blob: b0941186f867710a4f4ca4a9fdb486aa2d9b1455 [file] [log] [blame]
Linus Torvalds1da177e2005-04-16 15:20:36 -07001#
2# Traffic control configuration.
3#
4choice
5 prompt "Packet scheduler clock source"
6 depends on NET_SCHED
7 default NET_SCH_CLK_JIFFIES
8 help
9 Packet schedulers need a monotonic clock that increments at a static
10 rate. The kernel provides several suitable interfaces, each with
11 different properties:
12
13 - high resolution (us or better)
14 - fast to read (minimal locking, no i/o access)
15 - synchronized on all processors
16 - handles cpu clock frequency changes
17
18 but nothing provides all of the above.
19
20config NET_SCH_CLK_JIFFIES
21 bool "Timer interrupt"
22 help
23 Say Y here if you want to use the timer interrupt (jiffies) as clock
24 source. This clock source is fast, synchronized on all processors and
25 handles cpu clock frequency changes, but its resolution is too low
26 for accurate shaping except at very low speed.
27
28config NET_SCH_CLK_GETTIMEOFDAY
29 bool "gettimeofday"
30 help
31 Say Y here if you want to use gettimeofday as clock source. This clock
32 source has high resolution, is synchronized on all processors and
33 handles cpu clock frequency changes, but it is slow.
34
35 Choose this if you need a high resolution clock source but can't use
36 the CPU's cycle counter.
37
38config NET_SCH_CLK_CPU
39 bool "CPU cycle counter"
40 depends on X86_TSC || X86_64 || ALPHA || SPARC64 || PPC64 || IA64
41 help
42 Say Y here if you want to use the CPU's cycle counter as clock source.
43 This is a cheap and high resolution clock source, but on some
44 architectures it is not synchronized on all processors and doesn't
45 handle cpu clock frequency changes.
46
47 The useable cycle counters are:
48
49 x86/x86_64 - Timestamp Counter
50 alpha - Cycle Counter
51 sparc64 - %ticks register
52 ppc64 - Time base
53 ia64 - Interval Time Counter
54
55 Choose this if your CPU's cycle counter is working properly.
56
57endchoice
58
59config NET_SCH_CBQ
60 tristate "CBQ packet scheduler"
61 depends on NET_SCHED
62 ---help---
63 Say Y here if you want to use the Class-Based Queueing (CBQ) packet
64 scheduling algorithm for some of your network devices. This
65 algorithm classifies the waiting packets into a tree-like hierarchy
66 of classes; the leaves of this tree are in turn scheduled by
67 separate algorithms (called "disciplines" in this context).
68
69 See the top of <file:net/sched/sch_cbq.c> for references about the
70 CBQ algorithm.
71
72 CBQ is a commonly used scheduler, so if you're unsure, you should
73 say Y here. Then say Y to all the queueing algorithms below that you
74 want to use as CBQ disciplines. Then say Y to "Packet classifier
75 API" and say Y to all the classifiers you want to use; a classifier
76 is a routine that allows you to sort your outgoing traffic into
77 classes based on a certain criterion.
78
79 To compile this code as a module, choose M here: the
80 module will be called sch_cbq.
81
82config NET_SCH_HTB
83 tristate "HTB packet scheduler"
84 depends on NET_SCHED
85 ---help---
86 Say Y here if you want to use the Hierarchical Token Buckets (HTB)
87 packet scheduling algorithm for some of your network devices. See
88 <http://luxik.cdi.cz/~devik/qos/htb/> for complete manual and
89 in-depth articles.
90
91 HTB is very similar to the CBQ regarding its goals however is has
92 different properties and different algorithm.
93
94 To compile this code as a module, choose M here: the
95 module will be called sch_htb.
96
97config NET_SCH_HFSC
98 tristate "HFSC packet scheduler"
99 depends on NET_SCHED
100 ---help---
101 Say Y here if you want to use the Hierarchical Fair Service Curve
102 (HFSC) packet scheduling algorithm for some of your network devices.
103
104 To compile this code as a module, choose M here: the
105 module will be called sch_hfsc.
106
107#tristate ' H-PFQ packet scheduler' CONFIG_NET_SCH_HPFQ
108config NET_SCH_ATM
109 tristate "ATM pseudo-scheduler"
110 depends on NET_SCHED && ATM
111 ---help---
112 Say Y here if you want to use the ATM pseudo-scheduler. This
113 provides a framework for invoking classifiers (aka "filters"), which
114 in turn select classes of this queuing discipline. Each class maps
115 the flow(s) it is handling to a given virtual circuit (see the top of
116 <file:net/sched/sch_atm.c>).
117
118 To compile this code as a module, choose M here: the
119 module will be called sch_atm.
120
121config NET_SCH_PRIO
122 tristate "The simplest PRIO pseudoscheduler"
123 depends on NET_SCHED
124 help
125 Say Y here if you want to use an n-band priority queue packet
126 "scheduler" for some of your network devices or as a leaf discipline
127 for the CBQ scheduling algorithm. If unsure, say Y.
128
129 To compile this code as a module, choose M here: the
130 module will be called sch_prio.
131
132config NET_SCH_RED
133 tristate "RED queue"
134 depends on NET_SCHED
135 help
136 Say Y here if you want to use the Random Early Detection (RED)
137 packet scheduling algorithm for some of your network devices (see
138 the top of <file:net/sched/sch_red.c> for details and references
139 about the algorithm).
140
141 To compile this code as a module, choose M here: the
142 module will be called sch_red.
143
144config NET_SCH_SFQ
145 tristate "SFQ queue"
146 depends on NET_SCHED
147 ---help---
148 Say Y here if you want to use the Stochastic Fairness Queueing (SFQ)
149 packet scheduling algorithm for some of your network devices or as a
150 leaf discipline for the CBQ scheduling algorithm (see the top of
151 <file:net/sched/sch_sfq.c> for details and references about the SFQ
152 algorithm).
153
154 To compile this code as a module, choose M here: the
155 module will be called sch_sfq.
156
157config NET_SCH_TEQL
158 tristate "TEQL queue"
159 depends on NET_SCHED
160 ---help---
161 Say Y here if you want to use the True Link Equalizer (TLE) packet
162 scheduling algorithm for some of your network devices or as a leaf
163 discipline for the CBQ scheduling algorithm. This queueing
164 discipline allows the combination of several physical devices into
165 one virtual device. (see the top of <file:net/sched/sch_teql.c> for
166 details).
167
168 To compile this code as a module, choose M here: the
169 module will be called sch_teql.
170
171config NET_SCH_TBF
172 tristate "TBF queue"
173 depends on NET_SCHED
174 help
175 Say Y here if you want to use the Simple Token Bucket Filter (TBF)
176 packet scheduling algorithm for some of your network devices or as a
177 leaf discipline for the CBQ scheduling algorithm (see the top of
178 <file:net/sched/sch_tbf.c> for a description of the TBF algorithm).
179
180 To compile this code as a module, choose M here: the
181 module will be called sch_tbf.
182
183config NET_SCH_GRED
184 tristate "GRED queue"
185 depends on NET_SCHED
186 help
187 Say Y here if you want to use the Generic Random Early Detection
Lucas Correia Villa Real20cc6be2005-05-03 14:34:20 -0700188 (GRED) packet scheduling algorithm for some of your network devices
Linus Torvalds1da177e2005-04-16 15:20:36 -0700189 (see the top of <file:net/sched/sch_red.c> for details and
190 references about the algorithm).
191
192 To compile this code as a module, choose M here: the
193 module will be called sch_gred.
194
195config NET_SCH_DSMARK
196 tristate "Diffserv field marker"
197 depends on NET_SCHED
198 help
199 Say Y if you want to schedule packets according to the
200 Differentiated Services architecture proposed in RFC 2475.
201 Technical information on this method, with pointers to associated
202 RFCs, is available at <http://www.gta.ufrj.br/diffserv/>.
203
204 To compile this code as a module, choose M here: the
205 module will be called sch_dsmark.
206
207config NET_SCH_NETEM
208 tristate "Network emulator"
209 depends on NET_SCHED
210 help
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.
214
215 To compile this driver as a module, choose M here: the module
216 will be called sch_netem.
217
218 If unsure, say N.
219
220config NET_SCH_INGRESS
221 tristate "Ingress Qdisc"
222 depends on NET_SCHED
223 help
224 If you say Y here, you will be able to police incoming bandwidth
225 and drop packets when this bandwidth exceeds your desired rate.
226 If unsure, say Y.
227
228 To compile this code as a module, choose M here: the
229 module will be called sch_ingress.
230
231config NET_QOS
232 bool "QoS support"
233 depends on NET_SCHED
234 ---help---
235 Say Y here if you want to include Quality Of Service scheduling
236 features, which means that you will be able to request certain
237 rate-of-flow limits for your network devices.
238
239 This Quality of Service (QoS) support will enable you to use
240 Differentiated Services (diffserv) and Resource Reservation Protocol
241 (RSVP) on your Linux router if you also say Y to "Packet classifier
242 API" and to some classifiers below. Documentation and software is at
243 <http://diffserv.sourceforge.net/>.
244
245 Note that the answer to this question won't directly affect the
246 kernel: saying N will just cause the configurator to skip all
247 the questions about QoS support.
248
249config NET_ESTIMATOR
250 bool "Rate estimator"
251 depends on NET_QOS
252 help
253 In order for Quality of Service scheduling to work, the current
254 rate-of-flow for a network device has to be estimated; if you say Y
255 here, the kernel will do just that.
256
257config NET_CLS
258 bool "Packet classifier API"
259 depends on NET_SCHED
260 ---help---
261 The CBQ scheduling algorithm requires that network packets which are
262 scheduled to be sent out over a network device be classified
263 according to some criterion. If you say Y here, you will get a
264 choice of several different packet classifiers with the following
265 questions.
266
267 This will enable you to use Differentiated Services (diffserv) and
268 Resource Reservation Protocol (RSVP) on your Linux router.
269 Documentation and software is at
270 <http://diffserv.sourceforge.net/>.
271
272config NET_CLS_BASIC
273 tristate "Basic classifier"
274 depends on NET_CLS
275 ---help---
276 Say Y here if you want to be able to classify packets using
277 only extended matches and actions.
278
279 To compile this code as a module, choose M here: the
280 module will be called cls_basic.
281
282config NET_CLS_TCINDEX
283 tristate "TC index classifier"
284 depends on NET_CLS
285 help
286 If you say Y here, you will be able to classify outgoing packets
287 according to the tc_index field of the skb. You will want this
288 feature if you want to implement Differentiated Services using
289 sch_dsmark. If unsure, say Y.
290
291 To compile this code as a module, choose M here: the
292 module will be called cls_tcindex.
293
294config NET_CLS_ROUTE4
295 tristate "Routing table based classifier"
296 depends on NET_CLS
297 select NET_CLS_ROUTE
298 help
299 If you say Y here, you will be able to classify outgoing packets
300 according to the route table entry they matched. If unsure, say Y.
301
302 To compile this code as a module, choose M here: the
303 module will be called cls_route.
304
305config NET_CLS_ROUTE
306 bool
307 default n
308
309config NET_CLS_FW
310 tristate "Firewall based classifier"
311 depends on NET_CLS
312 help
313 If you say Y here, you will be able to classify outgoing packets
314 according to firewall criteria you specified.
315
316 To compile this code as a module, choose M here: the
317 module will be called cls_fw.
318
319config NET_CLS_U32
320 tristate "U32 classifier"
321 depends on NET_CLS
322 help
323 If you say Y here, you will be able to classify outgoing packets
324 according to their destination address. If unsure, say Y.
325
326 To compile this code as a module, choose M here: the
327 module will be called cls_u32.
328
329config CLS_U32_PERF
330 bool "U32 classifier performance counters"
331 depends on NET_CLS_U32
332 help
333 gathers stats that could be used to tune u32 classifier performance.
334 Requires a new iproute2
335 You MUST NOT turn this on if you dont have an update iproute2.
336
337config NET_CLS_IND
338 bool "classify input device (slows things u32/fw) "
339 depends on NET_CLS_U32 || NET_CLS_FW
340 help
341 This option will be killed eventually when a
342 metadata action appears because it slows things a little
343 Available only for u32 and fw classifiers.
344 Requires a new iproute2
345 You MUST NOT turn this on if you dont have an update iproute2.
346
347config CLS_U32_MARK
348 bool "Use nfmark as a key in U32 classifier"
349 depends on NET_CLS_U32 && NETFILTER
350 help
351 This allows you to match mark in a u32 filter.
352 Example:
353 tc filter add dev eth0 protocol ip parent 1:0 prio 5 u32 \
354 match mark 0x0090 0xffff \
355 match ip dst 4.4.4.4 \
356 flowid 1:90
357 You must use a new iproute2 to use this feature.
358
359config NET_CLS_RSVP
360 tristate "Special RSVP classifier"
361 depends on NET_CLS && NET_QOS
362 ---help---
363 The Resource Reservation Protocol (RSVP) permits end systems to
364 request a minimum and maximum data flow rate for a connection; this
365 is important for real time data such as streaming sound or video.
366
367 Say Y here if you want to be able to classify outgoing packets based
368 on their RSVP requests.
369
370 To compile this code as a module, choose M here: the
371 module will be called cls_rsvp.
372
373config NET_CLS_RSVP6
374 tristate "Special RSVP classifier for IPv6"
375 depends on NET_CLS && NET_QOS
376 ---help---
377 The Resource Reservation Protocol (RSVP) permits end systems to
378 request a minimum and maximum data flow rate for a connection; this
379 is important for real time data such as streaming sound or video.
380
381 Say Y here if you want to be able to classify outgoing packets based
382 on their RSVP requests and you are using the new Internet Protocol
383 IPv6 as opposed to the older and more common IPv4.
384
385 To compile this code as a module, choose M here: the
386 module will be called cls_rsvp6.
387
388config NET_EMATCH
389 bool "Extended Matches"
390 depends on NET_CLS
391 ---help---
392 Say Y here if you want to use extended matches on top of classifiers
393 and select the extended matches below.
394
395 Extended matches are small classification helpers not worth writing
396 a separate classifier.
397
398 You must have a recent version of the iproute2 tools in order to use
399 extended matches.
400
401config NET_EMATCH_STACK
402 int "Stack size"
403 depends on NET_EMATCH
404 default "32"
405 ---help---
406 Size of the local stack variable used while evaluating the tree of
407 ematches. Limits the depth of the tree, i.e. the number of
408 encapsulated precedences. Every level requires 4 bytes of addtional
409 stack space.
410
411config NET_EMATCH_CMP
412 tristate "Simple packet data comparison"
413 depends on NET_EMATCH
414 ---help---
415 Say Y here if you want to be able to classify packets based on
416 simple packet data comparisons for 8, 16, and 32bit values.
417
418 To compile this code as a module, choose M here: the
419 module will be called em_cmp.
420
421config NET_EMATCH_NBYTE
422 tristate "Multi byte comparison"
423 depends on NET_EMATCH
424 ---help---
425 Say Y here if you want to be able to classify packets based on
426 multiple byte comparisons mainly useful for IPv6 address comparisons.
427
428 To compile this code as a module, choose M here: the
429 module will be called em_nbyte.
430
431config NET_EMATCH_U32
432 tristate "U32 hashing key"
433 depends on NET_EMATCH
434 ---help---
435 Say Y here if you want to be able to classify packets using
436 the famous u32 key in combination with logic relations.
437
438 To compile this code as a module, choose M here: the
439 module will be called em_u32.
440
441config NET_EMATCH_META
442 tristate "Metadata"
443 depends on NET_EMATCH
444 ---help---
445 Say Y here if you want to be ablt to classify packets based on
446 metadata such as load average, netfilter attributes, socket
447 attributes and routing decisions.
448
449 To compile this code as a module, choose M here: the
450 module will be called em_meta.
451
452config NET_CLS_ACT
453 bool "Packet ACTION"
454 depends on EXPERIMENTAL && NET_CLS && NET_QOS
455 ---help---
456 This option requires you have a new iproute2. It enables
457 tc extensions which can be used with tc classifiers.
458 You MUST NOT turn this on if you dont have an update iproute2.
459
460config NET_ACT_POLICE
461 tristate "Policing Actions"
462 depends on NET_CLS_ACT
463 ---help---
464 If you are using a newer iproute2 select this one, otherwise use one
465 below to select a policer.
466 You MUST NOT turn this on if you dont have an update iproute2.
467
468config NET_ACT_GACT
469 tristate "generic Actions"
470 depends on NET_CLS_ACT
471 ---help---
472 You must have new iproute2 to use this feature.
473 This adds simple filtering actions like drop, accept etc.
474
475config GACT_PROB
476 bool "generic Actions probability"
477 depends on NET_ACT_GACT
478 ---help---
479 Allows generic actions to be randomly or deterministically used.
480
481config NET_ACT_MIRRED
482 tristate "Packet In/Egress redirecton/mirror Actions"
483 depends on NET_CLS_ACT
484 ---help---
485 requires new iproute2
486 This allows packets to be mirrored or redirected to netdevices
487
488config NET_ACT_IPT
489 tristate "iptables Actions"
490 depends on NET_CLS_ACT && NETFILTER && IP_NF_IPTABLES
491 ---help---
492 requires new iproute2
493 This allows iptables targets to be used by tc filters
494
495config NET_ACT_PEDIT
496 tristate "Generic Packet Editor Actions"
497 depends on NET_CLS_ACT
498 ---help---
499 requires new iproute2
500 This allows for packets to be generically edited
501
502config NET_CLS_POLICE
503 bool "Traffic policing (needed for in/egress)"
504 depends on NET_CLS && NET_QOS && NET_CLS_ACT!=y
505 help
506 Say Y to support traffic policing (bandwidth limits). Needed for
507 ingress and egress rate limiting.
508
Jamal Hadi Salimdb753072005-04-24 20:10:16 -0700509config NET_ACT_SIMP
510 tristate "Simple action"
511 depends on NET_CLS_ACT
512 ---help---
513 You must have new iproute2 to use this feature.
514 This adds a very simple action for demonstration purposes
515 The idea is to give action authors a basic example to look at.
516 All this action will do is print on the console the configured
517 policy string followed by _ then packet count.
518