5 bool "IP: multicasting"
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/>. For most people, it's safe to say N.
14 config IP_ADVANCED_ROUTER
15 bool "IP: advanced router"
17 If you intend to run your Linux box mostly as a router, i.e. as a
18 computer that forwards and redistributes network packets, say Y; you
19 will then be presented with several options that allow more precise
20 control about the routing process.
22 The answer to this question won't directly affect the kernel:
23 answering N will just cause the configurator to skip all the
24 questions about advanced routing.
26 Note that your box can only act as a router if you enable IP
27 forwarding in your kernel; you can do that by saying Y to "/proc
28 file system support" and "Sysctl support" below and executing the
31 echo "1" > /proc/sys/net/ipv4/ip_forward
33 at boot time after the /proc file system has been mounted.
35 If you turn on IP forwarding, you should consider the rp_filter, which
36 automatically rejects incoming packets if the routing table entry
37 for their source address doesn't match the network interface they're
38 arriving on. This has security advantages because it prevents the
39 so-called IP spoofing, however it can pose problems if you use
40 asymmetric routing (packets from you to a host take a different path
41 than packets from that host to you) or if you operate a non-routing
42 host which has several IP addresses on different interfaces. To turn
45 echo 1 > /proc/sys/net/ipv4/conf/<device>/rp_filter
47 echo 1 > /proc/sys/net/ipv4/conf/all/rp_filter
49 Note that some distributions enable it in startup scripts.
50 For details about rp_filter strict and loose mode read
51 <file:Documentation/networking/ip-sysctl.txt>.
53 If unsure, say N here.
55 config IP_FIB_TRIE_STATS
56 bool "FIB TRIE statistics"
57 depends on IP_ADVANCED_ROUTER
59 Keep track of statistics on structure of FIB TRIE table.
60 Useful for testing and measuring TRIE performance.
62 config IP_MULTIPLE_TABLES
63 bool "IP: policy routing"
64 depends on IP_ADVANCED_ROUTER
67 Normally, a router decides what to do with a received packet based
68 solely on the packet's final destination address. If you say Y here,
69 the Linux router will also be able to take the packet's source
70 address into account. Furthermore, the TOS (Type-Of-Service) field
71 of the packet can be used for routing decisions as well.
73 If you need more information, see the Linux Advanced
74 Routing and Traffic Control documentation at
75 <http://lartc.org/howto/lartc.rpdb.html>
79 config IP_ROUTE_MULTIPATH
80 bool "IP: equal cost multipath"
81 depends on IP_ADVANCED_ROUTER
83 Normally, the routing tables specify a single action to be taken in
84 a deterministic manner for a given packet. If you say Y here
85 however, it becomes possible to attach several actions to a packet
86 pattern, in effect specifying several alternative paths to travel
87 for those packets. The router considers all these paths to be of
88 equal "cost" and chooses one of them in a non-deterministic fashion
89 if a matching packet arrives.
91 config IP_ROUTE_VERBOSE
92 bool "IP: verbose route monitoring"
93 depends on IP_ADVANCED_ROUTER
95 If you say Y here, which is recommended, then the kernel will print
96 verbose messages regarding the routing, for example warnings about
97 received packets which look strange and could be evidence of an
98 attack or a misconfigured system somewhere. The information is
99 handled by the klogd daemon which is responsible for kernel messages
102 config IP_ROUTE_CLASSID
106 bool "IP: kernel level autoconfiguration"
108 This enables automatic configuration of IP addresses of devices and
109 of the routing table during kernel boot, based on either information
110 supplied on the kernel command line or by BOOTP or RARP protocols.
111 You need to say Y only for diskless machines requiring network
112 access to boot (in which case you want to say Y to "Root file system
113 on NFS" as well), because all other machines configure the network
114 in their startup scripts.
117 bool "IP: DHCP support"
120 If you want your Linux box to mount its whole root file system (the
121 one containing the directory /) from some other computer over the
122 net via NFS and you want the IP address of your computer to be
123 discovered automatically at boot time using the DHCP protocol (a
124 special protocol designed for doing this job), say Y here. In case
125 the boot ROM of your network card was designed for booting Linux and
126 does DHCP itself, providing all necessary information on the kernel
127 command line, you can say N here.
129 If unsure, say Y. Note that if you want to use DHCP, a DHCP server
130 must be operating on your network. Read
131 <file:Documentation/filesystems/nfs/nfsroot.txt> for details.
134 bool "IP: BOOTP support"
137 If you want your Linux box to mount its whole root file system (the
138 one containing the directory /) from some other computer over the
139 net via NFS and you want the IP address of your computer to be
140 discovered automatically at boot time using the BOOTP protocol (a
141 special protocol designed for doing this job), say Y here. In case
142 the boot ROM of your network card was designed for booting Linux and
143 does BOOTP itself, providing all necessary information on the kernel
144 command line, you can say N here. If unsure, say Y. Note that if you
145 want to use BOOTP, a BOOTP server must be operating on your network.
146 Read <file:Documentation/filesystems/nfs/nfsroot.txt> for details.
149 bool "IP: RARP support"
152 If you want your Linux box to mount its whole root file system (the
153 one containing the directory /) from some other computer over the
154 net via NFS and you want the IP address of your computer to be
155 discovered automatically at boot time using the RARP protocol (an
156 older protocol which is being obsoleted by BOOTP and DHCP), say Y
157 here. Note that if you want to use RARP, a RARP server must be
158 operating on your network. Read
159 <file:Documentation/filesystems/nfs/nfsroot.txt> for details.
162 tristate "IP: tunneling"
166 Tunneling means encapsulating data of one protocol type within
167 another protocol and sending it over a channel that understands the
168 encapsulating protocol. This particular tunneling driver implements
169 encapsulation of IP within IP, which sounds kind of pointless, but
170 can be useful if you want to make your (or some other) machine
171 appear on a different network than it physically is, or to use
172 mobile-IP facilities (allowing laptops to seamlessly move between
173 networks without changing their IP addresses).
175 Saying Y to this option will produce two modules ( = code which can
176 be inserted in and removed from the running kernel whenever you
177 want). Most people won't need this and can say N.
179 config NET_IPGRE_DEMUX
180 tristate "IP: GRE demultiplexer"
182 This is helper module to demultiplex GRE packets on GRE version field criteria.
183 Required by ip_gre and pptp modules.
192 tristate "IP: GRE tunnels over IP"
193 depends on (IPV6 || IPV6=n) && NET_IPGRE_DEMUX
196 Tunneling means encapsulating data of one protocol type within
197 another protocol and sending it over a channel that understands the
198 encapsulating protocol. This particular tunneling driver implements
199 GRE (Generic Routing Encapsulation) and at this time allows
200 encapsulating of IPv4 or IPv6 over existing IPv4 infrastructure.
201 This driver is useful if the other endpoint is a Cisco router: Cisco
202 likes GRE much better than the other Linux tunneling driver ("IP
203 tunneling" above). In addition, GRE allows multicast redistribution
206 config NET_IPGRE_BROADCAST
207 bool "IP: broadcast GRE over IP"
208 depends on IP_MULTICAST && NET_IPGRE
210 One application of GRE/IP is to construct a broadcast WAN (Wide Area
211 Network), which looks like a normal Ethernet LAN (Local Area
212 Network), but can be distributed all over the Internet. If you want
213 to do that, say Y here and to "IP multicast routing" below.
215 config IP_MROUTE_COMMON
217 depends on IP_MROUTE || IPV6_MROUTE
220 bool "IP: multicast routing"
221 depends on IP_MULTICAST
222 select IP_MROUTE_COMMON
224 This is used if you want your machine to act as a router for IP
225 packets that have several destination addresses. It is needed on the
226 MBONE, a high bandwidth network on top of the Internet which carries
227 audio and video broadcasts. In order to do that, you would most
228 likely run the program mrouted. If you haven't heard about it, you
231 config IP_MROUTE_MULTIPLE_TABLES
232 bool "IP: multicast policy routing"
233 depends on IP_MROUTE && IP_ADVANCED_ROUTER
236 Normally, a multicast router runs a userspace daemon and decides
237 what to do with a multicast packet based on the source and
238 destination addresses. If you say Y here, the multicast router
239 will also be able to take interfaces and packet marks into
240 account and run multiple instances of userspace daemons
241 simultaneously, each one handling a single table.
246 bool "IP: PIM-SM version 1 support"
249 Kernel side support for Sparse Mode PIM (Protocol Independent
250 Multicast) version 1. This multicast routing protocol is used widely
251 because Cisco supports it. You need special software to use it
252 (pimd-v1). Please see <http://netweb.usc.edu/pim/> for more
253 information about PIM.
255 Say Y if you want to use PIM-SM v1. Note that you can say N here if
256 you just want to use Dense Mode PIM.
259 bool "IP: PIM-SM version 2 support"
262 Kernel side support for Sparse Mode PIM version 2. In order to use
263 this, you need an experimental routing daemon supporting it (pimd or
264 gated-5). This routing protocol is not used widely, so say N unless
265 you want to play with it.
268 bool "IP: TCP syncookie support"
270 Normal TCP/IP networking is open to an attack known as "SYN
271 flooding". This denial-of-service attack prevents legitimate remote
272 users from being able to connect to your computer during an ongoing
273 attack and requires very little work from the attacker, who can
274 operate from anywhere on the Internet.
276 SYN cookies provide protection against this type of attack. If you
277 say Y here, the TCP/IP stack will use a cryptographic challenge
278 protocol known as "SYN cookies" to enable legitimate users to
279 continue to connect, even when your machine is under attack. There
280 is no need for the legitimate users to change their TCP/IP software;
281 SYN cookies work transparently to them. For technical information
282 about SYN cookies, check out <http://cr.yp.to/syncookies.html>.
284 If you are SYN flooded, the source address reported by the kernel is
285 likely to have been forged by the attacker; it is only reported as
286 an aid in tracing the packets to their actual source and should not
287 be taken as absolute truth.
289 SYN cookies may prevent correct error reporting on clients when the
290 server is really overloaded. If this happens frequently better turn
293 If you say Y here, you can disable SYN cookies at run time by
294 saying Y to "/proc file system support" and
295 "Sysctl support" below and executing the command
297 echo 0 > /proc/sys/net/ipv4/tcp_syncookies
299 after the /proc file system has been mounted.
304 tristate "Virtual (secure) IP: tunneling"
305 depends on IPV6 || IPV6=n
308 depends on INET_XFRM_MODE_TUNNEL
310 Tunneling means encapsulating data of one protocol type within
311 another protocol and sending it over a channel that understands the
312 encapsulating protocol. This can be used with xfrm mode tunnel to give
313 the notion of a secure tunnel for IPSEC and then use routing protocol
316 config NET_UDP_TUNNEL
322 tristate "IP: Foo (IP protocols) over UDP"
324 select NET_UDP_TUNNEL
326 Foo over UDP allows any IP protocol to be directly encapsulated
327 over UDP include tunnels (IPIP, GRE, SIT). By encapsulating in UDP
328 network mechanisms and optimizations for UDP (such as ECMP
329 and RSS) can be leveraged to provide better service.
331 config NET_FOU_IP_TUNNELS
332 bool "IP: FOU encapsulation of IP tunnels"
333 depends on NET_IPIP || NET_IPGRE || IPV6_SIT
336 Allow configuration of FOU or GUE encapsulation for IP tunnels.
337 When this option is enabled IP tunnels can be configured to use
338 FOU or GUE encapsulation.
341 tristate "IP: AH transformation"
348 Support for IPsec AH.
353 tristate "IP: ESP transformation"
356 select CRYPTO_AUTHENC
362 select CRYPTO_ECHAINIV
364 Support for IPsec ESP.
368 config INET_ESP_OFFLOAD
369 tristate "IP: ESP transformation offload"
374 Support for ESP transformation offload. This makes sense
375 only if this system really does IPsec and want to do it
376 with high throughput. A typical desktop system does not
377 need it, even if it does IPsec.
382 tristate "IP: IPComp transformation"
383 select INET_XFRM_TUNNEL
386 Support for IP Payload Compression Protocol (IPComp) (RFC3173),
387 typically needed for IPsec.
391 config INET_TABLE_PERTURB_ORDER
392 int "INET: Source port perturbation table size (as power of 2)" if EXPERT
395 Source port perturbation table size (as power of 2) for
396 RFC 6056 3.3.4. Algorithm 4: Double-Hash Port Selection Algorithm.
398 The default is almost always what you want.
399 Only change this if you know what you are doing.
401 config INET_XFRM_TUNNEL
410 config INET_XFRM_MODE_TRANSPORT
411 tristate "IP: IPsec transport mode"
415 Support for IPsec transport mode.
419 config INET_XFRM_MODE_TUNNEL
420 tristate "IP: IPsec tunnel mode"
424 Support for IPsec tunnel mode.
428 config INET_XFRM_MODE_BEET
429 tristate "IP: IPsec BEET mode"
433 Support for IPsec BEET mode.
438 tristate "INET: socket monitoring interface"
441 Support for INET (TCP, DCCP, etc) socket monitoring interface used by
442 native Linux tools such as ss. ss is included in iproute2, currently
445 http://www.linuxfoundation.org/collaborate/workgroups/networking/iproute2
451 def_tristate INET_DIAG
454 tristate "UDP: socket monitoring interface"
455 depends on INET_DIAG && (IPV6 || IPV6=n)
458 Support for UDP socket monitoring interface used by the ss tool.
462 tristate "RAW: socket monitoring interface"
463 depends on INET_DIAG && (IPV6 || IPV6=n)
466 Support for RAW socket monitoring interface used by the ss tool.
469 config INET_DIAG_DESTROY
470 bool "INET: allow privileged process to administratively close sockets"
474 Provides a SOCK_DESTROY operation that allows privileged processes
475 (e.g., a connection manager or a network administration tool such as
476 ss) to close sockets opened by other processes. Closing a socket in
477 this way interrupts any blocking read/write/connect operations on
478 the socket and causes future socket calls to behave as if the socket
479 had been disconnected.
482 menuconfig TCP_CONG_ADVANCED
483 bool "TCP: advanced congestion control"
485 Support for selection of various TCP congestion control
488 Nearly all users can safely say no here, and a safe default
489 selection will be made (CUBIC with new Reno as a fallback).
496 tristate "Binary Increase Congestion (BIC) control"
499 BIC-TCP is a sender-side only change that ensures a linear RTT
500 fairness under large windows while offering both scalability and
501 bounded TCP-friendliness. The protocol combines two schemes
502 called additive increase and binary search increase. When the
503 congestion window is large, additive increase with a large
504 increment ensures linear RTT fairness as well as good
505 scalability. Under small congestion windows, binary search
506 increase provides TCP friendliness.
507 See http://www.csc.ncsu.edu/faculty/rhee/export/bitcp/
509 config TCP_CONG_CUBIC
513 This is version 2.0 of BIC-TCP which uses a cubic growth function
514 among other techniques.
515 See http://www.csc.ncsu.edu/faculty/rhee/export/bitcp/cubic-paper.pdf
517 config TCP_CONG_WESTWOOD
518 tristate "TCP Westwood+"
521 TCP Westwood+ is a sender-side only modification of the TCP Reno
522 protocol stack that optimizes the performance of TCP congestion
523 control. It is based on end-to-end bandwidth estimation to set
524 congestion window and slow start threshold after a congestion
525 episode. Using this estimation, TCP Westwood+ adaptively sets a
526 slow start threshold and a congestion window which takes into
527 account the bandwidth used at the time congestion is experienced.
528 TCP Westwood+ significantly increases fairness wrt TCP Reno in
529 wired networks and throughput over wireless links.
535 H-TCP is a send-side only modifications of the TCP Reno
536 protocol stack that optimizes the performance of TCP
537 congestion control for high speed network links. It uses a
538 modeswitch to change the alpha and beta parameters of TCP Reno
539 based on network conditions and in a way so as to be fair with
540 other Reno and H-TCP flows.
542 config TCP_CONG_HSTCP
543 tristate "High Speed TCP"
546 Sally Floyd's High Speed TCP (RFC 3649) congestion control.
547 A modification to TCP's congestion control mechanism for use
548 with large congestion windows. A table indicates how much to
549 increase the congestion window by when an ACK is received.
550 For more detail see http://www.icir.org/floyd/hstcp.html
552 config TCP_CONG_HYBLA
553 tristate "TCP-Hybla congestion control algorithm"
556 TCP-Hybla is a sender-side only change that eliminates penalization of
557 long-RTT, large-bandwidth connections, like when satellite legs are
558 involved, especially when sharing a common bottleneck with normal
559 terrestrial connections.
561 config TCP_CONG_VEGAS
565 TCP Vegas is a sender-side only change to TCP that anticipates
566 the onset of congestion by estimating the bandwidth. TCP Vegas
567 adjusts the sending rate by modifying the congestion
568 window. TCP Vegas should provide less packet loss, but it is
569 not as aggressive as TCP Reno.
575 TCP NV is a follow up to TCP Vegas. It has been modified to deal with
576 10G networks, measurement noise introduced by LRO, GRO and interrupt
577 coalescence. In addition, it will decrease its cwnd multiplicatively
580 Note that in general congestion avoidance (cwnd decreased when # packets
581 queued grows) cannot coexist with congestion control (cwnd decreased only
582 when there is packet loss) due to fairness issues. One scenario when they
583 can coexist safely is when the CA flows have RTTs << CC flows RTTs.
585 For further details see http://www.brakmo.org/networking/tcp-nv/
587 config TCP_CONG_SCALABLE
588 tristate "Scalable TCP"
591 Scalable TCP is a sender-side only change to TCP which uses a
592 MIMD congestion control algorithm which has some nice scaling
593 properties, though is known to have fairness issues.
594 See http://www.deneholme.net/tom/scalable/
597 tristate "TCP Low Priority"
600 TCP Low Priority (TCP-LP), a distributed algorithm whose goal is
601 to utilize only the excess network bandwidth as compared to the
602 ``fair share`` of bandwidth as targeted by TCP.
603 See http://www-ece.rice.edu/networks/TCP-LP/
609 TCP Veno is a sender-side only enhancement of TCP to obtain better
610 throughput over wireless networks. TCP Veno makes use of state
611 distinguishing to circumvent the difficult judgment of the packet loss
612 type. TCP Veno cuts down less congestion window in response to random
614 See <http://ieeexplore.ieee.org/xpl/freeabs_all.jsp?arnumber=1177186>
618 select TCP_CONG_VEGAS
621 YeAH-TCP is a sender-side high-speed enabled TCP congestion control
622 algorithm, which uses a mixed loss/delay approach to compute the
623 congestion window. It's design goals target high efficiency,
624 internal, RTT and Reno fairness, resilience to link loss while
625 keeping network elements load as low as possible.
627 For further details look here:
628 http://wil.cs.caltech.edu/pfldnet2007/paper/YeAH_TCP.pdf
630 config TCP_CONG_ILLINOIS
631 tristate "TCP Illinois"
634 TCP-Illinois is a sender-side modification of TCP Reno for
635 high speed long delay links. It uses round-trip-time to
636 adjust the alpha and beta parameters to achieve a higher average
637 throughput and maintain fairness.
639 For further details see:
640 http://www.ews.uiuc.edu/~shaoliu/tcpillinois/index.html
642 config TCP_CONG_DCTCP
643 tristate "DataCenter TCP (DCTCP)"
646 DCTCP leverages Explicit Congestion Notification (ECN) in the network to
647 provide multi-bit feedback to the end hosts. It is designed to provide:
649 - High burst tolerance (incast due to partition/aggregate),
650 - Low latency (short flows, queries),
651 - High throughput (continuous data updates, large file transfers) with
652 commodity, shallow-buffered switches.
654 All switches in the data center network running DCTCP must support
655 ECN marking and be configured for marking when reaching defined switch
656 buffer thresholds. The default ECN marking threshold heuristic for
657 DCTCP on switches is 20 packets (30KB) at 1Gbps, and 65 packets
658 (~100KB) at 10Gbps, but might need further careful tweaking.
660 For further details see:
661 http://simula.stanford.edu/~alizade/Site/DCTCP_files/dctcp-final.pdf
664 tristate "CAIA Delay-Gradient (CDG)"
667 CAIA Delay-Gradient (CDG) is a TCP congestion control that modifies
668 the TCP sender in order to:
670 o Use the delay gradient as a congestion signal.
671 o Back off with an average probability that is independent of the RTT.
672 o Coexist with flows that use loss-based congestion control.
673 o Tolerate packet loss unrelated to congestion.
675 For further details see:
676 D.A. Hayes and G. Armitage. "Revisiting TCP congestion control using
677 delay gradients." In Networking 2011. Preprint: http://goo.gl/No3vdg
684 BBR (Bottleneck Bandwidth and RTT) TCP congestion control aims to
685 maximize network utilization and minimize queues. It builds an explicit
686 model of the the bottleneck delivery rate and path round-trip
687 propagation delay. It tolerates packet loss and delay unrelated to
688 congestion. It can operate over LAN, WAN, cellular, wifi, or cable
689 modem links. It can coexist with flows that use loss-based congestion
690 control, and can operate with shallow buffers, deep buffers,
691 bufferbloat, policers, or AQM schemes that do not provide a delay
692 signal. It requires the fq ("Fair Queue") pacing packet scheduler.
695 prompt "Default TCP congestion control"
696 default DEFAULT_CUBIC
698 Select the TCP congestion control that will be used by default
702 bool "Bic" if TCP_CONG_BIC=y
705 bool "Cubic" if TCP_CONG_CUBIC=y
708 bool "Htcp" if TCP_CONG_HTCP=y
711 bool "Hybla" if TCP_CONG_HYBLA=y
714 bool "Vegas" if TCP_CONG_VEGAS=y
717 bool "Veno" if TCP_CONG_VENO=y
719 config DEFAULT_WESTWOOD
720 bool "Westwood" if TCP_CONG_WESTWOOD=y
723 bool "DCTCP" if TCP_CONG_DCTCP=y
726 bool "CDG" if TCP_CONG_CDG=y
729 bool "BBR" if TCP_CONG_BBR=y
737 config TCP_CONG_CUBIC
739 depends on !TCP_CONG_ADVANCED
742 config DEFAULT_TCP_CONG
744 default "bic" if DEFAULT_BIC
745 default "cubic" if DEFAULT_CUBIC
746 default "htcp" if DEFAULT_HTCP
747 default "hybla" if DEFAULT_HYBLA
748 default "vegas" if DEFAULT_VEGAS
749 default "westwood" if DEFAULT_WESTWOOD
750 default "veno" if DEFAULT_VENO
751 default "reno" if DEFAULT_RENO
752 default "dctcp" if DEFAULT_DCTCP
753 default "cdg" if DEFAULT_CDG
754 default "bbr" if DEFAULT_BBR
758 bool "TCP: MD5 Signature Option support (RFC2385)"
762 RFC2385 specifies a method of giving MD5 protection to TCP sessions.
763 Its main (only?) use is to protect BGP sessions between core routers