Loading...
1#
2# IP configuration
3#
4config IP_MULTICAST
5 bool "IP: multicasting"
6 help
7 This is code for addressing several networked computers at once,
8 enlarging your kernel by about 2 KB. You need multicasting if you
9 intend to participate in the MBONE, a high bandwidth network on top
10 of the Internet which carries audio and video broadcasts. More
11 information about the MBONE is on the WWW at
12 <http://www.savetz.com/mbone/>. Information about the multicast
13 capabilities of the various network cards is contained in
14 <file:Documentation/networking/multicast.txt>. For most people, it's
15 safe to say N.
16
17config IP_ADVANCED_ROUTER
18 bool "IP: advanced router"
19 ---help---
20 If you intend to run your Linux box mostly as a router, i.e. as a
21 computer that forwards and redistributes network packets, say Y; you
22 will then be presented with several options that allow more precise
23 control about the routing process.
24
25 The answer to this question won't directly affect the kernel:
26 answering N will just cause the configurator to skip all the
27 questions about advanced routing.
28
29 Note that your box can only act as a router if you enable IP
30 forwarding in your kernel; you can do that by saying Y to "/proc
31 file system support" and "Sysctl support" below and executing the
32 line
33
34 echo "1" > /proc/sys/net/ipv4/ip_forward
35
36 at boot time after the /proc file system has been mounted.
37
38 If you turn on IP forwarding, you should consider the rp_filter, which
39 automatically rejects incoming packets if the routing table entry
40 for their source address doesn't match the network interface they're
41 arriving on. This has security advantages because it prevents the
42 so-called IP spoofing, however it can pose problems if you use
43 asymmetric routing (packets from you to a host take a different path
44 than packets from that host to you) or if you operate a non-routing
45 host which has several IP addresses on different interfaces. To turn
46 rp_filter on use:
47
48 echo 1 > /proc/sys/net/ipv4/conf/<device>/rp_filter
49 or
50 echo 1 > /proc/sys/net/ipv4/conf/all/rp_filter
51
52 Note that some distributions enable it in startup scripts.
53 For details about rp_filter strict and loose mode read
54 <file:Documentation/networking/ip-sysctl.txt>.
55
56 If unsure, say N here.
57
58config IP_FIB_TRIE_STATS
59 bool "FIB TRIE statistics"
60 depends on IP_ADVANCED_ROUTER
61 ---help---
62 Keep track of statistics on structure of FIB TRIE table.
63 Useful for testing and measuring TRIE performance.
64
65config IP_MULTIPLE_TABLES
66 bool "IP: policy routing"
67 depends on IP_ADVANCED_ROUTER
68 select FIB_RULES
69 ---help---
70 Normally, a router decides what to do with a received packet based
71 solely on the packet's final destination address. If you say Y here,
72 the Linux router will also be able to take the packet's source
73 address into account. Furthermore, the TOS (Type-Of-Service) field
74 of the packet can be used for routing decisions as well.
75
76 If you are interested in this, please see the preliminary
77 documentation at <http://www.compendium.com.ar/policy-routing.txt>
78 and <ftp://post.tepkom.ru/pub/vol2/Linux/docs/advanced-routing.tex>.
79 You will need supporting software from
80 <ftp://ftp.tux.org/pub/net/ip-routing/>.
81
82 If unsure, say N.
83
84config IP_ROUTE_MULTIPATH
85 bool "IP: equal cost multipath"
86 depends on IP_ADVANCED_ROUTER
87 help
88 Normally, the routing tables specify a single action to be taken in
89 a deterministic manner for a given packet. If you say Y here
90 however, it becomes possible to attach several actions to a packet
91 pattern, in effect specifying several alternative paths to travel
92 for those packets. The router considers all these paths to be of
93 equal "cost" and chooses one of them in a non-deterministic fashion
94 if a matching packet arrives.
95
96config IP_ROUTE_VERBOSE
97 bool "IP: verbose route monitoring"
98 depends on IP_ADVANCED_ROUTER
99 help
100 If you say Y here, which is recommended, then the kernel will print
101 verbose messages regarding the routing, for example warnings about
102 received packets which look strange and could be evidence of an
103 attack or a misconfigured system somewhere. The information is
104 handled by the klogd daemon which is responsible for kernel messages
105 ("man klogd").
106
107config IP_ROUTE_CLASSID
108 bool
109
110config IP_PNP
111 bool "IP: kernel level autoconfiguration"
112 help
113 This enables automatic configuration of IP addresses of devices and
114 of the routing table during kernel boot, based on either information
115 supplied on the kernel command line or by BOOTP or RARP protocols.
116 You need to say Y only for diskless machines requiring network
117 access to boot (in which case you want to say Y to "Root file system
118 on NFS" as well), because all other machines configure the network
119 in their startup scripts.
120
121config IP_PNP_DHCP
122 bool "IP: DHCP support"
123 depends on IP_PNP
124 ---help---
125 If you want your Linux box to mount its whole root file system (the
126 one containing the directory /) from some other computer over the
127 net via NFS and you want the IP address of your computer to be
128 discovered automatically at boot time using the DHCP protocol (a
129 special protocol designed for doing this job), say Y here. In case
130 the boot ROM of your network card was designed for booting Linux and
131 does DHCP itself, providing all necessary information on the kernel
132 command line, you can say N here.
133
134 If unsure, say Y. Note that if you want to use DHCP, a DHCP server
135 must be operating on your network. Read
136 <file:Documentation/filesystems/nfs/nfsroot.txt> for details.
137
138config IP_PNP_BOOTP
139 bool "IP: BOOTP support"
140 depends on IP_PNP
141 ---help---
142 If you want your Linux box to mount its whole root file system (the
143 one containing the directory /) from some other computer over the
144 net via NFS and you want the IP address of your computer to be
145 discovered automatically at boot time using the BOOTP protocol (a
146 special protocol designed for doing this job), say Y here. In case
147 the boot ROM of your network card was designed for booting Linux and
148 does BOOTP itself, providing all necessary information on the kernel
149 command line, you can say N here. If unsure, say Y. Note that if you
150 want to use BOOTP, a BOOTP server must be operating on your network.
151 Read <file:Documentation/filesystems/nfs/nfsroot.txt> for details.
152
153config IP_PNP_RARP
154 bool "IP: RARP support"
155 depends on IP_PNP
156 help
157 If you want your Linux box to mount its whole root file system (the
158 one containing the directory /) from some other computer over the
159 net via NFS and you want the IP address of your computer to be
160 discovered automatically at boot time using the RARP protocol (an
161 older protocol which is being obsoleted by BOOTP and DHCP), say Y
162 here. Note that if you want to use RARP, a RARP server must be
163 operating on your network. Read
164 <file:Documentation/filesystems/nfs/nfsroot.txt> for details.
165
166# not yet ready..
167# bool ' IP: ARP support' CONFIG_IP_PNP_ARP
168config NET_IPIP
169 tristate "IP: tunneling"
170 select INET_TUNNEL
171 ---help---
172 Tunneling means encapsulating data of one protocol type within
173 another protocol and sending it over a channel that understands the
174 encapsulating protocol. This particular tunneling driver implements
175 encapsulation of IP within IP, which sounds kind of pointless, but
176 can be useful if you want to make your (or some other) machine
177 appear on a different network than it physically is, or to use
178 mobile-IP facilities (allowing laptops to seamlessly move between
179 networks without changing their IP addresses).
180
181 Saying Y to this option will produce two modules ( = code which can
182 be inserted in and removed from the running kernel whenever you
183 want). Most people won't need this and can say N.
184
185config NET_IPGRE_DEMUX
186 tristate "IP: GRE demultiplexer"
187 help
188 This is helper module to demultiplex GRE packets on GRE version field criteria.
189 Required by ip_gre and pptp modules.
190
191config NET_IPGRE
192 tristate "IP: GRE tunnels over IP"
193 depends on (IPV6 || IPV6=n) && NET_IPGRE_DEMUX
194 help
195 Tunneling means encapsulating data of one protocol type within
196 another protocol and sending it over a channel that understands the
197 encapsulating protocol. This particular tunneling driver implements
198 GRE (Generic Routing Encapsulation) and at this time allows
199 encapsulating of IPv4 or IPv6 over existing IPv4 infrastructure.
200 This driver is useful if the other endpoint is a Cisco router: Cisco
201 likes GRE much better than the other Linux tunneling driver ("IP
202 tunneling" above). In addition, GRE allows multicast redistribution
203 through the tunnel.
204
205config NET_IPGRE_BROADCAST
206 bool "IP: broadcast GRE over IP"
207 depends on IP_MULTICAST && NET_IPGRE
208 help
209 One application of GRE/IP is to construct a broadcast WAN (Wide Area
210 Network), which looks like a normal Ethernet LAN (Local Area
211 Network), but can be distributed all over the Internet. If you want
212 to do that, say Y here and to "IP multicast routing" below.
213
214config IP_MROUTE
215 bool "IP: multicast routing"
216 depends on IP_MULTICAST
217 help
218 This is used if you want your machine to act as a router for IP
219 packets that have several destination addresses. It is needed on the
220 MBONE, a high bandwidth network on top of the Internet which carries
221 audio and video broadcasts. In order to do that, you would most
222 likely run the program mrouted. Information about the multicast
223 capabilities of the various network cards is contained in
224 <file:Documentation/networking/multicast.txt>. If you haven't heard
225 about it, you don't need it.
226
227config IP_MROUTE_MULTIPLE_TABLES
228 bool "IP: multicast policy routing"
229 depends on IP_MROUTE && IP_ADVANCED_ROUTER
230 select FIB_RULES
231 help
232 Normally, a multicast router runs a userspace daemon and decides
233 what to do with a multicast packet based on the source and
234 destination addresses. If you say Y here, the multicast router
235 will also be able to take interfaces and packet marks into
236 account and run multiple instances of userspace daemons
237 simultaneously, each one handling a single table.
238
239 If unsure, say N.
240
241config IP_PIMSM_V1
242 bool "IP: PIM-SM version 1 support"
243 depends on IP_MROUTE
244 help
245 Kernel side support for Sparse Mode PIM (Protocol Independent
246 Multicast) version 1. This multicast routing protocol is used widely
247 because Cisco supports it. You need special software to use it
248 (pimd-v1). Please see <http://netweb.usc.edu/pim/> for more
249 information about PIM.
250
251 Say Y if you want to use PIM-SM v1. Note that you can say N here if
252 you just want to use Dense Mode PIM.
253
254config IP_PIMSM_V2
255 bool "IP: PIM-SM version 2 support"
256 depends on IP_MROUTE
257 help
258 Kernel side support for Sparse Mode PIM version 2. In order to use
259 this, you need an experimental routing daemon supporting it (pimd or
260 gated-5). This routing protocol is not used widely, so say N unless
261 you want to play with it.
262
263config ARPD
264 bool "IP: ARP daemon support"
265 ---help---
266 The kernel maintains an internal cache which maps IP addresses to
267 hardware addresses on the local network, so that Ethernet/Token Ring/
268 etc. frames are sent to the proper address on the physical networking
269 layer. Normally, kernel uses the ARP protocol to resolve these
270 mappings.
271
272 Saying Y here adds support to have an user space daemon to do this
273 resolution instead. This is useful for implementing an alternate
274 address resolution protocol (e.g. NHRP on mGRE tunnels) and also for
275 testing purposes.
276
277 If unsure, say N.
278
279config SYN_COOKIES
280 bool "IP: TCP syncookie support"
281 ---help---
282 Normal TCP/IP networking is open to an attack known as "SYN
283 flooding". This denial-of-service attack prevents legitimate remote
284 users from being able to connect to your computer during an ongoing
285 attack and requires very little work from the attacker, who can
286 operate from anywhere on the Internet.
287
288 SYN cookies provide protection against this type of attack. If you
289 say Y here, the TCP/IP stack will use a cryptographic challenge
290 protocol known as "SYN cookies" to enable legitimate users to
291 continue to connect, even when your machine is under attack. There
292 is no need for the legitimate users to change their TCP/IP software;
293 SYN cookies work transparently to them. For technical information
294 about SYN cookies, check out <http://cr.yp.to/syncookies.html>.
295
296 If you are SYN flooded, the source address reported by the kernel is
297 likely to have been forged by the attacker; it is only reported as
298 an aid in tracing the packets to their actual source and should not
299 be taken as absolute truth.
300
301 SYN cookies may prevent correct error reporting on clients when the
302 server is really overloaded. If this happens frequently better turn
303 them off.
304
305 If you say Y here, you can disable SYN cookies at run time by
306 saying Y to "/proc file system support" and
307 "Sysctl support" below and executing the command
308
309 echo 0 > /proc/sys/net/ipv4/tcp_syncookies
310
311 after the /proc file system has been mounted.
312
313 If unsure, say N.
314
315config INET_AH
316 tristate "IP: AH transformation"
317 select XFRM
318 select CRYPTO
319 select CRYPTO_HMAC
320 select CRYPTO_MD5
321 select CRYPTO_SHA1
322 ---help---
323 Support for IPsec AH.
324
325 If unsure, say Y.
326
327config INET_ESP
328 tristate "IP: ESP transformation"
329 select XFRM
330 select CRYPTO
331 select CRYPTO_AUTHENC
332 select CRYPTO_HMAC
333 select CRYPTO_MD5
334 select CRYPTO_CBC
335 select CRYPTO_SHA1
336 select CRYPTO_DES
337 ---help---
338 Support for IPsec ESP.
339
340 If unsure, say Y.
341
342config INET_IPCOMP
343 tristate "IP: IPComp transformation"
344 select INET_XFRM_TUNNEL
345 select XFRM_IPCOMP
346 ---help---
347 Support for IP Payload Compression Protocol (IPComp) (RFC3173),
348 typically needed for IPsec.
349
350 If unsure, say Y.
351
352config INET_XFRM_TUNNEL
353 tristate
354 select INET_TUNNEL
355 default n
356
357config INET_TUNNEL
358 tristate
359 default n
360
361config INET_XFRM_MODE_TRANSPORT
362 tristate "IP: IPsec transport mode"
363 default y
364 select XFRM
365 ---help---
366 Support for IPsec transport mode.
367
368 If unsure, say Y.
369
370config INET_XFRM_MODE_TUNNEL
371 tristate "IP: IPsec tunnel mode"
372 default y
373 select XFRM
374 ---help---
375 Support for IPsec tunnel mode.
376
377 If unsure, say Y.
378
379config INET_XFRM_MODE_BEET
380 tristate "IP: IPsec BEET mode"
381 default y
382 select XFRM
383 ---help---
384 Support for IPsec BEET mode.
385
386 If unsure, say Y.
387
388config INET_LRO
389 tristate "Large Receive Offload (ipv4/tcp)"
390 default y
391 ---help---
392 Support for Large Receive Offload (ipv4/tcp).
393
394 If unsure, say Y.
395
396config INET_DIAG
397 tristate "INET: socket monitoring interface"
398 default y
399 ---help---
400 Support for INET (TCP, DCCP, etc) socket monitoring interface used by
401 native Linux tools such as ss. ss is included in iproute2, currently
402 downloadable at:
403
404 http://www.linuxfoundation.org/collaborate/workgroups/networking/iproute2
405
406 If unsure, say Y.
407
408config INET_TCP_DIAG
409 depends on INET_DIAG
410 def_tristate INET_DIAG
411
412menuconfig TCP_CONG_ADVANCED
413 bool "TCP: advanced congestion control"
414 ---help---
415 Support for selection of various TCP congestion control
416 modules.
417
418 Nearly all users can safely say no here, and a safe default
419 selection will be made (CUBIC with new Reno as a fallback).
420
421 If unsure, say N.
422
423if TCP_CONG_ADVANCED
424
425config TCP_CONG_BIC
426 tristate "Binary Increase Congestion (BIC) control"
427 default m
428 ---help---
429 BIC-TCP is a sender-side only change that ensures a linear RTT
430 fairness under large windows while offering both scalability and
431 bounded TCP-friendliness. The protocol combines two schemes
432 called additive increase and binary search increase. When the
433 congestion window is large, additive increase with a large
434 increment ensures linear RTT fairness as well as good
435 scalability. Under small congestion windows, binary search
436 increase provides TCP friendliness.
437 See http://www.csc.ncsu.edu/faculty/rhee/export/bitcp/
438
439config TCP_CONG_CUBIC
440 tristate "CUBIC TCP"
441 default y
442 ---help---
443 This is version 2.0 of BIC-TCP which uses a cubic growth function
444 among other techniques.
445 See http://www.csc.ncsu.edu/faculty/rhee/export/bitcp/cubic-paper.pdf
446
447config TCP_CONG_WESTWOOD
448 tristate "TCP Westwood+"
449 default m
450 ---help---
451 TCP Westwood+ is a sender-side only modification of the TCP Reno
452 protocol stack that optimizes the performance of TCP congestion
453 control. It is based on end-to-end bandwidth estimation to set
454 congestion window and slow start threshold after a congestion
455 episode. Using this estimation, TCP Westwood+ adaptively sets a
456 slow start threshold and a congestion window which takes into
457 account the bandwidth used at the time congestion is experienced.
458 TCP Westwood+ significantly increases fairness wrt TCP Reno in
459 wired networks and throughput over wireless links.
460
461config TCP_CONG_HTCP
462 tristate "H-TCP"
463 default m
464 ---help---
465 H-TCP is a send-side only modifications of the TCP Reno
466 protocol stack that optimizes the performance of TCP
467 congestion control for high speed network links. It uses a
468 modeswitch to change the alpha and beta parameters of TCP Reno
469 based on network conditions and in a way so as to be fair with
470 other Reno and H-TCP flows.
471
472config TCP_CONG_HSTCP
473 tristate "High Speed TCP"
474 depends on EXPERIMENTAL
475 default n
476 ---help---
477 Sally Floyd's High Speed TCP (RFC 3649) congestion control.
478 A modification to TCP's congestion control mechanism for use
479 with large congestion windows. A table indicates how much to
480 increase the congestion window by when an ACK is received.
481 For more detail see http://www.icir.org/floyd/hstcp.html
482
483config TCP_CONG_HYBLA
484 tristate "TCP-Hybla congestion control algorithm"
485 depends on EXPERIMENTAL
486 default n
487 ---help---
488 TCP-Hybla is a sender-side only change that eliminates penalization of
489 long-RTT, large-bandwidth connections, like when satellite legs are
490 involved, especially when sharing a common bottleneck with normal
491 terrestrial connections.
492
493config TCP_CONG_VEGAS
494 tristate "TCP Vegas"
495 depends on EXPERIMENTAL
496 default n
497 ---help---
498 TCP Vegas is a sender-side only change to TCP that anticipates
499 the onset of congestion by estimating the bandwidth. TCP Vegas
500 adjusts the sending rate by modifying the congestion
501 window. TCP Vegas should provide less packet loss, but it is
502 not as aggressive as TCP Reno.
503
504config TCP_CONG_SCALABLE
505 tristate "Scalable TCP"
506 depends on EXPERIMENTAL
507 default n
508 ---help---
509 Scalable TCP is a sender-side only change to TCP which uses a
510 MIMD congestion control algorithm which has some nice scaling
511 properties, though is known to have fairness issues.
512 See http://www.deneholme.net/tom/scalable/
513
514config TCP_CONG_LP
515 tristate "TCP Low Priority"
516 depends on EXPERIMENTAL
517 default n
518 ---help---
519 TCP Low Priority (TCP-LP), a distributed algorithm whose goal is
520 to utilize only the excess network bandwidth as compared to the
521 ``fair share`` of bandwidth as targeted by TCP.
522 See http://www-ece.rice.edu/networks/TCP-LP/
523
524config TCP_CONG_VENO
525 tristate "TCP Veno"
526 depends on EXPERIMENTAL
527 default n
528 ---help---
529 TCP Veno is a sender-side only enhancement of TCP to obtain better
530 throughput over wireless networks. TCP Veno makes use of state
531 distinguishing to circumvent the difficult judgment of the packet loss
532 type. TCP Veno cuts down less congestion window in response to random
533 loss packets.
534 See <http://ieeexplore.ieee.org/xpl/freeabs_all.jsp?arnumber=1177186>
535
536config TCP_CONG_YEAH
537 tristate "YeAH TCP"
538 depends on EXPERIMENTAL
539 select TCP_CONG_VEGAS
540 default n
541 ---help---
542 YeAH-TCP is a sender-side high-speed enabled TCP congestion control
543 algorithm, which uses a mixed loss/delay approach to compute the
544 congestion window. It's design goals target high efficiency,
545 internal, RTT and Reno fairness, resilience to link loss while
546 keeping network elements load as low as possible.
547
548 For further details look here:
549 http://wil.cs.caltech.edu/pfldnet2007/paper/YeAH_TCP.pdf
550
551config TCP_CONG_ILLINOIS
552 tristate "TCP Illinois"
553 depends on EXPERIMENTAL
554 default n
555 ---help---
556 TCP-Illinois is a sender-side modification of TCP Reno for
557 high speed long delay links. It uses round-trip-time to
558 adjust the alpha and beta parameters to achieve a higher average
559 throughput and maintain fairness.
560
561 For further details see:
562 http://www.ews.uiuc.edu/~shaoliu/tcpillinois/index.html
563
564choice
565 prompt "Default TCP congestion control"
566 default DEFAULT_CUBIC
567 help
568 Select the TCP congestion control that will be used by default
569 for all connections.
570
571 config DEFAULT_BIC
572 bool "Bic" if TCP_CONG_BIC=y
573
574 config DEFAULT_CUBIC
575 bool "Cubic" if TCP_CONG_CUBIC=y
576
577 config DEFAULT_HTCP
578 bool "Htcp" if TCP_CONG_HTCP=y
579
580 config DEFAULT_HYBLA
581 bool "Hybla" if TCP_CONG_HYBLA=y
582
583 config DEFAULT_VEGAS
584 bool "Vegas" if TCP_CONG_VEGAS=y
585
586 config DEFAULT_VENO
587 bool "Veno" if TCP_CONG_VENO=y
588
589 config DEFAULT_WESTWOOD
590 bool "Westwood" if TCP_CONG_WESTWOOD=y
591
592 config DEFAULT_RENO
593 bool "Reno"
594
595endchoice
596
597endif
598
599config TCP_CONG_CUBIC
600 tristate
601 depends on !TCP_CONG_ADVANCED
602 default y
603
604config DEFAULT_TCP_CONG
605 string
606 default "bic" if DEFAULT_BIC
607 default "cubic" if DEFAULT_CUBIC
608 default "htcp" if DEFAULT_HTCP
609 default "hybla" if DEFAULT_HYBLA
610 default "vegas" if DEFAULT_VEGAS
611 default "westwood" if DEFAULT_WESTWOOD
612 default "veno" if DEFAULT_VENO
613 default "reno" if DEFAULT_RENO
614 default "cubic"
615
616config TCP_MD5SIG
617 bool "TCP: MD5 Signature Option support (RFC2385) (EXPERIMENTAL)"
618 depends on EXPERIMENTAL
619 select CRYPTO
620 select CRYPTO_MD5
621 ---help---
622 RFC2385 specifies a method of giving MD5 protection to TCP sessions.
623 Its main (only?) use is to protect BGP sessions between core routers
624 on the Internet.
625
626 If unsure, say N.
1#
2# IP configuration
3#
4config IP_MULTICAST
5 bool "IP: multicasting"
6 help
7 This is code for addressing several networked computers at once,
8 enlarging your kernel by about 2 KB. You need multicasting if you
9 intend to participate in the MBONE, a high bandwidth network on top
10 of the Internet which carries audio and video broadcasts. More
11 information about the MBONE is on the WWW at
12 <http://www.savetz.com/mbone/>. Information about the multicast
13 capabilities of the various network cards is contained in
14 <file:Documentation/networking/multicast.txt>. For most people, it's
15 safe to say N.
16
17config IP_ADVANCED_ROUTER
18 bool "IP: advanced router"
19 ---help---
20 If you intend to run your Linux box mostly as a router, i.e. as a
21 computer that forwards and redistributes network packets, say Y; you
22 will then be presented with several options that allow more precise
23 control about the routing process.
24
25 The answer to this question won't directly affect the kernel:
26 answering N will just cause the configurator to skip all the
27 questions about advanced routing.
28
29 Note that your box can only act as a router if you enable IP
30 forwarding in your kernel; you can do that by saying Y to "/proc
31 file system support" and "Sysctl support" below and executing the
32 line
33
34 echo "1" > /proc/sys/net/ipv4/ip_forward
35
36 at boot time after the /proc file system has been mounted.
37
38 If you turn on IP forwarding, you should consider the rp_filter, which
39 automatically rejects incoming packets if the routing table entry
40 for their source address doesn't match the network interface they're
41 arriving on. This has security advantages because it prevents the
42 so-called IP spoofing, however it can pose problems if you use
43 asymmetric routing (packets from you to a host take a different path
44 than packets from that host to you) or if you operate a non-routing
45 host which has several IP addresses on different interfaces. To turn
46 rp_filter on use:
47
48 echo 1 > /proc/sys/net/ipv4/conf/<device>/rp_filter
49 or
50 echo 1 > /proc/sys/net/ipv4/conf/all/rp_filter
51
52 Note that some distributions enable it in startup scripts.
53 For details about rp_filter strict and loose mode read
54 <file:Documentation/networking/ip-sysctl.txt>.
55
56 If unsure, say N here.
57
58config IP_FIB_TRIE_STATS
59 bool "FIB TRIE statistics"
60 depends on IP_ADVANCED_ROUTER
61 ---help---
62 Keep track of statistics on structure of FIB TRIE table.
63 Useful for testing and measuring TRIE performance.
64
65config IP_MULTIPLE_TABLES
66 bool "IP: policy routing"
67 depends on IP_ADVANCED_ROUTER
68 select FIB_RULES
69 ---help---
70 Normally, a router decides what to do with a received packet based
71 solely on the packet's final destination address. If you say Y here,
72 the Linux router will also be able to take the packet's source
73 address into account. Furthermore, the TOS (Type-Of-Service) field
74 of the packet can be used for routing decisions as well.
75
76 If you are interested in this, please see the preliminary
77 documentation at <http://www.compendium.com.ar/policy-routing.txt>
78 and <ftp://post.tepkom.ru/pub/vol2/Linux/docs/advanced-routing.tex>.
79 You will need supporting software from
80 <ftp://ftp.tux.org/pub/net/ip-routing/>.
81
82 If unsure, say N.
83
84config IP_ROUTE_MULTIPATH
85 bool "IP: equal cost multipath"
86 depends on IP_ADVANCED_ROUTER
87 help
88 Normally, the routing tables specify a single action to be taken in
89 a deterministic manner for a given packet. If you say Y here
90 however, it becomes possible to attach several actions to a packet
91 pattern, in effect specifying several alternative paths to travel
92 for those packets. The router considers all these paths to be of
93 equal "cost" and chooses one of them in a non-deterministic fashion
94 if a matching packet arrives.
95
96config IP_ROUTE_VERBOSE
97 bool "IP: verbose route monitoring"
98 depends on IP_ADVANCED_ROUTER
99 help
100 If you say Y here, which is recommended, then the kernel will print
101 verbose messages regarding the routing, for example warnings about
102 received packets which look strange and could be evidence of an
103 attack or a misconfigured system somewhere. The information is
104 handled by the klogd daemon which is responsible for kernel messages
105 ("man klogd").
106
107config IP_ROUTE_CLASSID
108 bool
109
110config IP_PNP
111 bool "IP: kernel level autoconfiguration"
112 help
113 This enables automatic configuration of IP addresses of devices and
114 of the routing table during kernel boot, based on either information
115 supplied on the kernel command line or by BOOTP or RARP protocols.
116 You need to say Y only for diskless machines requiring network
117 access to boot (in which case you want to say Y to "Root file system
118 on NFS" as well), because all other machines configure the network
119 in their startup scripts.
120
121config IP_PNP_DHCP
122 bool "IP: DHCP support"
123 depends on IP_PNP
124 ---help---
125 If you want your Linux box to mount its whole root file system (the
126 one containing the directory /) from some other computer over the
127 net via NFS and you want the IP address of your computer to be
128 discovered automatically at boot time using the DHCP protocol (a
129 special protocol designed for doing this job), say Y here. In case
130 the boot ROM of your network card was designed for booting Linux and
131 does DHCP itself, providing all necessary information on the kernel
132 command line, you can say N here.
133
134 If unsure, say Y. Note that if you want to use DHCP, a DHCP server
135 must be operating on your network. Read
136 <file:Documentation/filesystems/nfs/nfsroot.txt> for details.
137
138config IP_PNP_BOOTP
139 bool "IP: BOOTP support"
140 depends on IP_PNP
141 ---help---
142 If you want your Linux box to mount its whole root file system (the
143 one containing the directory /) from some other computer over the
144 net via NFS and you want the IP address of your computer to be
145 discovered automatically at boot time using the BOOTP protocol (a
146 special protocol designed for doing this job), say Y here. In case
147 the boot ROM of your network card was designed for booting Linux and
148 does BOOTP itself, providing all necessary information on the kernel
149 command line, you can say N here. If unsure, say Y. Note that if you
150 want to use BOOTP, a BOOTP server must be operating on your network.
151 Read <file:Documentation/filesystems/nfs/nfsroot.txt> for details.
152
153config IP_PNP_RARP
154 bool "IP: RARP support"
155 depends on IP_PNP
156 help
157 If you want your Linux box to mount its whole root file system (the
158 one containing the directory /) from some other computer over the
159 net via NFS and you want the IP address of your computer to be
160 discovered automatically at boot time using the RARP protocol (an
161 older protocol which is being obsoleted by BOOTP and DHCP), say Y
162 here. Note that if you want to use RARP, a RARP server must be
163 operating on your network. Read
164 <file:Documentation/filesystems/nfs/nfsroot.txt> for details.
165
166config NET_IPIP
167 tristate "IP: tunneling"
168 select INET_TUNNEL
169 ---help---
170 Tunneling means encapsulating data of one protocol type within
171 another protocol and sending it over a channel that understands the
172 encapsulating protocol. This particular tunneling driver implements
173 encapsulation of IP within IP, which sounds kind of pointless, but
174 can be useful if you want to make your (or some other) machine
175 appear on a different network than it physically is, or to use
176 mobile-IP facilities (allowing laptops to seamlessly move between
177 networks without changing their IP addresses).
178
179 Saying Y to this option will produce two modules ( = code which can
180 be inserted in and removed from the running kernel whenever you
181 want). Most people won't need this and can say N.
182
183config NET_IPGRE_DEMUX
184 tristate "IP: GRE demultiplexer"
185 help
186 This is helper module to demultiplex GRE packets on GRE version field criteria.
187 Required by ip_gre and pptp modules.
188
189config NET_IPGRE
190 tristate "IP: GRE tunnels over IP"
191 depends on (IPV6 || IPV6=n) && NET_IPGRE_DEMUX
192 help
193 Tunneling means encapsulating data of one protocol type within
194 another protocol and sending it over a channel that understands the
195 encapsulating protocol. This particular tunneling driver implements
196 GRE (Generic Routing Encapsulation) and at this time allows
197 encapsulating of IPv4 or IPv6 over existing IPv4 infrastructure.
198 This driver is useful if the other endpoint is a Cisco router: Cisco
199 likes GRE much better than the other Linux tunneling driver ("IP
200 tunneling" above). In addition, GRE allows multicast redistribution
201 through the tunnel.
202
203config NET_IPGRE_BROADCAST
204 bool "IP: broadcast GRE over IP"
205 depends on IP_MULTICAST && NET_IPGRE
206 help
207 One application of GRE/IP is to construct a broadcast WAN (Wide Area
208 Network), which looks like a normal Ethernet LAN (Local Area
209 Network), but can be distributed all over the Internet. If you want
210 to do that, say Y here and to "IP multicast routing" below.
211
212config IP_MROUTE
213 bool "IP: multicast routing"
214 depends on IP_MULTICAST
215 help
216 This is used if you want your machine to act as a router for IP
217 packets that have several destination addresses. It is needed on the
218 MBONE, a high bandwidth network on top of the Internet which carries
219 audio and video broadcasts. In order to do that, you would most
220 likely run the program mrouted. Information about the multicast
221 capabilities of the various network cards is contained in
222 <file:Documentation/networking/multicast.txt>. If you haven't heard
223 about it, you don't need it.
224
225config IP_MROUTE_MULTIPLE_TABLES
226 bool "IP: multicast policy routing"
227 depends on IP_MROUTE && IP_ADVANCED_ROUTER
228 select FIB_RULES
229 help
230 Normally, a multicast router runs a userspace daemon and decides
231 what to do with a multicast packet based on the source and
232 destination addresses. If you say Y here, the multicast router
233 will also be able to take interfaces and packet marks into
234 account and run multiple instances of userspace daemons
235 simultaneously, each one handling a single table.
236
237 If unsure, say N.
238
239config IP_PIMSM_V1
240 bool "IP: PIM-SM version 1 support"
241 depends on IP_MROUTE
242 help
243 Kernel side support for Sparse Mode PIM (Protocol Independent
244 Multicast) version 1. This multicast routing protocol is used widely
245 because Cisco supports it. You need special software to use it
246 (pimd-v1). Please see <http://netweb.usc.edu/pim/> for more
247 information about PIM.
248
249 Say Y if you want to use PIM-SM v1. Note that you can say N here if
250 you just want to use Dense Mode PIM.
251
252config IP_PIMSM_V2
253 bool "IP: PIM-SM version 2 support"
254 depends on IP_MROUTE
255 help
256 Kernel side support for Sparse Mode PIM version 2. In order to use
257 this, you need an experimental routing daemon supporting it (pimd or
258 gated-5). This routing protocol is not used widely, so say N unless
259 you want to play with it.
260
261config ARPD
262 bool "IP: ARP daemon support"
263 ---help---
264 The kernel maintains an internal cache which maps IP addresses to
265 hardware addresses on the local network, so that Ethernet
266 frames are sent to the proper address on the physical networking
267 layer. Normally, kernel uses the ARP protocol to resolve these
268 mappings.
269
270 Saying Y here adds support to have an user space daemon to do this
271 resolution instead. This is useful for implementing an alternate
272 address resolution protocol (e.g. NHRP on mGRE tunnels) and also for
273 testing purposes.
274
275 If unsure, say N.
276
277config SYN_COOKIES
278 bool "IP: TCP syncookie support"
279 ---help---
280 Normal TCP/IP networking is open to an attack known as "SYN
281 flooding". This denial-of-service attack prevents legitimate remote
282 users from being able to connect to your computer during an ongoing
283 attack and requires very little work from the attacker, who can
284 operate from anywhere on the Internet.
285
286 SYN cookies provide protection against this type of attack. If you
287 say Y here, the TCP/IP stack will use a cryptographic challenge
288 protocol known as "SYN cookies" to enable legitimate users to
289 continue to connect, even when your machine is under attack. There
290 is no need for the legitimate users to change their TCP/IP software;
291 SYN cookies work transparently to them. For technical information
292 about SYN cookies, check out <http://cr.yp.to/syncookies.html>.
293
294 If you are SYN flooded, the source address reported by the kernel is
295 likely to have been forged by the attacker; it is only reported as
296 an aid in tracing the packets to their actual source and should not
297 be taken as absolute truth.
298
299 SYN cookies may prevent correct error reporting on clients when the
300 server is really overloaded. If this happens frequently better turn
301 them off.
302
303 If you say Y here, you can disable SYN cookies at run time by
304 saying Y to "/proc file system support" and
305 "Sysctl support" below and executing the command
306
307 echo 0 > /proc/sys/net/ipv4/tcp_syncookies
308
309 after the /proc file system has been mounted.
310
311 If unsure, say N.
312
313config INET_AH
314 tristate "IP: AH transformation"
315 select XFRM_ALGO
316 select CRYPTO
317 select CRYPTO_HMAC
318 select CRYPTO_MD5
319 select CRYPTO_SHA1
320 ---help---
321 Support for IPsec AH.
322
323 If unsure, say Y.
324
325config INET_ESP
326 tristate "IP: ESP transformation"
327 select XFRM_ALGO
328 select CRYPTO
329 select CRYPTO_AUTHENC
330 select CRYPTO_HMAC
331 select CRYPTO_MD5
332 select CRYPTO_CBC
333 select CRYPTO_SHA1
334 select CRYPTO_DES
335 ---help---
336 Support for IPsec ESP.
337
338 If unsure, say Y.
339
340config INET_IPCOMP
341 tristate "IP: IPComp transformation"
342 select INET_XFRM_TUNNEL
343 select XFRM_IPCOMP
344 ---help---
345 Support for IP Payload Compression Protocol (IPComp) (RFC3173),
346 typically needed for IPsec.
347
348 If unsure, say Y.
349
350config INET_XFRM_TUNNEL
351 tristate
352 select INET_TUNNEL
353 default n
354
355config INET_TUNNEL
356 tristate
357 default n
358
359config INET_XFRM_MODE_TRANSPORT
360 tristate "IP: IPsec transport mode"
361 default y
362 select XFRM
363 ---help---
364 Support for IPsec transport mode.
365
366 If unsure, say Y.
367
368config INET_XFRM_MODE_TUNNEL
369 tristate "IP: IPsec tunnel mode"
370 default y
371 select XFRM
372 ---help---
373 Support for IPsec tunnel mode.
374
375 If unsure, say Y.
376
377config INET_XFRM_MODE_BEET
378 tristate "IP: IPsec BEET mode"
379 default y
380 select XFRM
381 ---help---
382 Support for IPsec BEET mode.
383
384 If unsure, say Y.
385
386config INET_LRO
387 tristate "Large Receive Offload (ipv4/tcp)"
388 default y
389 ---help---
390 Support for Large Receive Offload (ipv4/tcp).
391
392 If unsure, say Y.
393
394config INET_DIAG
395 tristate "INET: socket monitoring interface"
396 default y
397 ---help---
398 Support for INET (TCP, DCCP, etc) socket monitoring interface used by
399 native Linux tools such as ss. ss is included in iproute2, currently
400 downloadable at:
401
402 http://www.linuxfoundation.org/collaborate/workgroups/networking/iproute2
403
404 If unsure, say Y.
405
406config INET_TCP_DIAG
407 depends on INET_DIAG
408 def_tristate INET_DIAG
409
410config INET_UDP_DIAG
411 tristate "UDP: socket monitoring interface"
412 depends on INET_DIAG && (IPV6 || IPV6=n)
413 default n
414 ---help---
415 Support for UDP socket monitoring interface used by the ss tool.
416 If unsure, say Y.
417
418menuconfig TCP_CONG_ADVANCED
419 bool "TCP: advanced congestion control"
420 ---help---
421 Support for selection of various TCP congestion control
422 modules.
423
424 Nearly all users can safely say no here, and a safe default
425 selection will be made (CUBIC with new Reno as a fallback).
426
427 If unsure, say N.
428
429if TCP_CONG_ADVANCED
430
431config TCP_CONG_BIC
432 tristate "Binary Increase Congestion (BIC) control"
433 default m
434 ---help---
435 BIC-TCP is a sender-side only change that ensures a linear RTT
436 fairness under large windows while offering both scalability and
437 bounded TCP-friendliness. The protocol combines two schemes
438 called additive increase and binary search increase. When the
439 congestion window is large, additive increase with a large
440 increment ensures linear RTT fairness as well as good
441 scalability. Under small congestion windows, binary search
442 increase provides TCP friendliness.
443 See http://www.csc.ncsu.edu/faculty/rhee/export/bitcp/
444
445config TCP_CONG_CUBIC
446 tristate "CUBIC TCP"
447 default y
448 ---help---
449 This is version 2.0 of BIC-TCP which uses a cubic growth function
450 among other techniques.
451 See http://www.csc.ncsu.edu/faculty/rhee/export/bitcp/cubic-paper.pdf
452
453config TCP_CONG_WESTWOOD
454 tristate "TCP Westwood+"
455 default m
456 ---help---
457 TCP Westwood+ is a sender-side only modification of the TCP Reno
458 protocol stack that optimizes the performance of TCP congestion
459 control. It is based on end-to-end bandwidth estimation to set
460 congestion window and slow start threshold after a congestion
461 episode. Using this estimation, TCP Westwood+ adaptively sets a
462 slow start threshold and a congestion window which takes into
463 account the bandwidth used at the time congestion is experienced.
464 TCP Westwood+ significantly increases fairness wrt TCP Reno in
465 wired networks and throughput over wireless links.
466
467config TCP_CONG_HTCP
468 tristate "H-TCP"
469 default m
470 ---help---
471 H-TCP is a send-side only modifications of the TCP Reno
472 protocol stack that optimizes the performance of TCP
473 congestion control for high speed network links. It uses a
474 modeswitch to change the alpha and beta parameters of TCP Reno
475 based on network conditions and in a way so as to be fair with
476 other Reno and H-TCP flows.
477
478config TCP_CONG_HSTCP
479 tristate "High Speed TCP"
480 depends on EXPERIMENTAL
481 default n
482 ---help---
483 Sally Floyd's High Speed TCP (RFC 3649) congestion control.
484 A modification to TCP's congestion control mechanism for use
485 with large congestion windows. A table indicates how much to
486 increase the congestion window by when an ACK is received.
487 For more detail see http://www.icir.org/floyd/hstcp.html
488
489config TCP_CONG_HYBLA
490 tristate "TCP-Hybla congestion control algorithm"
491 depends on EXPERIMENTAL
492 default n
493 ---help---
494 TCP-Hybla is a sender-side only change that eliminates penalization of
495 long-RTT, large-bandwidth connections, like when satellite legs are
496 involved, especially when sharing a common bottleneck with normal
497 terrestrial connections.
498
499config TCP_CONG_VEGAS
500 tristate "TCP Vegas"
501 depends on EXPERIMENTAL
502 default n
503 ---help---
504 TCP Vegas is a sender-side only change to TCP that anticipates
505 the onset of congestion by estimating the bandwidth. TCP Vegas
506 adjusts the sending rate by modifying the congestion
507 window. TCP Vegas should provide less packet loss, but it is
508 not as aggressive as TCP Reno.
509
510config TCP_CONG_SCALABLE
511 tristate "Scalable TCP"
512 depends on EXPERIMENTAL
513 default n
514 ---help---
515 Scalable TCP is a sender-side only change to TCP which uses a
516 MIMD congestion control algorithm which has some nice scaling
517 properties, though is known to have fairness issues.
518 See http://www.deneholme.net/tom/scalable/
519
520config TCP_CONG_LP
521 tristate "TCP Low Priority"
522 depends on EXPERIMENTAL
523 default n
524 ---help---
525 TCP Low Priority (TCP-LP), a distributed algorithm whose goal is
526 to utilize only the excess network bandwidth as compared to the
527 ``fair share`` of bandwidth as targeted by TCP.
528 See http://www-ece.rice.edu/networks/TCP-LP/
529
530config TCP_CONG_VENO
531 tristate "TCP Veno"
532 depends on EXPERIMENTAL
533 default n
534 ---help---
535 TCP Veno is a sender-side only enhancement of TCP to obtain better
536 throughput over wireless networks. TCP Veno makes use of state
537 distinguishing to circumvent the difficult judgment of the packet loss
538 type. TCP Veno cuts down less congestion window in response to random
539 loss packets.
540 See <http://ieeexplore.ieee.org/xpl/freeabs_all.jsp?arnumber=1177186>
541
542config TCP_CONG_YEAH
543 tristate "YeAH TCP"
544 depends on EXPERIMENTAL
545 select TCP_CONG_VEGAS
546 default n
547 ---help---
548 YeAH-TCP is a sender-side high-speed enabled TCP congestion control
549 algorithm, which uses a mixed loss/delay approach to compute the
550 congestion window. It's design goals target high efficiency,
551 internal, RTT and Reno fairness, resilience to link loss while
552 keeping network elements load as low as possible.
553
554 For further details look here:
555 http://wil.cs.caltech.edu/pfldnet2007/paper/YeAH_TCP.pdf
556
557config TCP_CONG_ILLINOIS
558 tristate "TCP Illinois"
559 depends on EXPERIMENTAL
560 default n
561 ---help---
562 TCP-Illinois is a sender-side modification of TCP Reno for
563 high speed long delay links. It uses round-trip-time to
564 adjust the alpha and beta parameters to achieve a higher average
565 throughput and maintain fairness.
566
567 For further details see:
568 http://www.ews.uiuc.edu/~shaoliu/tcpillinois/index.html
569
570choice
571 prompt "Default TCP congestion control"
572 default DEFAULT_CUBIC
573 help
574 Select the TCP congestion control that will be used by default
575 for all connections.
576
577 config DEFAULT_BIC
578 bool "Bic" if TCP_CONG_BIC=y
579
580 config DEFAULT_CUBIC
581 bool "Cubic" if TCP_CONG_CUBIC=y
582
583 config DEFAULT_HTCP
584 bool "Htcp" if TCP_CONG_HTCP=y
585
586 config DEFAULT_HYBLA
587 bool "Hybla" if TCP_CONG_HYBLA=y
588
589 config DEFAULT_VEGAS
590 bool "Vegas" if TCP_CONG_VEGAS=y
591
592 config DEFAULT_VENO
593 bool "Veno" if TCP_CONG_VENO=y
594
595 config DEFAULT_WESTWOOD
596 bool "Westwood" if TCP_CONG_WESTWOOD=y
597
598 config DEFAULT_RENO
599 bool "Reno"
600
601endchoice
602
603endif
604
605config TCP_CONG_CUBIC
606 tristate
607 depends on !TCP_CONG_ADVANCED
608 default y
609
610config DEFAULT_TCP_CONG
611 string
612 default "bic" if DEFAULT_BIC
613 default "cubic" if DEFAULT_CUBIC
614 default "htcp" if DEFAULT_HTCP
615 default "hybla" if DEFAULT_HYBLA
616 default "vegas" if DEFAULT_VEGAS
617 default "westwood" if DEFAULT_WESTWOOD
618 default "veno" if DEFAULT_VENO
619 default "reno" if DEFAULT_RENO
620 default "cubic"
621
622config TCP_MD5SIG
623 bool "TCP: MD5 Signature Option support (RFC2385) (EXPERIMENTAL)"
624 depends on EXPERIMENTAL
625 select CRYPTO
626 select CRYPTO_MD5
627 ---help---
628 RFC2385 specifies a method of giving MD5 protection to TCP sessions.
629 Its main (only?) use is to protect BGP sessions between core routers
630 on the Internet.
631
632 If unsure, say N.