2 # Traffic control configuration.
6 bool "QoS and/or fair queueing"
9 When the kernel has several packets to send out over a network
10 device, it has to decide which ones to send first, which ones to
11 delay, and which ones to drop. This is the job of the queueing
12 disciplines, several different algorithms for how to do this
13 "fairly" have been proposed.
15 If you say N here, you will get the standard packet scheduler, which
16 is a FIFO (first come, first served). If you say Y here, you will be
17 able to choose from among several alternative algorithms which can
18 then be attached to different network devices. This is useful for
19 example if some of your network devices are real time devices that
20 need a certain minimum data flow rate, or if you need to limit the
21 maximum data flow rate for traffic which matches specified criteria.
22 This code is considered to be experimental.
24 To administer these schedulers, you'll need the user-level utilities
25 from the package iproute2+tc at
26 <https://www.kernel.org/pub/linux/utils/net/iproute2/>. That package
27 also contains some documentation; for more, check out
28 <http://www.linuxfoundation.org/collaborate/workgroups/networking/iproute2>.
30 This Quality of Service (QoS) support will enable you to use
31 Differentiated Services (diffserv) and Resource Reservation Protocol
32 (RSVP) on your Linux router if you also say Y to the corresponding
33 classifiers below. Documentation and software is at
34 <http://diffserv.sourceforge.net/>.
36 If you say Y here and to "/proc file system" below, you will be able
37 to read status information about packet schedulers from the file
40 The available schedulers are listed in the following questions; you
41 can say Y to as many as you like. If unsure, say N now.
45 comment "Queueing/Scheduling"
48 tristate "Class Based Queueing (CBQ)"
50 Say Y here if you want to use the Class-Based Queueing (CBQ) packet
51 scheduling algorithm. This algorithm classifies the waiting packets
52 into a tree-like hierarchy of classes; the leaves of this tree are
53 in turn scheduled by separate algorithms.
55 See the top of <file:net/sched/sch_cbq.c> for more details.
57 CBQ is a commonly used scheduler, so if you're unsure, you should
58 say Y here. Then say Y to all the queueing algorithms below that you
59 want to use as leaf disciplines.
61 To compile this code as a module, choose M here: the
62 module will be called sch_cbq.
65 tristate "Hierarchical Token Bucket (HTB)"
67 Say Y here if you want to use the Hierarchical Token Buckets (HTB)
68 packet scheduling algorithm. See
69 <http://luxik.cdi.cz/~devik/qos/htb/> for complete manual and
72 HTB is very similar to CBQ regarding its goals however is has
73 different properties and different algorithm.
75 To compile this code as a module, choose M here: the
76 module will be called sch_htb.
79 tristate "Hierarchical Fair Service Curve (HFSC)"
81 Say Y here if you want to use the Hierarchical Fair Service Curve
82 (HFSC) packet scheduling algorithm.
84 To compile this code as a module, choose M here: the
85 module will be called sch_hfsc.
88 tristate "ATM Virtual Circuits (ATM)"
91 Say Y here if you want to use the ATM pseudo-scheduler. This
92 provides a framework for invoking classifiers, which in turn
93 select classes of this queuing discipline. Each class maps
94 the flow(s) it is handling to a given virtual circuit.
96 See the top of <file:net/sched/sch_atm.c> for more details.
98 To compile this code as a module, choose M here: the
99 module will be called sch_atm.
102 tristate "Multi Band Priority Queueing (PRIO)"
104 Say Y here if you want to use an n-band priority queue packet
107 To compile this code as a module, choose M here: the
108 module will be called sch_prio.
110 config NET_SCH_MULTIQ
111 tristate "Hardware Multiqueue-aware Multi Band Queuing (MULTIQ)"
113 Say Y here if you want to use an n-band queue packet scheduler
114 to support devices that have multiple hardware transmit queues.
116 To compile this code as a module, choose M here: the
117 module will be called sch_multiq.
120 tristate "Random Early Detection (RED)"
122 Say Y here if you want to use the Random Early Detection (RED)
123 packet scheduling algorithm.
125 See the top of <file:net/sched/sch_red.c> for more details.
127 To compile this code as a module, choose M here: the
128 module will be called sch_red.
131 tristate "Stochastic Fair Blue (SFB)"
133 Say Y here if you want to use the Stochastic Fair Blue (SFB)
134 packet scheduling algorithm.
136 See the top of <file:net/sched/sch_sfb.c> for more details.
138 To compile this code as a module, choose M here: the
139 module will be called sch_sfb.
142 tristate "Stochastic Fairness Queueing (SFQ)"
144 Say Y here if you want to use the Stochastic Fairness Queueing (SFQ)
145 packet scheduling algorithm.
147 See the top of <file:net/sched/sch_sfq.c> for more details.
149 To compile this code as a module, choose M here: the
150 module will be called sch_sfq.
153 tristate "True Link Equalizer (TEQL)"
155 Say Y here if you want to use the True Link Equalizer (TLE) packet
156 scheduling algorithm. This queueing discipline allows the combination
157 of several physical devices into one virtual device.
159 See the top of <file:net/sched/sch_teql.c> for more details.
161 To compile this code as a module, choose M here: the
162 module will be called sch_teql.
165 tristate "Token Bucket Filter (TBF)"
167 Say Y here if you want to use the Token Bucket Filter (TBF) packet
168 scheduling algorithm.
170 See the top of <file:net/sched/sch_tbf.c> for more details.
172 To compile this code as a module, choose M here: the
173 module will be called sch_tbf.
176 tristate "Credit Based Shaper (CBS)"
178 Say Y here if you want to use the Credit Based Shaper (CBS) packet
179 scheduling algorithm.
181 See the top of <file:net/sched/sch_cbs.c> for more details.
183 To compile this code as a module, choose M here: the
184 module will be called sch_cbs.
187 tristate "Generic Random Early Detection (GRED)"
189 Say Y here if you want to use the Generic Random Early Detection
190 (GRED) packet scheduling algorithm for some of your network devices
191 (see the top of <file:net/sched/sch_red.c> for details and
192 references about the algorithm).
194 To compile this code as a module, choose M here: the
195 module will be called sch_gred.
197 config NET_SCH_DSMARK
198 tristate "Differentiated Services marker (DSMARK)"
200 Say Y if you want to schedule packets according to the
201 Differentiated Services architecture proposed in RFC 2475.
202 Technical information on this method, with pointers to associated
203 RFCs, is available at <http://www.gta.ufrj.br/diffserv/>.
205 To compile this code as a module, choose M here: the
206 module will be called sch_dsmark.
209 tristate "Network emulator (NETEM)"
211 Say Y if you want to emulate network delay, loss, and packet
212 re-ordering. This is often useful to simulate networks when
213 testing applications or protocols.
215 To compile this driver as a module, choose M here: the module
216 will be called sch_netem.
221 tristate "Deficit Round Robin scheduler (DRR)"
223 Say Y here if you want to use the Deficit Round Robin (DRR) packet
224 scheduling algorithm.
226 To compile this driver as a module, choose M here: the module
227 will be called sch_drr.
231 config NET_SCH_MQPRIO
232 tristate "Multi-queue priority scheduler (MQPRIO)"
234 Say Y here if you want to use the Multi-queue Priority scheduler.
235 This scheduler allows QOS to be offloaded on NICs that have support
236 for offloading QOS schedulers.
238 To compile this driver as a module, choose M here: the module will
239 be called sch_mqprio.
244 tristate "CHOose and Keep responsive flow scheduler (CHOKE)"
246 Say Y here if you want to use the CHOKe packet scheduler (CHOose
247 and Keep for responsive flows, CHOose and Kill for unresponsive
248 flows). This is a variation of RED which trys to penalize flows
249 that monopolize the queue.
251 To compile this code as a module, choose M here: the
252 module will be called sch_choke.
255 tristate "Quick Fair Queueing scheduler (QFQ)"
257 Say Y here if you want to use the Quick Fair Queueing Scheduler (QFQ)
258 packet scheduling algorithm.
260 To compile this driver as a module, choose M here: the module
261 will be called sch_qfq.
266 tristate "Controlled Delay AQM (CODEL)"
268 Say Y here if you want to use the Controlled Delay (CODEL)
269 packet scheduling algorithm.
271 To compile this driver as a module, choose M here: the module
272 will be called sch_codel.
276 config NET_SCH_FQ_CODEL
277 tristate "Fair Queue Controlled Delay AQM (FQ_CODEL)"
279 Say Y here if you want to use the FQ Controlled Delay (FQ_CODEL)
280 packet scheduling algorithm.
282 To compile this driver as a module, choose M here: the module
283 will be called sch_fq_codel.
288 tristate "Fair Queue"
290 Say Y here if you want to use the FQ packet scheduling algorithm.
292 FQ does flow separation, and is able to respect pacing requirements
293 set by TCP stack into sk->sk_pacing_rate (for localy generated
296 To compile this driver as a module, choose M here: the module
297 will be called sch_fq.
302 tristate "Heavy-Hitter Filter (HHF)"
304 Say Y here if you want to use the Heavy-Hitter Filter (HHF)
305 packet scheduling algorithm.
307 To compile this driver as a module, choose M here: the module
308 will be called sch_hhf.
311 tristate "Proportional Integral controller Enhanced (PIE) scheduler"
313 Say Y here if you want to use the Proportional Integral controller
314 Enhanced scheduler packet scheduling algorithm.
315 For more information, please see
316 http://tools.ietf.org/html/draft-pan-tsvwg-pie-00
318 To compile this driver as a module, choose M here: the module
319 will be called sch_pie.
323 config NET_SCH_INGRESS
324 tristate "Ingress/classifier-action Qdisc"
325 depends on NET_CLS_ACT
329 Say Y here if you want to use classifiers for incoming and/or outgoing
330 packets. This qdisc doesn't do anything else besides running classifiers,
331 which can also have actions attached to them. In case of outgoing packets,
332 classifiers that this qdisc holds are executed in the transmit path
333 before real enqueuing to an egress qdisc happens.
337 To compile this code as a module, choose M here: the module will be
338 called sch_ingress with alias of sch_clsact.
341 tristate "Plug network traffic until release (PLUG)"
344 This queuing discipline allows userspace to plug/unplug a network
345 output queue, using the netlink interface. When it receives an
346 enqueue command it inserts a plug into the outbound queue that
347 causes following packets to enqueue until a dequeue command arrives
348 over netlink, causing the plug to be removed and resuming the normal
351 This module also provides a generic "network output buffering"
352 functionality (aka output commit), wherein upon arrival of a dequeue
353 command, only packets up to the first plug are released for delivery.
354 The Remus HA project uses this module to enable speculative execution
355 of virtual machines by allowing the generated network output to be rolled
358 For more information, please refer to <http://wiki.xenproject.org/wiki/Remus>
360 Say Y here if you are using this kernel for Xen dom0 and
361 want to protect Xen guests with Remus.
363 To compile this code as a module, choose M here: the
364 module will be called sch_plug.
366 menuconfig NET_SCH_DEFAULT
367 bool "Allow override default queue discipline"
369 Support for selection of default queuing discipline.
371 Nearly all users can safely say no here, and the default
372 of pfifo_fast will be used. Many distributions already set
373 the default value via /proc/sys/net/core/default_qdisc.
380 prompt "Default queuing discipline"
381 default DEFAULT_PFIFO_FAST
383 Select the queueing discipline that will be used by default
384 for all network devices.
387 bool "Fair Queue" if NET_SCH_FQ
390 bool "Controlled Delay" if NET_SCH_CODEL
392 config DEFAULT_FQ_CODEL
393 bool "Fair Queue Controlled Delay" if NET_SCH_FQ_CODEL
396 bool "Stochastic Fair Queue" if NET_SCH_SFQ
398 config DEFAULT_PFIFO_FAST
399 bool "Priority FIFO Fast"
402 config DEFAULT_NET_SCH
404 default "pfifo_fast" if DEFAULT_PFIFO_FAST
405 default "fq" if DEFAULT_FQ
406 default "fq_codel" if DEFAULT_FQ_CODEL
407 default "sfq" if DEFAULT_SFQ
411 comment "Classification"
417 tristate "Elementary classification (BASIC)"
420 Say Y here if you want to be able to classify packets using
421 only extended matches and actions.
423 To compile this code as a module, choose M here: the
424 module will be called cls_basic.
426 config NET_CLS_TCINDEX
427 tristate "Traffic-Control Index (TCINDEX)"
430 Say Y here if you want to be able to classify packets based on
431 traffic control indices. You will want this feature if you want
432 to implement Differentiated Services together with DSMARK.
434 To compile this code as a module, choose M here: the
435 module will be called cls_tcindex.
437 config NET_CLS_ROUTE4
438 tristate "Routing decision (ROUTE)"
440 select IP_ROUTE_CLASSID
443 If you say Y here, you will be able to classify packets
444 according to the route table entry they matched.
446 To compile this code as a module, choose M here: the
447 module will be called cls_route.
450 tristate "Netfilter mark (FW)"
453 If you say Y here, you will be able to classify packets
454 according to netfilter/firewall marks.
456 To compile this code as a module, choose M here: the
457 module will be called cls_fw.
460 tristate "Universal 32bit comparisons w/ hashing (U32)"
463 Say Y here to be able to classify packets using a universal
464 32bit pieces based comparison scheme.
466 To compile this code as a module, choose M here: the
467 module will be called cls_u32.
470 bool "Performance counters support"
471 depends on NET_CLS_U32
473 Say Y here to make u32 gather additional statistics useful for
474 fine tuning u32 classifiers.
477 bool "Netfilter marks support"
478 depends on NET_CLS_U32
480 Say Y here to be able to use netfilter marks as u32 key.
483 tristate "IPv4 Resource Reservation Protocol (RSVP)"
486 The Resource Reservation Protocol (RSVP) permits end systems to
487 request a minimum and maximum data flow rate for a connection; this
488 is important for real time data such as streaming sound or video.
490 Say Y here if you want to be able to classify outgoing packets based
491 on their RSVP requests.
493 To compile this code as a module, choose M here: the
494 module will be called cls_rsvp.
497 tristate "IPv6 Resource Reservation Protocol (RSVP6)"
500 The Resource Reservation Protocol (RSVP) permits end systems to
501 request a minimum and maximum data flow rate for a connection; this
502 is important for real time data such as streaming sound or video.
504 Say Y here if you want to be able to classify outgoing packets based
505 on their RSVP requests and you are using the IPv6 protocol.
507 To compile this code as a module, choose M here: the
508 module will be called cls_rsvp6.
511 tristate "Flow classifier"
514 If you say Y here, you will be able to classify packets based on
515 a configurable combination of packet keys. This is mostly useful
516 in combination with SFQ.
518 To compile this code as a module, choose M here: the
519 module will be called cls_flow.
521 config NET_CLS_CGROUP
522 tristate "Control Group Classifier"
524 select CGROUP_NET_CLASSID
527 Say Y here if you want to classify packets based on the control
528 cgroup of their process.
530 To compile this code as a module, choose M here: the
531 module will be called cls_cgroup.
534 tristate "BPF-based classifier"
537 If you say Y here, you will be able to classify packets based on
538 programmable BPF (JIT'ed) filters as an alternative to ematches.
540 To compile this code as a module, choose M here: the module will
543 config NET_CLS_FLOWER
544 tristate "Flower classifier"
547 If you say Y here, you will be able to classify packets based on
548 a configurable combination of packet keys and masks.
550 To compile this code as a module, choose M here: the module will
551 be called cls_flower.
553 config NET_CLS_MATCHALL
554 tristate "Match-all classifier"
557 If you say Y here, you will be able to classify packets based on
558 nothing. Every packet will match.
560 To compile this code as a module, choose M here: the module will
561 be called cls_matchall.
564 bool "Extended Matches"
567 Say Y here if you want to use extended matches on top of classifiers
568 and select the extended matches below.
570 Extended matches are small classification helpers not worth writing
571 a separate classifier for.
573 A recent version of the iproute2 package is required to use
576 config NET_EMATCH_STACK
578 depends on NET_EMATCH
581 Size of the local stack variable used while evaluating the tree of
582 ematches. Limits the depth of the tree, i.e. the number of
583 encapsulated precedences. Every level requires 4 bytes of additional
586 config NET_EMATCH_CMP
587 tristate "Simple packet data comparison"
588 depends on NET_EMATCH
590 Say Y here if you want to be able to classify packets based on
591 simple packet data comparisons for 8, 16, and 32bit values.
593 To compile this code as a module, choose M here: the
594 module will be called em_cmp.
596 config NET_EMATCH_NBYTE
597 tristate "Multi byte comparison"
598 depends on NET_EMATCH
600 Say Y here if you want to be able to classify packets based on
601 multiple byte comparisons mainly useful for IPv6 address comparisons.
603 To compile this code as a module, choose M here: the
604 module will be called em_nbyte.
606 config NET_EMATCH_U32
608 depends on NET_EMATCH
610 Say Y here if you want to be able to classify packets using
611 the famous u32 key in combination with logic relations.
613 To compile this code as a module, choose M here: the
614 module will be called em_u32.
616 config NET_EMATCH_META
618 depends on NET_EMATCH
620 Say Y here if you want to be able to classify packets based on
621 metadata such as load average, netfilter attributes, socket
622 attributes and routing decisions.
624 To compile this code as a module, choose M here: the
625 module will be called em_meta.
627 config NET_EMATCH_TEXT
628 tristate "Textsearch"
629 depends on NET_EMATCH
631 select TEXTSEARCH_KMP
633 select TEXTSEARCH_FSM
635 Say Y here if you want to be able to classify packets based on
636 textsearch comparisons.
638 To compile this code as a module, choose M here: the
639 module will be called em_text.
641 config NET_EMATCH_CANID
642 tristate "CAN Identifier"
643 depends on NET_EMATCH && (CAN=y || CAN=m)
645 Say Y here if you want to be able to classify CAN frames based
648 To compile this code as a module, choose M here: the
649 module will be called em_canid.
651 config NET_EMATCH_IPSET
653 depends on NET_EMATCH && IP_SET
655 Say Y here if you want to be able to classify packets based on
658 To compile this code as a module, choose M here: the
659 module will be called em_ipset.
665 Say Y here if you want to use traffic control actions. Actions
666 get attached to classifiers and are invoked after a successful
667 classification. They are used to overwrite the classification
668 result, instantly drop or redirect packets, etc.
670 A recent version of the iproute2 package is required to use
673 config NET_ACT_POLICE
674 tristate "Traffic Policing"
675 depends on NET_CLS_ACT
677 Say Y here if you want to do traffic policing, i.e. strict
678 bandwidth limiting. This action replaces the existing policing
681 To compile this code as a module, choose M here: the
682 module will be called act_police.
685 tristate "Generic actions"
686 depends on NET_CLS_ACT
688 Say Y here to take generic actions such as dropping and
691 To compile this code as a module, choose M here: the
692 module will be called act_gact.
695 bool "Probability support"
696 depends on NET_ACT_GACT
698 Say Y here to use the generic action randomly or deterministically.
700 config NET_ACT_MIRRED
701 tristate "Redirecting and Mirroring"
702 depends on NET_CLS_ACT
704 Say Y here to allow packets to be mirrored or redirected to
707 To compile this code as a module, choose M here: the
708 module will be called act_mirred.
710 config NET_ACT_SAMPLE
711 tristate "Traffic Sampling"
712 depends on NET_CLS_ACT
715 Say Y here to allow packet sampling tc action. The packet sample
716 action consists of statistically choosing packets and sampling
717 them using the psample module.
719 To compile this code as a module, choose M here: the
720 module will be called act_sample.
723 tristate "IPtables targets"
724 depends on NET_CLS_ACT && NETFILTER && IP_NF_IPTABLES
726 Say Y here to be able to invoke iptables targets after successful
729 To compile this code as a module, choose M here: the
730 module will be called act_ipt.
733 tristate "Stateless NAT"
734 depends on NET_CLS_ACT
736 Say Y here to do stateless NAT on IPv4 packets. You should use
737 netfilter for NAT unless you know what you are doing.
739 To compile this code as a module, choose M here: the
740 module will be called act_nat.
743 tristate "Packet Editing"
744 depends on NET_CLS_ACT
746 Say Y here if you want to mangle the content of packets.
748 To compile this code as a module, choose M here: the
749 module will be called act_pedit.
752 tristate "Simple Example (Debug)"
753 depends on NET_CLS_ACT
755 Say Y here to add a simple action for demonstration purposes.
756 It is meant as an example and for debugging purposes. It will
757 print a configured policy string followed by the packet count
758 to the console for every packet that passes by.
762 To compile this code as a module, choose M here: the
763 module will be called act_simple.
765 config NET_ACT_SKBEDIT
766 tristate "SKB Editing"
767 depends on NET_CLS_ACT
769 Say Y here to change skb priority or queue_mapping settings.
773 To compile this code as a module, choose M here: the
774 module will be called act_skbedit.
777 tristate "Checksum Updating"
778 depends on NET_CLS_ACT && INET
781 Say Y here to update some common checksum after some direct
784 To compile this code as a module, choose M here: the
785 module will be called act_csum.
788 tristate "Vlan manipulation"
789 depends on NET_CLS_ACT
791 Say Y here to push or pop vlan headers.
795 To compile this code as a module, choose M here: the
796 module will be called act_vlan.
799 tristate "BPF based action"
800 depends on NET_CLS_ACT
802 Say Y here to execute BPF code on packets. The BPF code will decide
803 if the packet should be dropped or not.
807 To compile this code as a module, choose M here: the
808 module will be called act_bpf.
810 config NET_ACT_CONNMARK
811 tristate "Netfilter Connection Mark Retriever"
812 depends on NET_CLS_ACT && NETFILTER && IP_NF_IPTABLES
813 depends on NF_CONNTRACK && NF_CONNTRACK_MARK
815 Say Y here to allow retrieving of conn mark
819 To compile this code as a module, choose M here: the
820 module will be called act_connmark.
822 config NET_ACT_SKBMOD
823 tristate "skb data modification action"
824 depends on NET_CLS_ACT
826 Say Y here to allow modification of skb data
830 To compile this code as a module, choose M here: the
831 module will be called act_skbmod.
834 tristate "Inter-FE action based on IETF ForCES InterFE LFB"
835 depends on NET_CLS_ACT
838 Say Y here to allow for sourcing and terminating metadata
839 For details refer to netdev01 paper:
840 "Distributing Linux Traffic Control Classifier-Action Subsystem"
841 Authors: Jamal Hadi Salim and Damascene M. Joachimpillai
843 To compile this code as a module, choose M here: the
844 module will be called act_ife.
846 config NET_ACT_TUNNEL_KEY
847 tristate "IP tunnel metadata manipulation"
848 depends on NET_CLS_ACT
850 Say Y here to set/release ip tunnel metadata.
854 To compile this code as a module, choose M here: the
855 module will be called act_tunnel_key.
857 config NET_IFE_SKBMARK
858 tristate "Support to encoding decoding skb mark on IFE action"
859 depends on NET_ACT_IFE
861 config NET_IFE_SKBPRIO
862 tristate "Support to encoding decoding skb prio on IFE action"
863 depends on NET_ACT_IFE
865 config NET_IFE_SKBTCINDEX
866 tristate "Support to encoding decoding skb tcindex on IFE action"
867 depends on NET_ACT_IFE
870 bool "Incoming device classification"
871 depends on NET_CLS_U32 || NET_CLS_FW
873 Say Y here to extend the u32 and fw classifier to support
874 classification based on the incoming device. This option is
875 likely to disappear in favour of the metadata ematch.