* ``n_shared_shapers``: Number of shared shapers.
2464* ``shared_shaper_id``: Shared shaper id.
2465
2466Add port traffic management hierarchy leaf node
2467~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
2468
2469Add leaf node to port traffic management hiearchy::
2470
2471   testpmd> add port tm leaf node (port_id) (node_id) (parent_node_id) \
2472   (priority) (weight) (level_id) (shaper_profile_id) \
2473   (cman_mode) (wred_profile_id) (stats_mask) (n_shared_shapers) \
2474   [(shared_shaper_id) (shared_shaper_id) ...] \
2475
2476where:
2477
2478* ``parent_node_id``: Node ID of the parent.
2479* ``priority``: Node priority (highest node priority is zero). This is used by
2480  the SP algorithm running on the parent node for scheduling this node.
2481* ``weight``: Node weight (lowest weight is one). The node weight is relative
2482  to the weight sum of all siblings that have the same priority. It is used by
2483  the WFQ algorithm running on the parent node for scheduling this node.
2484* ``level_id``: Hiearchy level of the node.
2485* ``shaper_profile_id``: Shaper profile ID of the private shaper to be used by
2486  the node.
2487* ``cman_mode``: Congestion management mode to be enabled for this node.
2488* ``wred_profile_id``: WRED profile id to be enabled for this node.
2489* ``stats_mask``: Mask of statistics counter types to be enabled for this node.
2490* ``n_shared_shapers``: Number of shared shapers.
2491* ``shared_shaper_id``: Shared shaper id.
2492
2493Delete port traffic management hierarchy node
2494~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
2495
2496Delete node from port traffic management hiearchy::
2497
2498   testpmd> del port tm node (port_id) (node_id)
2499
2500Update port traffic management hierarchy parent node
2501~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
2502
2503Update port traffic management hierarchy parent node::
2504
2505   testpmd> set port tm node parent (port_id) (node_id) (parent_node_id) \
2506   (priority) (weight)
2507
2508This function can only be called after the hierarchy commit invocation. Its
2509success depends on the port support for this operation, as advertised through
2510the port capability set. This function is valid for all nodes of the traffic
2511management hierarchy except root node.
2512
2513Commit port traffic management hierarchy
2514~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
2515
2516Commit the traffic management hierarchy on the port::
2517
2518   testpmd> port tm hierarchy commit (port_id) (clean_on_fail)
2519
2520where:
2521
2522* ``clean_on_fail``: When set to non-zero, hierarchy is cleared on function
2523  call failure. On the other hand, hierarchy is preserved when this parameter
2524  is equal to zero.
2525
2526Set port traffic management default hierarchy (tm forwarding mode)
2527~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
2528
2529set the traffic management default hierarchy on the port::
2530
2531   testpmd> set port tm hierarchy default (port_id)
2532
2533Filter Functions
2534----------------
2535
2536This section details the available filter functions that are available.
2537
2538Note these functions interface the deprecated legacy filtering framework,
2539superseded by *rte_flow*. See `Flow rules management`_.
2540
2541ethertype_filter
2542~~~~~~~~~~~~~~~~~~~~
2543
2544Add or delete a L2 Ethertype filter, which identify packets by their L2 Ethertype mainly assign them to a receive queue::
2545
2546   ethertype_filter (port_id) (add|del) (mac_addr|mac_ignr) (mac_address) \
2547                    ethertype (ether_type) (drop|fwd) queue (queue_id)
2548
2549The available information parameters are:
2550
2551* ``port_id``: The port which the Ethertype filter assigned on.
2552
2553* ``mac_addr``: Compare destination mac address.
2554
2555* ``mac_ignr``: Ignore destination mac address match.
2556
2557* ``mac_address``: Destination mac address to match.
2558
2559* ``ether_type``: The EtherType value want to match,
2560  for example 0x0806 for ARP packet. 0x0800 (IPv4) and 0x86DD (IPv6) are invalid.
2561
2562* ``queue_id``: The receive queue associated with this EtherType filter.
2563  It is meaningless when deleting or dropping.
2564
2565Example, to add/remove an ethertype filter rule::
2566
2567   testpmd> ethertype_filter 0 add mac_ignr 00:11:22:33:44:55 \
2568                             ethertype 0x0806 fwd queue 3
2569
2570   testpmd> ethertype_filter 0 del mac_ignr 00:11:22:33:44:55 \
2571                             ethertype 0x0806 fwd queue 3
2572
25732tuple_filter
2574~~~~~~~~~~~~~~~~~
2575
2576Add or delete a 2-tuple filter,
2577which identifies packets by specific protocol and destination TCP/UDP port
2578and forwards packets into one of the receive queues::
2579
2580   2tuple_filter (port_id) (add|del) dst_port (dst_port_value) \
2581                 protocol (protocol_value) mask (mask_value) \
2582                 tcp_flags (tcp_flags_value) priority (prio_value) \
2583                 queue (queue_id)
2584
2585The available information parameters are:
2586
2587* ``port_id``: The port which the 2-tuple filter assigned on.
2588
2589* ``dst_port_value``: Destination port in L4.
2590
2591* ``protocol_value``: IP L4 protocol.
2592
2593* ``mask_value``: Participates in the match or not by bit for field above, 1b means participate.
2594
2595* ``tcp_flags_value``: TCP control bits. The non-zero value is invalid, when the pro_value is not set to 0x06 (TCP).
2596
2597* ``prio_value``: Priority of this filter.
2598
2599* ``queue_id``: The receive queue associated with this 2-tuple filter.
2600
2601Example, to add/remove an 2tuple filter rule::
2602
2603   testpmd> 2tuple_filter 0 add dst_port 32 protocol 0x06 mask 0x03 \
2604                          tcp_flags 0x02 priority 3 queue 3
2605
2606   testpmd> 2tuple_filter 0 del dst_port 32 protocol 0x06 mask 0x03 \
2607                          tcp_flags 0x02 priority 3 queue 3
2608
26095tuple_filter
2610~~~~~~~~~~~~~~~~~
2611
2612Add or delete a 5-tuple filter,
2613which consists of a 5-tuple (protocol, source and destination IP addresses, source and destination TCP/UDP/SCTP port)
2614and routes packets into one of the receive queues::
2615
2616   5tuple_filter (port_id) (add|del) dst_ip (dst_address) src_ip \
2617                 (src_address) dst_port (dst_port_value) \
2618                 src_port (src_port_value) protocol (protocol_value) \
2619                 mask (mask_value) tcp_flags (tcp_flags_value) \
2620                 priority (prio_value) queue (queue_id)
2621
2622The available information parameters are:
2623
2624* ``port_id``: The port which the 5-tuple filter assigned on.
2625
2626* ``dst_address``: Destination IP address.
2627
2628* ``src_address``: Source IP address.
2629
2630* ``dst_port_value``: TCP/UDP destination port.
2631
2632* ``src_port_value``: TCP/UDP source port.
2633
2634* ``protocol_value``: L4 protocol.
2635
2636* ``mask_value``: Participates in the match or not by bit for field above, 1b means participate
2637
2638* ``tcp_flags_value``: TCP control bits. The non-zero value is invalid, when the protocol_value is not set to 0x06 (TCP).
2639
2640* ``prio_value``: The priority of this filter.
2641
2642* ``queue_id``: The receive queue associated with this 5-tuple filter.
2643
2644Example, to add/remove an 5tuple filter rule::
2645
2646   testpmd> 5tuple_filter 0 add dst_ip 2.2.2.5 src_ip 2.2.2.4 \
2647            dst_port 64 src_port 32 protocol 0x06 mask 0x1F \
2648            flags 0x0 priority 3 queue 3
2649
2650   testpmd> 5tuple_filter 0 del dst_ip 2.2.2.5 src_ip 2.2.2.4 \
2651            dst_port 64 src_port 32 protocol 0x06 mask 0x1F \
2652            flags 0x0 priority 3 queue 3
2653
2654syn_filter
2655~~~~~~~~~~
2656
2657Using the  SYN filter, TCP packets whose *SYN* flag is set can be forwarded to a separate queue::
2658
2659   syn_filter (port_id) (add|del) priority (high|low) queue (queue_id)
2660
2661The available information parameters are:
2662
2663* ``port_id``: The port which the SYN filter assigned on.
2664
2665* ``high``: This SYN filter has higher priority than other filters.
2666
2667* ``low``: This SYN filter has lower priority than other filters.
2668
2669* ``queue_id``: The receive queue associated with this SYN filter
2670
2671Example::
2672
2673   testpmd> syn_filter 0 add priority high queue 3
2674
2675flex_filter
2676~~~~~~~~~~~
2677
2678With flex filter, packets can be recognized by any arbitrary pattern within the first 128 bytes of the packet
2679and routed into one of the receive queues::
2680
2681   flex_filter (port_id) (add|del) len (len_value) bytes (bytes_value) \
2682               mask (mask_value) priority (prio_value) queue (queue_id)
2683
2684The available information parameters are:
2685
2686* ``port_id``: The port which the Flex filter is assigned on.
2687
2688* ``len_value``: Filter length in bytes, no greater than 128.
2689
2690* ``bytes_value``: A string in hexadecimal, means the value the flex filter needs to match.
2691
2692* ``mask_value``: A string in hexadecimal, bit 1 means corresponding byte participates in the match.
2693
2694* ``prio_value``: The priority of this filter.
2695
2696* ``queue_id``: The receive queue associated with this Flex filter.
2697
2698Example::
2699
2700   testpmd> flex_filter 0 add len 16 bytes 0x00000000000000000000000008060000 \
2701                          mask 000C priority 3 queue 3
2702
2703   testpmd> flex_filter 0 del len 16 bytes 0x00000000000000000000000008060000 \
2704                          mask 000C priority 3 queue 3
2705
2706
2707.. _testpmd_flow_director:
2708
2709flow_director_filter
2710~~~~~~~~~~~~~~~~~~~~
2711
2712The Flow Director works in receive mode to identify specific flows or sets of flows and route them to specific queues.
2713
2714Four types of filtering are supported which are referred to as Perfect Match, Signature, Perfect-mac-vlan and
2715Perfect-tunnel filters, the match mode is set by the ``--pkt-filter-mode`` command-line parameter:
2716
2717* Perfect match filters.
2718  The hardware checks a match between the masked fields of the received packets and the programmed filters.
2719  The masked fields are for IP flow.
2720
2721* Signature filters.
2722  The hardware checks a match between a hash-based signature of the masked fields of the received packet.
2723
2724* Perfect-mac-vlan match filters.
2725  The hardware checks a match between the masked fields of the received packets and the programmed filters.
2726  The masked fields are for MAC VLAN flow.
2727
2728* Perfect-tunnel match filters.
2729  The hardware checks a match between the masked fields of the received packets and the programmed filters.
2730  The masked fields are for tunnel flow.
2731
2732* Perfect-raw-flow-type match filters.
2733  The hardware checks a match between the masked fields of the received packets and pre-loaded raw (template) packet.
2734  The masked fields are specified by input sets.
2735
2736The Flow Director filters can match the different fields for different type of packet: flow type, specific input set
2737per flow type and the flexible payload.
2738
2739The Flow Director can also mask out parts of all of these fields so that filters
2740are only applied to certain fields or parts of the fields.
2741
2742Note that for raw flow type mode the source and destination fields in the
2743raw packet buffer need to be presented in a reversed order with respect
2744to the expected received packets.
2745For example: IP source and destination addresses or TCP/UDP/SCTP
2746source and destination ports
2747
2748Different NICs may have different capabilities, command show port fdir (port_id) can be used to acquire the information.
2749
2750# Commands to add flow director filters of different flow types::
2751
2752   flow_director_filter (port_id) mode IP (add|del|update) \
2753                        flow (ipv4-other|ipv4-frag|ipv6-other|ipv6-frag) \
2754                        src (src_ip_address) dst (dst_ip_address) \
2755                        tos (tos_value) proto (proto_value) ttl (ttl_value) \
2756                        vlan (vlan_value) flexbytes (flexbytes_value) \
2757                        (drop|fwd) pf|vf(vf_id) queue (queue_id) \
2758                        fd_id (fd_id_value)
2759
2760   flow_director_filter (port_id) mode IP (add|del|update) \
2761                        flow (ipv4-tcp|ipv4-udp|ipv6-tcp|ipv6-udp) \
2762                        src (src_ip_address) (src_port) \
2763                        dst (dst_ip_address) (dst_port) \
2764                        tos (tos_value) ttl (ttl_value) \
2765                        vlan (vlan_value) flexbytes (flexbytes_value) \
2766                        (drop|fwd) queue pf|vf(vf_id) (queue_id) \
2767                        fd_id (fd_id_value)
2768
2769   flow_director_filter (port_id) mode IP (add|del|update) \
2770                        flow (ipv4-sctp|ipv6-sctp) \
2771                        src (src_ip_address) (src_port) \
2772                        dst (dst_ip_address) (dst_port) \
2773                        tos (tos_value) ttl (ttl_value) \
2774                        tag (verification_tag) vlan (vlan_value) \
2775                        flexbytes (flexbytes_value) (drop|fwd) \
2776                        pf|vf(vf_id) queue (queue_id) fd_id (fd_id_value)
2777
2778   flow_director_filter (port_id) mode IP (add|del|update) flow l2_payload \
2779                        ether (ethertype) flexbytes (flexbytes_value) \
2780                        (drop|fwd) pf|vf(vf_id) queue (queue_id)
2781                        fd_id (fd_id_value)
2782
2783   flow_director_filter (port_id) mode MAC-VLAN (add|del|update) \
2784                        mac (mac_address) vlan (vlan_value) \
2785                        flexbytes (flexbytes_value) (drop|fwd) \
2786                        queue (queue_id) fd_id (fd_id_value)
2787
2788   flow_director_filter (port_id) mode Tunnel (add|del|update) \
2789                        mac (mac_address) vlan (vlan_value) \
2790                        tunnel (NVGRE|VxLAN) tunnel-id (tunnel_id_value) \
2791                        flexbytes (flexbytes_value) (drop|fwd) \
2792                        queue (queue_id) fd_id (fd_id_value)
2793
2794   flow_director_filter (port_id) mode raw (add|del|update) flow (flow_id) \
2795                        (drop|fwd) queue (queue_id) fd_id (fd_id_value) \
2796                        packet (packet file name)
2797
2798For example, to add an ipv4-udp flow type filter::
2799
2800   testpmd> flow_director_filter 0 mode IP add flow ipv4-udp src 2.2.2.3 32 \
2801            dst 2.2.2.5 33 tos 2 ttl 40 vlan 0x1 flexbytes (0x88,0x48) \
2802            fwd pf queue 1 fd_id 1
2803
2804For example, add an ipv4-other flow type filter::
2805
2806   testpmd> flow_director_filter 0 mode IP add flow ipv4-other src 2.2.2.3 \
2807             dst 2.2.2.5 tos 2 proto 20 ttl 40 vlan 0x1 \
2808             flexbytes (0x88,0x48) fwd pf queue 1 fd_id 1
2809
2810flush_flow_director
2811~~~~~~~~~~~~~~~~~~~
2812
2813Flush all flow director filters on a device::
2814
2815   testpmd> flush_flow_director (port_id)
2816
2817Example, to flush all flow director filter on port 0::
2818
2819   testpmd> flush_flow_director 0
2820
2821flow_director_mask
2822~~~~~~~~~~~~~~~~~~
2823
2824Set flow director's input masks::
2825
2826   flow_director_mask (port_id) mode IP vlan (vlan_value) \
2827                      src_mask (ipv4_src) (ipv6_src) (src_port) \
2828                      dst_mask (ipv4_dst) (ipv6_dst) (dst_port)
2829
2830   flow_director_mask (port_id) mode MAC-VLAN vlan (vlan_value)
2831
2832   flow_director_mask (port_id) mode Tunnel vlan (vlan_value) \
2833                      mac (mac_value) tunnel-type (tunnel_type_value) \
2834                      tunnel-id (tunnel_id_value)
2835
2836Example, to set flow director mask on port 0::
2837
2838   testpmd> flow_director_mask 0 mode IP vlan 0xefff \
2839            src_mask 255.255.255.255 \
2840                FFFF:FFFF:FFFF:FFFF:FFFF:FFFF:FFFF:FFFF 0xFFFF \
2841            dst_mask 255.255.255.255 \
2842                FFFF:FFFF:FFFF:FFFF:FFFF:FFFF:FFFF:FFFF 0xFFFF
2843
2844flow_director_flex_mask
2845~~~~~~~~~~~~~~~~~~~~~~~
2846
2847set masks of flow director's flexible payload based on certain flow type::
2848
2849   testpmd> flow_director_flex_mask (port_id) \
2850            flow (none|ipv4-other|ipv4-frag|ipv4-tcp|ipv4-udp|ipv4-sctp| \
2851                  ipv6-other|ipv6-frag|ipv6-tcp|ipv6-udp|ipv6-sctp| \
2852                  l2_payload|all) (mask)
2853
2854Example, to set flow director's flex mask for all flow type on port 0::
2855
2856   testpmd> flow_director_flex_mask 0 flow all \
2857            (0xff,0xff,0,0,0,0,0,0,0,0,0,0,0,0,0,0)
2858
2859
2860flow_director_flex_payload
2861~~~~~~~~~~~~~~~~~~~~~~~~~~
2862
2863Configure flexible payload selection::
2864
2865   flow_director_flex_payload (port_id) (raw|l2|l3|l4) (config)
2866
2867For example, to select the first 16 bytes from the offset 4 (bytes) of packet's payload as flexible payload::
2868
2869   testpmd> flow_director_flex_payload 0 l4 \
2870            (4,5,6,7,8,9,10,11,12,13,14,15,16,17,18,19)
2871
2872get_sym_hash_ena_per_port
2873~~~~~~~~~~~~~~~~~~~~~~~~~
2874
2875Get symmetric hash enable configuration per port::
2876
2877   get_sym_hash_ena_per_port (port_id)
2878
2879For example, to get symmetric hash enable configuration of port 1::
2880
2881   testpmd> get_sym_hash_ena_per_port 1
2882
2883set_sym_hash_ena_per_port
2884~~~~~~~~~~~~~~~~~~~~~~~~~
2885
2886Set symmetric hash enable configuration per port to enable or disable::
2887
2888   set_sym_hash_ena_per_port (port_id) (enable|disable)
2889
2890For example, to set symmetric hash enable configuration of port 1 to enable::
2891
2892   testpmd> set_sym_hash_ena_per_port 1 enable
2893
2894get_hash_global_config
2895~~~~~~~~~~~~~~~~~~~~~~
2896
2897Get the global configurations of hash filters::
2898
2899   get_hash_global_config (port_id)
2900
2901For example, to get the global configurations of hash filters of port 1::
2902
2903   testpmd> get_hash_global_config 1
2904
2905set_hash_global_config
2906~~~~~~~~~~~~~~~~~~~~~~
2907
2908Set the global configurations of hash filters::
2909
2910   set_hash_global_config (port_id) (toeplitz|simple_xor|default) \
2911   (ipv4|ipv4-frag|ipv4-tcp|ipv4-udp|ipv4-sctp|ipv4-other|ipv6|ipv6-frag| \
2912   ipv6-tcp|ipv6-udp|ipv6-sctp|ipv6-other|l2_payload|<flow_id>) \
2913   (enable|disable)
2914
2915For example, to enable simple_xor for flow type of ipv6 on port 2::
2916
2917   testpmd> set_hash_global_config 2 simple_xor ipv6 enable
2918
2919set_hash_input_set
2920~~~~~~~~~~~~~~~~~~
2921
2922Set the input set for hash::
2923
2924   set_hash_input_set (port_id) (ipv4-frag|ipv4-tcp|ipv4-udp|ipv4-sctp| \
2925   ipv4-other|ipv6-frag|ipv6-tcp|ipv6-udp|ipv6-sctp|ipv6-other| \
2926   l2_payload|<flow_id>) (ovlan|ivlan|src-ipv4|dst-ipv4|src-ipv6|dst-ipv6| \
2927   ipv4-tos|ipv4-proto|ipv6-tc|ipv6-next-header|udp-src-port|udp-dst-port| \
2928   tcp-src-port|tcp-dst-port|sctp-src-port|sctp-dst-port|sctp-veri-tag| \
2929   udp-key|gre-key|fld-1st|fld-2nd|fld-3rd|fld-4th|fld-5th|fld-6th|fld-7th| \
2930   fld-8th|none) (select|add)
2931
2932For example, to add source IP to hash input set for flow type of ipv4-udp on port 0::
2933
2934   testpmd> set_hash_input_set 0 ipv4-udp src-ipv4 add
2935
2936set_fdir_input_set
2937~~~~~~~~~~~~~~~~~~
2938
2939The Flow Director filters can match the different fields for different type of packet, i.e. specific input set
2940on per flow type and the flexible payload. This command can be used to change input set for each flow type.
2941
2942Set the input set for flow director::
2943
2944   set_fdir_input_set (port_id) (ipv4-frag|ipv4-tcp|ipv4-udp|ipv4-sctp| \
2945   ipv4-other|ipv6|ipv6-frag|ipv6-tcp|ipv6-udp|ipv6-sctp|ipv6-other| \
2946   l2_payload|<flow_id>) (ivlan|ethertype|src-ipv4|dst-ipv4|src-ipv6|dst-ipv6| \
2947   ipv4-tos|ipv4-proto|ipv4-ttl|ipv6-tc|ipv6-next-header|ipv6-hop-limits| \
2948   tudp-src-port|udp-dst-port|cp-src-port|tcp-dst-port|sctp-src-port| \
2949   sctp-dst-port|sctp-veri-tag|none) (select|add)
2950
2951For example to add source IP to FD input set for flow type of ipv4-udp on port 0::
2952
2953   testpmd> set_fdir_input_set 0 ipv4-udp src-ipv4 add
2954
2955global_config
2956~~~~~~~~~~~~~
2957
2958Set different GRE key length for input set::
2959
2960   global_config (port_id) gre-key-len (number in bytes)
2961
2962For example to set GRE key length for input set to 4 bytes on port 0::
2963
2964   testpmd> global_config 0 gre-key-len 4
2965
2966
2967.. _testpmd_rte_flow:
2968
2969Flow rules management
2970---------------------
2971
2972Control of the generic flow API (*rte_flow*) is fully exposed through the
2973``flow`` command (validation, creation, destruction, queries and operation
2974modes).
2975
2976Considering *rte_flow* overlaps with all `Filter Functions`_, using both
2977features simultaneously may cause undefined side-effects and is therefore
2978not recommended.
2979
2980``flow`` syntax
2981~~~~~~~~~~~~~~~
2982
2983Because the ``flow`` command uses dynamic tokens to handle the large number
2984of possible flow rules combinations, its behavior differs slightly from
2985other commands, in particular:
2986
2987- Pressing *?* or the *<tab>* key displays contextual help for the current
2988  token, not that of the entire command.
2989
2990- Optional and repeated parameters are supported (provided they are listed
2991  in the contextual help).
2992
2993The first parameter stands for the operation mode. Possible operations and
2994their general syntax are described below. They are covered in detail in the
2995following sections.
2996
2997- Check whether a flow rule can be created::
2998
2999   flow validate {port_id}
3000       [group {group_id}] [priority {level}] [ingress] [egress]
3001       pattern {item} [/ {item} [...]] / end
3002       actions {action} [/ {action} [...]] / end
3003
3004- Create a flow rule::
3005
3006   flow create {port_id}
3007       [group {group_id}] [priority {level}] [ingress] [egress]
3008       pattern {item} [/ {item} [...]] / end
3009       actions {action} [/ {action} [...]] / end
3010
3011- Destroy specific flow rules::
3012
3013   flow destroy {port_id} rule {rule_id} [...]
3014
3015- Destroy all flow rules::
3016
3017   flow flush {port_id}
3018
3019- Query an existing flow rule::
3020
3021   flow query {port_id} {rule_id} {action}
3022
3023- List existing flow rules sorted by priority, filtered by group
3024  identifiers::
3025
3026   flow list {port_id} [group {group_id}] [...]
3027
3028- Restrict ingress traffic to the defined flow rules::
3029
3030   flow isolate {port_id} {boolean}
3031
3032Validating flow rules
3033~~~~~~~~~~~~~~~~~~~~~
3034
3035``flow validate`` reports whether a flow rule would be accepted by the
3036underlying device in its current state but stops short of creating it. It is
3037bound to ``rte_flow_validate()``::
3038
3039   flow validate {port_id}
3040      [group {group_id}] [priority {level}] [ingress] [egress]
3041      pattern {item} [/ {item} [...]] / end
3042      actions {action} [/ {action} [...]] / end
3043
3044If successful, it will show::
3045
3046   Flow rule validated
3047
3048Otherwise it will show an error message of the form::
3049
3050   Caught error type [...] ([...]): [...]
3051
3052This command uses the same parameters as ``flow create``, their format is
3053described in `Creating flow rules`_.
3054
3055Check whether redirecting any Ethernet packet received on port 0 to RX queue
3056index 6 is supported::
3057
3058   testpmd> flow validate 0 ingress pattern eth / end
3059      actions queue index 6 / end
3060   Flow rule validated
3061   testpmd>
3062
3063Port 0 does not support TCPv6 rules::
3064
3065   testpmd> flow validate 0 ingress pattern eth / ipv6 / tcp / end
3066      actions drop / end
3067   Caught error type 9 (specific pattern item): Invalid argument
3068   testpmd>
3069
3070Creating flow rules
3071~~~~~~~~~~~~~~~~~~~
3072
3073``flow create`` validates and creates the specified flow rule. It is bound
3074to ``rte_flow_create()``::
3075
3076   flow create {port_id}
3077      [group {group_id}] [priority {level}] [ingress] [egress]
3078      pattern {item} [/ {item} [...]] / end
3079      actions {action} [/ {action} [...]] / end
3080
3081If successful, it will return a flow rule ID usable with other commands::
3082
3083   Flow rule #[...] created
3084
3085Otherwise it will show an error message of the form::
3086
3087   Caught error type [...] ([...]): [...]
3088
3089Parameters describe in the following order:
3090
3091- Attributes (*group*, *priority*, *ingress*, *egress* tokens).
3092- A matching pattern, starting with the *pattern* token and terminated by an
3093  *end* pattern item.
3094- Actions, starting with the *actions* token and terminated by an *end*
3095  action.
3096
3097These translate directly to *rte_flow* objects provided as-is to the
3098underlying functions.
3099
3100The shortest valid definition only comprises mandatory tokens::
3101
3102   testpmd> flow create 0 pattern end actions end
3103
3104Note that PMDs may refuse rules that essentially do nothing such as this
3105one.
3106
3107**All unspecified object values are automatically initialized to 0.**
3108
3109Attributes
3110^^^^^^^^^^
3111
3112These tokens affect flow rule attributes (``struct rte_flow_attr``) and are
3113specified before the ``pattern`` token.
3114
3115- ``group {group id}``: priority group.
3116- ``priority {level}``: priority level within group.
3117- ``ingress``: rule applies to ingress traffic.
3118- ``egress``: rule applies to egress traffic.
3119
3120Each instance of an attribute specified several times overrides the previous
3121value as shown below (group 4 is used)::
3122
3123   testpmd> flow create 0 group 42 group 24 group 4 [...]
3124
3125Note that once enabled, ``ingress`` and ``egress`` cannot be disabled.
3126
3127While not specifying a direction is an error, some rules may allow both
3128simultaneously.
3129
3130Most rules affect RX therefore contain the ``ingress`` token::
3131
3132   testpmd> flow create 0 ingress pattern [...]
3133
3134Matching pattern
3135^^^^^^^^^^^^^^^^
3136
3137A matching pattern starts after the ``pattern`` token. It is made of pattern
3138items and is terminated by a mandatory ``end`` item.
3139
3140Items are named after their type (*RTE_FLOW_ITEM_TYPE_* from ``enum
3141rte_flow_item_type``).
3142
3143The ``/`` token is used as a separator between pattern items as shown
3144below::
3145
3146   testpmd> flow create 0 ingress pattern eth / ipv4 / udp / end [...]
3147
3148Note that protocol items like these must be stacked from lowest to highest
3149layer to make sense. For instance, the following rule is either invalid or
3150unlikely to match any packet::
3151
3152   testpmd> flow create 0 ingress pattern eth / udp / ipv4 / end [...]
3153
3154More information on these restrictions can be found in the *rte_flow*
3155documentation.
3156
3157Several items support additional specification structures, for example
3158``ipv4`` allows specifying source and destination addresses as follows::
3159
3160   testpmd> flow create 0 ingress pattern eth / ipv4 src is 10.1.1.1
3161      dst is 10.2.0.0 / end [...]
3162
3163This rule matches all IPv4 traffic with the specified properties.
3164
3165In this example, ``src`` and ``dst`` are field names of the underlying
3166``struct rte_flow_item_ipv4`` object. All item properties can be specified
3167in a similar fashion.
3168
3169The ``is`` token means that the subsequent value must be matched exactly,
3170and assigns ``spec`` and ``mask`` fields in ``struct rte_flow_item``
3171accordingly. Possible assignment tokens are:
3172
3173- ``is``: match value perfectly (with full bit-mask).
3174- ``spec``: match value according to configured bit-mask.
3175- ``last``: specify upper bound to establish a range.
3176- ``mask``: specify bit-mask with relevant bits set to one.
3177- ``prefix``: generate bit-mask from a prefix length.
3178
3179These yield identical results::
3180
3181   ipv4 src is 10.1.1.1
3182
3183::
3184
3185   ipv4 src spec 10.1.1.1 src mask 255.255.255.255
3186
3187::
3188
3189   ipv4 src spec 10.1.1.1 src prefix 32
3190
3191::
3192
3193   ipv4 src is 10.1.1.1 src last 10.1.1.1 # range with a single value
3194
3195::
3196
3197   ipv4 src is 10.1.1.1 src last 0 # 0 disables range
3198
3199Inclusive ranges can be defined with ``last``::
3200
3201   ipv4 src is 10.1.1.1 src last 10.2.3.4 # 10.1.1.1 to 10.2.3.4
3202
3203Note that ``mask`` affects both ``spec`` and ``last``::
3204
3205   ipv4 src is 10.1.1.1 src last 10.2.3.4 src mask 255.255.0.0
3206      # matches 10.1.0.0 to 10.2.255.255
3207
3208Properties can be modified multiple times::
3209
3210   ipv4 src is 10.1.1.1 src is 10.1.2.3 src is 10.2.3.4 # matches 10.2.3.4
3211
3212::
3213
3214   ipv4 src is 10.1.1.1 src prefix 24 src prefix 16 # matches 10.1.0.0/16
3215
3216Pattern items
3217^^^^^^^^^^^^^
3218
3219This section lists supported pattern items and their attributes, if any.
3220
3221- ``end``: end list of pattern items.
3222
3223- ``void``: no-op pattern item.
3224
3225- ``invert``: perform actions when pattern does not match.
3226
3227- ``any``: match any protocol for the current layer.
3228
3229  - ``num {unsigned}``: number of layers covered.
3230
3231- ``pf``: match packets addressed to the physical function.
3232
3233- ``vf``: match packets addressed to a virtual function ID.
3234
3235  - ``id {unsigned}``: destination VF ID.
3236
3237- ``port``: device-specific physical port index to use.
3238
3239  - ``index {unsigned}``: physical port index.
3240
3241- ``raw``: match an arbitrary byte string.
3242
3243  - ``relative {boolean}``: look for pattern after the previous item.
3244  - ``search {boolean}``: search pattern from offset (see also limit).
3245  - ``offset {integer}``: absolute or relative offset for pattern.
3246  - ``limit {unsigned}``: search area limit for start of pattern.
3247  - ``pattern {string}``: byte string to look for.
3248
3249- ``eth``: match Ethernet header.
3250
3251  - ``dst {MAC-48}``: destination MAC.
3252  - ``src {MAC-48}``: source MAC.
3253  - ``type {unsigned}``: EtherType.
3254
3255- ``vlan``: match 802.1Q/ad VLAN tag.
3256
3257  - ``tpid {unsigned}``: tag protocol identifier.
3258  - ``tci {unsigned}``: tag control information.
3259  - ``pcp {unsigned}``: priority code point.
3260  - ``dei {unsigned}``: drop eligible indicator.
3261  - ``vid {unsigned}``: VLAN identifier.
3262
3263- ``ipv4``: match IPv4 header.
3264
3265  - ``tos {unsigned}``: type of service.
3266  - ``ttl {unsigned}``: time to live.
3267  - ``proto {unsigned}``: next protocol ID.
3268  - ``src {ipv4 address}``: source address.
3269  - ``dst {ipv4 address}``: destination address.
3270
3271- ``ipv6``: match IPv6 header.
3272
3273  - ``tc {unsigned}``: traffic class.
3274  - ``flow {unsigned}``: flow label.
3275  - ``proto {unsigned}``: protocol (next header).
3276  - ``hop {unsigned}``: hop limit.
3277  - ``src {ipv6 address}``: source address.
3278  - ``dst {ipv6 address}``: destination address.
3279
3280- ``icmp``: match ICMP header.
3281
3282  - ``type {unsigned}``: ICMP packet type.
3283  - ``code {unsigned}``: ICMP packet code.
3284
3285- ``udp``: match UDP header.
3286
3287  - ``src {unsigned}``: UDP source port.
3288  - ``dst {unsigned}``: UDP destination port.
3289
3290- ``tcp``: match TCP header.
3291
3292  - ``src {unsigned}``: TCP source port.
3293  - ``dst {unsigned}``: TCP destination port.
3294
3295- ``sctp``: match SCTP header.
3296
3297  - ``src {unsigned}``: SCTP source port.
3298  - ``dst {unsigned}``: SCTP destination port.
3299  - ``tag {unsigned}``: validation tag.
3300  - ``cksum {unsigned}``: checksum.
3301
3302- ``vxlan``: match VXLAN header.
3303
3304  - ``vni {unsigned}``: VXLAN identifier.
3305
3306- ``e_tag``: match IEEE 802.1BR E-Tag header.
3307
3308  - ``grp_ecid_b {unsigned}``: GRP and E-CID base.
3309
3310- ``nvgre``: match NVGRE header.
3311
3312  - ``tni {unsigned}``: virtual subnet ID.
3313
3314- ``mpls``: match MPLS header.
3315
3316  - ``label {unsigned}``: MPLS label.
3317
3318- ``gre``: match GRE header.
3319
3320  - ``protocol {unsigned}``: protocol type.
3321
3322- ``fuzzy``: fuzzy pattern match, expect faster than default.
3323
3324  - ``thresh {unsigned}``: accuracy threshold.
3325
3326- ``gtp``, ``gtpc``, ``gtpu``: match GTPv1 header.
3327
3328  - ``teid {unsigned}``: tunnel endpoint identifier.
3329
3330- ``geneve``: match GENEVE header.
3331
3332  - ``vni {unsigned}``: virtual network identifier.
3333  - ``protocol {unsigned}``: protocol type.
3334
3335Actions list
3336^^^^^^^^^^^^
3337
3338A list of actions starts after the ``actions`` token in the same fashion as
3339`Matching pattern`_; actions are separated by ``/`` tokens and the list is
3340terminated by a mandatory ``end`` action.
3341
3342Actions are named after their type (*RTE_FLOW_ACTION_TYPE_* from ``enum
3343rte_flow_action_type``).
3344
3345Dropping all incoming UDPv4 packets can be expressed as follows::
3346
3347   testpmd> flow create 0 ingress pattern eth / ipv4 / udp / end
3348      actions drop / end
3349
3350Several actions have configurable properties which must be specified when
3351there is no valid default value. For example, ``queue`` requires a target
3352queue index.
3353
3354This rule redirects incoming UDPv4 traffic to queue index 6::
3355
3356   testpmd> flow create 0 ingress pattern eth / ipv4 / udp / end
3357      actions queue index 6 / end
3358
3359While this one could be rejected by PMDs (unspecified queue index)::
3360
3361   testpmd> flow create 0 ingress pattern eth / ipv4 / udp / end
3362      actions queue / end
3363
3364As defined by *rte_flow*, the list is not ordered, all actions of a given
3365rule are performed simultaneously. These are equivalent::
3366
3367   queue index 6 / void / mark id 42 / end
3368
3369::
3370
3371   void / mark id 42 / queue index 6 / end
3372
3373All actions in a list should have different types, otherwise only the last
3374action of a given type is taken into account::
3375
3376   queue index 4 / queue index 5 / queue index 6 / end # will use queue 6
3377
3378::
3379
3380   drop / drop / drop / end # drop is performed only once
3381
3382::
3383
3384   mark id 42 / queue index 3 / mark id 24 / end # mark will be 24
3385
3386Considering they are performed simultaneously, opposite and overlapping
3387actions can sometimes be combined when the end result is unambiguous::
3388
3389   drop / queue index 6 / end # drop has no effect
3390
3391::
3392
3393   drop / dup index 6 / end # same as above
3394
3395::
3396
3397   queue index 6 / rss queues 6 7 8 / end # queue has no effect
3398
3399::
3400
3401   drop / passthru / end # drop has no effect
3402
3403Note that PMDs may still refuse such combinations.
3404
3405Actions
3406^^^^^^^
3407
3408This section lists supported actions and their attributes, if any.
3409
3410- ``end``: end list of actions.
3411
3412- ``void``: no-op action.
3413
3414- ``passthru``: let subsequent rule process matched packets.
3415
3416- ``mark``: attach 32 bit value to packets.
3417
3418  - ``id {unsigned}``: 32 bit value to return with packets.
3419
3420- ``flag``: flag packets.
3421
3422- ``queue``: assign packets to a given queue index.
3423
3424  - ``index {unsigned}``: queue index to use.
3425
3426- ``drop``: drop packets (note: passthru has priority).
3427
3428- ``count``: enable counters for this rule.
3429
3430- ``dup``: duplicate packets to a given queue index.
3431
3432  - ``index {unsigned}``: queue index to duplicate packets to.
3433
3434- ``rss``: spread packets among several queues.
3435
3436  - ``queues [{unsigned} [...]] end``: queue indices to use.
3437
3438- ``pf``: redirect packets to physical device function.
3439
3440- ``vf``: redirect packets to virtual device function.
3441
3442  - ``original {boolean}``: use original VF ID if possible.
3443  - ``id {unsigned}``: VF ID to redirect packets to.
3444
3445Destroying flow rules
3446~~~~~~~~~~~~~~~~~~~~~
3447
3448``flow destroy`` destroys one or more rules from their rule ID (as returned
3449by ``flow create``), this command calls ``rte_flow_destroy()`` as many
3450times as necessary::
3451
3452   flow destroy {port_id} rule {rule_id} [...]
3453
3454If successful, it will show::
3455
3456   Flow rule #[...] destroyed
3457
3458It does not report anything for rule IDs that do not exist. The usual error
3459message is shown when a rule cannot be destroyed::
3460
3461   Caught error type [...] ([...]): [...]
3462
3463``flow flush`` destroys all rules on a device and does not take extra
3464arguments. It is bound to ``rte_flow_flush()``::
3465
3466   flow flush {port_id}
3467
3468Any errors are reported as above.
3469
3470Creating several rules and destroying them::
3471
3472   testpmd> flow create 0 ingress pattern eth / ipv6 / end
3473      actions queue index 2 / end
3474   Flow rule #0 created
3475   testpmd> flow create 0 ingress pattern eth / ipv4 / end
3476      actions queue index 3 / end
3477   Flow rule #1 created
3478   testpmd> flow destroy 0 rule 0 rule 1
3479   Flow rule #1 destroyed
3480   Flow rule #0 destroyed
3481   testpmd>
3482
3483The same result can be achieved using ``flow flush``::
3484
3485   testpmd> flow create 0 ingress pattern eth / ipv6 / end
3486      actions queue index 2 / end
3487   Flow rule #0 created
3488   testpmd> flow create 0 ingress pattern eth / ipv4 / end
3489      actions queue index 3 / end
3490   Flow rule #1 created
3491   testpmd> flow flush 0
3492   testpmd>
3493
3494Non-existent rule IDs are ignored::
3495
3496   testpmd> flow create 0 ingress pattern eth / ipv6 / end
3497      actions queue index 2 / end
3498   Flow rule #0 created
3499   testpmd> flow create 0 ingress pattern eth / ipv4 / end
3500      actions queue index 3 / end
3501   Flow rule #1 created
3502   testpmd> flow destroy 0 rule 42 rule 10 rule 2
3503   testpmd>
3504   testpmd> flow destroy 0 rule 0
3505   Flow rule #0 destroyed
3506   testpmd>
3507
3508Querying flow rules
3509~~~~~~~~~~~~~~~~~~~
3510
3511``flow query`` queries a specific action of a flow rule having that
3512ability. Such actions collect information that can be reported using this
3513command. It is bound to ``rte_flow_query()``::
3514
3515   flow query {port_id} {rule_id} {action}
3516
3517If successful, it will display either the retrieved data for known actions
3518or the following message::
3519
3520   Cannot display result for action type [...] ([...])
3521
3522Otherwise, it will complain either that the rule does not exist or that some
3523error occurred::
3524
3525   Flow rule #[...] not found
3526
3527::
3528
3529   Caught error type [...] ([...]): [...]
3530
3531Currently only the ``count`` action is supported. This action reports the
3532number of packets that hit the flow rule and the total number of bytes. Its
3533output has the following format::
3534
3535   count:
3536    hits_set: [...] # whether "hits" contains a valid value
3537    bytes_set: [...] # whether "bytes" contains a valid value
3538    hits: [...] # number of packets
3539    bytes: [...] # number of bytes
3540
3541Querying counters for TCPv6 packets redirected to queue 6::
3542
3543   testpmd> flow create 0 ingress pattern eth / ipv6 / tcp / end
3544      actions queue index 6 / count / end
3545   Flow rule #4 created
3546   testpmd> flow query 0 4 count
3547   count:
3548    hits_set: 1
3549    bytes_set: 0
3550    hits: 386446
3551    bytes: 0
3552   testpmd>
3553
3554Listing flow rules
3555~~~~~~~~~~~~~~~~~~
3556
3557``flow list`` lists existing flow rules sorted by priority and optionally
3558filtered by group identifiers::
3559
3560   flow list {port_id} [group {group_id}] [...]
3561
3562This command only fails with the following message if the device does not
3563exist::
3564
3565   Invalid port [...]
3566
3567Output consists of a header line followed by a short description of each
3568flow rule, one per line. There is no output at all when no flow rules are
3569configured on the device::
3570
3571   ID      Group   Prio    Attr    Rule
3572   [...]   [...]   [...]   [...]   [...]
3573
3574``Attr`` column flags:
3575
3576- ``i`` for ``ingress``.
3577- ``e`` for ``egress``.
3578
3579Creating several flow rules and listing them::
3580
3581   testpmd> flow create 0 ingress pattern eth / ipv4 / end
3582      actions queue index 6 / end
3583   Flow rule #0 created
3584   testpmd> flow create 0 ingress pattern eth / ipv6 / end
3585      actions queue index 2 / end
3586   Flow rule #1 created
3587   testpmd> flow create 0 priority 5 ingress pattern eth / ipv4 / udp / end
3588      actions rss queues 6 7 8 end / end
3589   Flow rule #2 created
3590   testpmd> flow list 0
3591   ID      Group   Prio    Attr    Rule
3592   0       0       0       i-      ETH IPV4 => QUEUE
3593   1       0       0       i-      ETH IPV6 => QUEUE
3594   2       0       5       i-      ETH IPV4 UDP => RSS
3595   testpmd>
3596
3597Rules are sorted by priority (i.e. group ID first, then priority level)::
3598
3599   testpmd> flow list 1
3600   ID      Group   Prio    Attr    Rule
3601   0       0       0       i-      ETH => COUNT
3602   6       0       500     i-      ETH IPV6 TCP => DROP COUNT
3603   5       0       1000    i-      ETH IPV6 ICMP => QUEUE
3604   1       24      0       i-      ETH IPV4 UDP => QUEUE
3605   4       24      10      i-      ETH IPV4 TCP => DROP
3606   3       24      20      i-      ETH IPV4 => DROP
3607   2       24      42      i-      ETH IPV4 UDP => QUEUE
3608   7       63      0       i-      ETH IPV6 UDP VXLAN => MARK QUEUE
3609   testpmd>
3610
3611Output can be limited to specific groups::
3612
3613   testpmd> flow list 1 group 0 group 63
3614   ID      Group   Prio    Attr    Rule
3615   0       0       0       i-      ETH => COUNT
3616   6       0       500     i-      ETH IPV6 TCP => DROP COUNT
3617   5       0       1000    i-      ETH IPV6 ICMP => QUEUE
3618   7       63      0       i-      ETH IPV6 UDP VXLAN => MARK QUEUE
3619   testpmd>
3620
3621Toggling isolated mode
3622~~~~~~~~~~~~~~~~~~~~~~
3623
3624``flow isolate`` can be used to tell the underlying PMD that ingress traffic
3625must only be injected from the defined flow rules; that no default traffic
3626is expected outside those rules and the driver is free to assign more
3627resources to handle them. It is bound to ``rte_flow_isolate()``::
3628
3629 flow isolate {port_id} {boolean}
3630
3631If successful, enabling or disabling isolated mode shows either::
3632
3633 Ingress traffic on port [...]
3634    is now restricted to the defined flow rules
3635
3636Or::
3637
3638 Ingress traffic on port [...]
3639    is not restricted anymore to the defined flow rules
3640
3641Otherwise, in case of error::
3642
3643   Caught error type [...] ([...]): [...]
3644
3645Mainly due to its side effects, PMDs supporting this mode may not have the
3646ability to toggle it more than once without reinitializing affected ports
3647first (e.g. by exiting testpmd).
3648
3649Enabling isolated mode::
3650
3651 testpmd> flow isolate 0 true
3652 Ingress traffic on port 0 is now restricted to the defined flow rules
3653 testpmd>
3654
3655Disabling isolated mode::
3656
3657 testpmd> flow isolate 0 false
3658 Ingress traffic on port 0 is not restricted anymore to the defined flow rules
3659 testpmd>
3660
3661Sample QinQ flow rules
3662~~~~~~~~~~~~~~~~~~~~~~
3663
3664Before creating QinQ rule(s) the following commands should be issued to enable QinQ::
3665
3666   testpmd> port stop 0
3667   testpmd> vlan set qinq on 0
3668
3669The above command sets the inner and outer TPID's to 0x8100.
3670
3671To change the TPID's the following commands should be used::
3672
3673   testpmd> vlan set outer tpid 0xa100 0
3674   testpmd> vlan set inner tpid 0x9100 0
3675   testpmd> port start 0
3676
3677Validate and create a QinQ rule on port 0 to steer traffic to a VF queue in a VM.
3678
3679::
3680
3681   testpmd> flow validate 0 ingress pattern eth / vlan tci is 123 /
3682       vlan tci is 456 / end actions vf id 1 / queue index 0 / end
3683   Flow rule #0 validated
3684
3685   testpmd> flow create 0 ingress pattern eth / vlan tci is 4 /
3686       vlan tci is 456 / end actions vf id 123 / queue index 0 / end
3687   Flow rule #0 created
3688
3689   testpmd> flow list 0
3690   ID      Group   Prio    Attr    Rule
3691   0       0       0       i-      ETH VLAN VLAN=>VF QUEUE
3692
3693Validate and create a QinQ rule on port 0 to steer traffic to a queue on the host.
3694
3695::
3696
3697   testpmd> flow validate 0 ingress pattern eth / vlan tci is 321 /
3698        vlan tci is 654 / end actions pf / queue index 0 / end
3699   Flow rule #1 validated
3700
3701   testpmd> flow create 0 ingress pattern eth / vlan tci is 321 /
3702        vlan tci is 654 / end actions pf / queue index 1 / end
3703   Flow rule #1 created
3704
3705   testpmd> flow list 0
3706   ID      Group   Prio    Attr    Rule
3707   0       0       0       i-      ETH VLAN VLAN=>VF QUEUE
3708   1       0       0       i-      ETH VLAN VLAN=>PF QUEUE
3709