path: root/man7/ip.7

'\" t
.\" This man page is Copyright (C) 1999 Andi Kleen <ak@muc.de>.
.\"
.\" %%%LICENSE_START(VERBATIM_ONE_PARA)
.\" Permission is granted to distribute possibly modified copies
.\" of this page provided the header is included verbatim,
.\" and in case of nontrivial modification author and date
.\" of the modification is added to the header.
.\" %%%LICENSE_END
.\"
.\" $Id: ip.7,v 1.19 2000/12/20 18:10:31 ak Exp $
.\"
.\" FIXME: The following socket options are yet to be documented
.\" 	IP_XFRM_POLICY (2.5.48)
.\"	    Needs CAP_NET_ADMIN
.\" 	IP_IPSEC_POLICY (2.5.47)
.\"	    Needs CAP_NET_ADMIN
.\" 	IP_PASSSEC (2.6.17)
.\" 	    Boolean
.\"	    commit 2c7946a7bf45ae86736ab3b43d0085e43947945c
.\"	    Author: Catherine Zhang <cxzhang@watson.ibm.com>
.\"	IP_MINTTL (2.6.34)
.\"	    commit d218d11133d888f9745802146a50255a4781d37a
.\"	    Author: Stephen Hemminger <shemminger@vyatta.com>
.\"	MCAST_JOIN_GROUP (2.4.22 / 2.6)
.\"	MCAST_BLOCK_SOURCE (2.4.22 / 2.6)
.\"	MCAST_UNBLOCK_SOURCE (2.4.22 / 2.6)
.\"	MCAST_LEAVE_GROUP (2.4.22 / 2.6)
.\"	MCAST_JOIN_SOURCE_GROUP (2.4.22 / 2.6)
.\"	MCAST_LEAVE_SOURCE_GROUP (2.4.22 / 2.6)
.\"	MCAST_MSFILTER (2.4.22 / 2.6)
.\"	IP_MULTICAST_ALL (2.6.31)
.\"	    commit f771bef98004d9d141b085d987a77d06669d4f4f
.\"	    Author: Nivedita Singhvi <niv@us.ibm.com>
.\"	IP_UNICAST_IF (3.4)
.\"	    commit 76e21053b5bf33a07c76f99d27a74238310e3c71
.\"	    Author: Erich E. Hoover <ehoover@mines.edu>
.\"
.TH IP  7 2013-02-12 "Linux" "Linux Programmer's Manual"
.SH NAME
ip \- Linux IPv4 protocol implementation
.SH SYNOPSIS
.B #include <sys/socket.h>
.br
.\" .B #include <net/netinet.h> -- does not exist anymore
.\" .B #include <linux/errqueue.h> -- never include <linux/foo.h>
.B #include <netinet/in.h>
.br
.B #include <netinet/ip.h>        \fR/* superset of previous */
.sp
.IB tcp_socket " = socket(AF_INET, SOCK_STREAM, 0);"
.br
.IB udp_socket " = socket(AF_INET, SOCK_DGRAM, 0);"
.br
.IB raw_socket " = socket(AF_INET, SOCK_RAW, " protocol ");"
.SH DESCRIPTION
Linux implements the Internet Protocol, version 4,
described in RFC\ 791 and RFC\ 1122.
.B ip
contains a level 2 multicasting implementation conforming to RFC\ 1112.
It also contains an IP router including a packet filter.
.\" FIXME has someone verified that 2.1 is really 1812 compliant?
.PP
The programming interface is BSD-sockets compatible.
For more information on sockets, see
.BR socket (7).
.PP
An IP socket is created by calling the
.BR socket (2)
function as
.BI "socket(AF_INET, " socket_type ", " protocol ) \fR.
Valid socket types are
.B SOCK_STREAM
to open a
.BR tcp (7)
socket,
.B SOCK_DGRAM
to open a
.BR udp (7)
socket, or
.B SOCK_RAW
to open a
.BR raw (7)
socket to access the IP protocol directly.
.I protocol
is the IP protocol in the IP header to be received or sent.
The only valid values for
.I protocol
are 0 and
.B IPPROTO_TCP
for TCP sockets, and 0 and
.B IPPROTO_UDP
for UDP sockets.
For
.B SOCK_RAW
you may specify a valid IANA IP protocol defined in
RFC\ 1700 assigned numbers.
.PP
When a process wants to receive new incoming packets or connections, it
should bind a socket to a local interface address using
.BR bind (2).
In this case, only one IP socket may be bound to any given local
(address, port) pair.
When
.B INADDR_ANY
is specified in the bind call, the socket will be bound to
.I all
local interfaces.
When
.BR listen (2)
is called on an unbound socket, the socket is automatically bound
to a random free port with the local address set to
.BR INADDR_ANY .
When
.BR connect (2)
is called on an unbound socket, the socket is automatically bound
to a random free port or to a usable shared port with the local address
set to
.BR INADDR_ANY .

A TCP local socket address that has been bound is unavailable for
some time after closing, unless the
.B SO_REUSEADDR
flag has been set.
Care should be taken when using this flag as it makes TCP less reliable.
.SS Address format
An IP socket address is defined as a combination of an IP interface
address and a 16-bit port number.
The basic IP protocol does not supply port numbers, they
are implemented by higher level protocols like
.BR udp (7)
and
.BR tcp (7).
On raw sockets
.I sin_port
is set to the IP protocol.
.PP
.in +4n
.nf
struct sockaddr_in {
    sa_family_t    sin_family; /* address family: AF_INET */
    in_port_t      sin_port;   /* port in network byte order */
    struct in_addr sin_addr;   /* internet address */
};

/* Internet address. */
struct in_addr {
    uint32_t       s_addr;     /* address in network byte order */
};
.fi
.in
.PP
.I sin_family
is always set to
.BR AF_INET .
This is required; in Linux 2.2 most networking functions return
.B EINVAL
when this setting is missing.
.I sin_port
contains the port in network byte order.
The port numbers below 1024 are called
.IR "privileged ports"
(or sometimes:
.IR "reserved ports" ).
Only privileged processes (i.e., those having the
.B CAP_NET_BIND_SERVICE
capability) may
.BR bind (2)
to these sockets.
Note that the raw IPv4 protocol as such has no concept of a
port, they are implemented only by higher protocols like
.BR tcp (7)
and
.BR udp (7).
.PP
.I sin_addr
is the IP host address.
The
.I s_addr
member of
.I struct in_addr
contains the host interface address in network byte order.
.I in_addr
should be assigned one of the
.BR INADDR_*
values (e.g.,
.BR INADDR_ANY )
or set using the
.BR inet_aton (3),
.BR inet_addr (3),
.BR inet_makeaddr (3)
library functions or directly with the name resolver (see
.BR gethostbyname (3)).

IPv4 addresses are divided into unicast, broadcast
and multicast addresses.
Unicast addresses specify a single interface of a host,
broadcast addresses specify all hosts on a network and multicast
addresses address all hosts in a multicast group.
Datagrams to broadcast addresses can be sent or received only when the
.B SO_BROADCAST
socket flag is set.
In the current implementation, connection-oriented sockets are allowed
to use only unicast addresses.
.\" Leave a loophole for XTP @)

Note that the address and the port are always stored in
network byte order.
In particular, this means that you need to call
.BR htons (3)
on the number that is assigned to a port.
All address/port manipulation
functions in the standard library work in network byte order.

There are several special addresses:
.B INADDR_LOOPBACK
(127.0.0.1)
always refers to the local host via the loopback device;
.B INADDR_ANY
(0.0.0.0)
means any address for binding;
.B INADDR_BROADCAST
(255.255.255.255)
means any host and has the same effect on bind as
.B INADDR_ANY
for historical reasons.
.SS Socket options
IP supports some protocol-specific socket options that can be set with
.BR setsockopt (2)
and read with
.BR getsockopt (2).
The socket option level for IP is
.BR IPPROTO_IP .
.\" or SOL_IP on Linux
A boolean integer flag is zero when it is false, otherwise true.
.TP
.BR IP_ADD_MEMBERSHIP " (since Linux 1.2)"
Join a multicast group.
Argument is an
.I ip_mreqn
structure.
.sp
.in +4n
.nf
struct ip_mreqn {
    struct in_addr imr_multiaddr; /* IP multicast group
                                     address */
    struct in_addr imr_address;   /* IP address of local
                                     interface */
    int            imr_ifindex;   /* interface index */
};
.fi
.in
.sp
.I imr_multiaddr
contains the address of the multicast group the application
wants to join or leave.
It must be a valid multicast address
.\" (i.e., within the 224.0.0.0-239.255.255.255 range)
(or
.BR setsockopt (2)
fails with the error
.BR EINVAL ).
.I imr_address
is the address of the local interface with which the system
should join the multicast group; if it is equal to
.B INADDR_ANY
an appropriate interface is chosen by the system.
.I imr_ifindex
is the interface index of the interface that should join/leave the
.I imr_multiaddr
group, or 0 to indicate any interface.
.IP
The
.I ip_mreqn
structure is available only since Linux 2.2.
For compatibility, the old
.I ip_mreq
structure (present since Linux 1.2) is still supported;
it differs from
.I ip_mreqn
only by not including the
.I imr_ifindex
field.
Only valid as a
.BR setsockopt (2).
.\"
.TP
.BR IP_ADD_SOURCE_MEMBERSHIP " (since Linux 2.4.22 / 2.5.68)"
Join a multicast group and allow receiving data only
from a specified source.
Argument is an
.I ip_mreq_source
structure.
.sp
.in +4n
.nf
struct ip_mreq_source {
    struct in_addr imr_multiaddr;  /* IP multicast group
                                      address */
    struct in_addr imr_interface;  /* IP address of local
                                      interface */
    struct in_addr imr_sourceaddr; /* IP address of
                                      multicast source */
};
.fi
.in
.sp
The
.I ip_mreq_source
structure is similar to
.I ip_mreqn
described under
.BR IP_ADD_MEMBERSIP .
The
.I imr_multiaddr
field contains the address of the multicast group the application
wants to join or leave.
The
.I imr_interface
field is the address of the local interface with which
the system should join the multicast group.
Finally, the
.I imr_sourceaddr
field contains the address of the source the
application wants to receive data from.
.IP
This option can be used multiple times to allow
receiving data from more than one source.
.TP
.BR IP_BLOCK_SOURCE " (since Linux 2.4.22 / 2.5.68)"
Stop receiving multicast data from a specific source in a given group.
This is valid only after the application has subscribed
to the multicast group using either
.BR IP_ADD_MEMBERSHIP
or
.BR IP_ADD_SOURCE_MEMBERSHIP .
.IP
Argument is an
.I ip_mreq_source
structure as described under
.BR IP_ADD_SOURCE_MEMBERSHIP .
.TP
.BR IP_DROP_MEMBERSHIP " (since Linux 1.2)"
Leave a multicast group.
Argument is an
.I ip_mreqn
or
.I ip_mreq
structure similar to
.BR IP_ADD_MEMBERSHIP .
.TP
.BR IP_DROP_SOURCE_MEMBERSHIP " (since Linux 2.4.22 / 2.5.68)"
Leave a source-specific group\(emthat is, stop receiving data from
a given multicast group that come from a given source.
If the application has subscribed to multiple sources within
the same group, data from the remaining sources will still be delivered.
To stop receiving data from all sources at once, use
.BR IP_LEAVE_GROUP .
.IP
Argument is an
.I ip_mreq_source
structure as described under
.BR IP_ADD_SOURCE_MEMBERSHIP .
.TP
.BR IP_FREEBIND " (since Linux 2.4)"
.\" Precisely: 2.4.0-test10
If enabled, this boolean option allows binding to an IP address
that is nonlocal or does not (yet) exist.
This permits listening on a socket,
without requiring the underlying network interface or the
specified dynamic IP address to be up at the time that
the application is trying to bind to it.
This option is the per-socket equivalent of the
.IR ip_nonlocal_bind
.I /proc
interface described below.
.TP
.BR IP_HDRINCL " (since Linux 2.0)"
If enabled,
the user supplies an IP header in front of the user data.
Only valid for
.B SOCK_RAW
sockets.
See
.BR raw (7)
for more information.
When this flag is enabled the values set by
.BR IP_OPTIONS ,
.B IP_TTL
and
.B IP_TOS
are ignored.
.TP
.BR IP_MSFILTER " (since Linux 2.4.22 / 2.5.68)"
This option provides access to the advanced full-state filtering API.
Argument is an
.I ip_msfilter
structure.
.sp
.in +4n
.nf
struct ip_msfilter {
    struct in_addr imsf_multiaddr; /* IP multicast group
                                      address */
    struct in_addr imsf_interface; /* IP address of local
                                      interface */
    uint32_t       imsf_fmode;     /* Filter-mode */

    uint32_t       imsf_numsrc;    /* Number of sources in
                                      the following array */
    struct in_addr imsf_slist[1];  /* Array of source
                                      addresses */
};
.fi
.in
.sp
There are two macros,
.BR MCAST_INCLUDE
and
.BR MCAST_EXCLUDE ,
which can be used to specify the filtering mode.
Additionally, the
.BR IP_MSFILTER_SIZE (n)
macro exists to determine how much memory is needed to store
.I ip_msfilter
structure with
.I n
sources in the source list.
.IP
For the full description of multicast source filtering
refer to RFC 3376.
.TP
.BR IP_MTU " (since Linux 2.2)"
.\" Precisely: 2.1.124
Retrieve the current known path MTU of the current socket.
Valid only when the socket has been connected.
Returns an integer.
Only valid as a
.BR getsockopt (2).
.TP
.BR IP_MTU_DISCOVER " (since Linux 2.2)"
.\" Precisely: 2.1.124
Set or receive the Path MTU Discovery setting for a socket.
When enabled, Linux will perform Path MTU Discovery
as defined in RFC\ 1191 on
.B SOCK_STREAM
sockets.
For
.RB non- SOCK_STREAM
sockets,
.B IP_PMTUDISC_DO
forces the don't-fragment flag to be set on all outgoing packets.
It is the user's responsibility to packetize the data
in MTU-sized chunks and to do the retransmits if necessary.
The kernel will reject (with
.BR EMSGSIZE )
datagrams that are bigger than the known path MTU.
.B IP_PMTUDISC_WANT
will fragment a datagram if needed according to the path MTU,
or will set the don't-fragment flag otherwise.

The system-wide default can be toggled between
.B IP_PMTUDISC_WANT
and
.B IP_PMTUDISC_DONT
by writing (respectively, zero and nonzero values) to the
.I /proc/sys/net/ipv4/ip_no_pmtu_disc
file.
.TS
tab(:);
c l
l l.
Path MTU discovery value:Meaning
IP_PMTUDISC_WANT:Use per-route settings.
IP_PMTUDISC_DONT:Never do Path MTU Discovery.
IP_PMTUDISC_DO:Always do Path MTU Discovery.
IP_PMTUDISC_PROBE:Set DF but ignore Path MTU.
.TE

When PMTU discovery is enabled, the kernel automatically keeps track of
the path MTU per destination host.
When it is connected to a specific peer with
.BR connect (2),
the currently known path MTU can be retrieved conveniently using the
.B IP_MTU
socket option (e.g., after an
.B EMSGSIZE
error occurred).
The path MTU may change over time.
For connectionless sockets with many destinations,
the new MTU for a given destination can also be accessed using the
error queue (see
.BR IP_RECVERR ).
A new error will be queued for every incoming MTU update.

While MTU discovery is in progress, initial packets from datagram sockets
may be dropped.
Applications using UDP should be aware of this and not
take it into account for their packet retransmit strategy.

To bootstrap the path MTU discovery process on unconnected sockets, it
is possible to start with a big datagram size
(up to 64K-headers bytes long) and let it shrink by updates of the path MTU.
.\" FIXME this is an ugly hack

To get an initial estimate of the
path MTU, connect a datagram socket to the destination address using
.BR connect (2)
and retrieve the MTU by calling
.BR getsockopt (2)
with the
.B IP_MTU
option.

It is possible to implement RFC 4821 MTU probing with
.B SOCK_DGRAM
or
.B SOCK_RAW
sockets by setting a value of
.BR IP_PMTUDISC_PROBE
(available since Linux 2.6.22).
This is also particularly useful for diagnostic tools such as
.BR tracepath (8)
that wish to deliberately send probe packets larger than
the observed Path MTU.
.TP
.BR IP_MULTICAST_IF " (since Linux 1.2)"
Set the local device for a multicast socket.
Argument is an
.I ip_mreqn
or
.I ip_mreq
structure similar to
.BR IP_ADD_MEMBERSHIP .
.IP
When an invalid socket option is passed,
.B ENOPROTOOPT
is returned.
.TP
.BR IP_MULTICAST_LOOP " (since Linux 1.2)"
Set or read a boolean integer argument that determines whether
sent multicast packets should be looped back to the local sockets.
.TP
.BR IP_MULTICAST_TTL " (since Linux 1.2)"
Set or read the time-to-live value of outgoing multicast packets for this
socket.
It is very important for multicast packets to set the smallest TTL possible.
The default is 1 which means that multicast packets don't leave the local
network unless the user program explicitly requests it.
Argument is an integer.
.TP
.BR IP_NODEFRAG " (since Linux 2.6.36)"
If enabled (argument is nonzero),
the reassembly of outgoing packets is disabled in the netfilter layer.
This option is valid only for
.B SOCK_RAW
sockets.
The argument is an integer.
.TP
.BR IP_OPTIONS " (since Linux 2.0)"
.\" Precisely: 1.3.30
Set or get the IP options to be sent with every packet from this socket.
The arguments are a pointer to a memory buffer containing the options
and the option length.
The
.BR setsockopt (2)
call sets the IP options associated with a socket.
The maximum option size for IPv4 is 40 bytes.
See RFC\ 791 for the allowed options.
When the initial connection request packet for a
.B SOCK_STREAM
socket contains IP options, the IP options will be set automatically
to the options from the initial packet with routing headers reversed.
Incoming packets are not allowed to change options after the connection
is established.
The processing of all incoming source routing options
is disabled by default and can be enabled by using the
.I accept_source_route
.I /proc
interface.
Other options like timestamps are still handled.
For datagram sockets, IP options can be only set by the local user.
Calling
.BR getsockopt (2)
with
.B IP_OPTIONS
puts the current IP options used for sending into the supplied buffer.
.TP
.BR IP_PKTINFO " (since Linux 2.2)"
.\" Precisely: 2.1.68
Pass an
.B IP_PKTINFO
ancillary message that contains a
.I pktinfo
structure that supplies some information about the incoming packet.
This only works for datagram oriented sockets.
The argument is a flag that tells the socket whether the
.B IP_PKTINFO
message should be passed or not.
The message itself can only be sent/retrieved
as control message with a packet using
.BR recvmsg (2)
or
.BR sendmsg (2).
.IP
.in +4n
.nf
struct in_pktinfo {
    unsigned int   ipi_ifindex;  /* Interface index */
    struct in_addr ipi_spec_dst; /* Local address */
    struct in_addr ipi_addr;     /* Header Destination
                                    address */
};
.fi
.in
.IP
.\" FIXME elaborate on that.
.I ipi_ifindex
is the unique index of the interface the packet was received on.
.I ipi_spec_dst
is the local address of the packet and
.I ipi_addr
is the destination address in the packet header.
If
.B IP_PKTINFO
is passed to
.BR sendmsg (2)
and
.\" This field is grossly misnamed
.I ipi_spec_dst
is not zero, then it is used as the local source address for the routing
table lookup and for setting up IP source route options.
When
.I ipi_ifindex
is not zero, the primary local address of the interface specified by the
index overwrites
.I ipi_spec_dst
for the routing table lookup.
.TP
.BR IP_RECVERR " (since Linux 2.2)"
.\" Precisely: 2.1.15
Enable extended reliable error message passing.
When enabled on a datagram socket, all
generated errors will be queued in a per-socket error queue.
When the user receives an error from a socket operation,
the errors can be received by calling
.BR recvmsg (2)
with the
.B MSG_ERRQUEUE
flag set.
The
.I sock_extended_err
structure describing the error will be passed in an ancillary message with
the type
.B IP_RECVERR
and the level
.BR IPPROTO_IP .
.\" or SOL_IP on Linux
This is useful for reliable error handling on unconnected sockets.
The received data portion of the error queue contains the error packet.
.IP
The
.B IP_RECVERR
control message contains a
.I sock_extended_err
structure:
.IP
.in +4n
.ne 18
.nf
#define SO_EE_ORIGIN_NONE    0
#define SO_EE_ORIGIN_LOCAL   1
#define SO_EE_ORIGIN_ICMP    2
#define SO_EE_ORIGIN_ICMP6   3

struct sock_extended_err {
    uint32_t ee_errno;   /* error number */
    uint8_t  ee_origin;  /* where the error originated */
    uint8_t  ee_type;    /* type */
    uint8_t  ee_code;    /* code */
    uint8_t  ee_pad;
    uint32_t ee_info;    /* additional information */
    uint32_t ee_data;    /* other data */
    /* More data may follow */
};

struct sockaddr *SO_EE_OFFENDER(struct sock_extended_err *);
.fi
.in
.IP
.I ee_errno
contains the
.I errno
number of the queued error.
.I ee_origin
is the origin code of where the error originated.
The other fields are protocol-specific.
The macro
.B SO_EE_OFFENDER
returns a pointer to the address of the network object
where the error originated from given a pointer to the ancillary message.
If this address is not known, the
.I sa_family
member of the
.I sockaddr
contains
.B AF_UNSPEC
and the other fields of the
.I sockaddr
are undefined.
.IP
IP uses the
.I sock_extended_err
structure as follows:
.I ee_origin
is set to
.B SO_EE_ORIGIN_ICMP
for errors received as an ICMP packet, or
.B SO_EE_ORIGIN_LOCAL
for locally generated errors.
Unknown values should be ignored.
.I ee_type
and
.I ee_code
are set from the type and code fields of the ICMP header.
.I ee_info
contains the discovered MTU for
.B EMSGSIZE
errors.
The message also contains the
.I sockaddr_in of the node
caused the error, which can be accessed with the
.B SO_EE_OFFENDER
macro.
The
.I sin_family
field of the
.B SO_EE_OFFENDER
address is
.B AF_UNSPEC
when the source was unknown.
When the error originated from the network, all IP options
.RB ( IP_OPTIONS ", " IP_TTL ", "
etc.) enabled on the socket and contained in the
error packet are passed as control messages.
The payload of the packet causing the error is returned as normal payload.
.\" FIXME . Is it a good idea to document that? It is a dubious feature.
.\" On
.\" .B SOCK_STREAM
.\" sockets,
.\" .B IP_RECVERR
.\" has slightly different semantics. Instead of
.\" saving the errors for the next timeout, it passes all incoming
.\" errors immediately to the user.
.\" This might be useful for very short-lived TCP connections which
.\" need fast error handling. Use this option with care:
.\" it makes TCP unreliable
.\" by not allowing it to recover properly from routing
.\" shifts and other normal
.\" conditions and breaks the protocol specification.
Note that TCP has no error queue;
.B MSG_ERRQUEUE
is not permitted on
.B SOCK_STREAM
sockets.
.B IP_RECVERR
is valid for TCP, but all errors are returned by socket function return or
.B SO_ERROR
only.
.IP
For raw sockets,
.B IP_RECVERR
enables passing of all received ICMP errors to the
application, otherwise errors are only reported on connected sockets
.IP
It sets or retrieves an integer boolean flag.
.B IP_RECVERR
defaults to off.
.TP
.BR IP_RECVOPTS " (since Linux 2.2)"
.\" Precisely: 2.1.15
Pass all incoming IP options to the user in a
.B IP_OPTIONS
control message.
The routing header and other options are already filled in
for the local host.
Not supported for
.B SOCK_STREAM
sockets.
.TP
.BR IP_RECVORIGDSTADDR " (since Linux 2.6.29)"
.\" commit e8b2dfe9b4501ed0047459b2756ba26e5a940a69
This boolean option enables the
.B IP_ORIGDSTADDR
ancillary message in
.BR recvmsg (2),
in which the kernel returns the original destination address
of the datagram being received.
The ancillary message contains a
.IR "struct sockaddr_in" .
.TP
.BR IP_RECVTOS " (since Linux 2.2)"
.\" Precisely: 2.1.68
If enabled the
.B IP_TOS
ancillary message is passed with incoming packets.
It contains a byte which specifies the Type of Service/Precedence
field of the packet header.
Expects a boolean integer flag.
.TP
.BR IP_RECVTTL " (since Linux 2.2)"
.\" Precisely: 2.1.68
When this flag is set, pass a
.B IP_TTL
control message with the time to live
field of the received packet as a byte.
Not supported for
.B SOCK_STREAM
sockets.
.TP
.BR IP_RETOPTS " (since Linux 2.2)"
.\" Precisely: 2.1.15
Identical to
.BR IP_RECVOPTS ,
but returns raw unprocessed options with timestamp and route record
options not filled in for this hop.
.TP
.BR IP_ROUTER_ALERT " (since Linux 2.2)"
.\" Precisely: 2.1.68
Pass all to-be forwarded packets with the
IP Router Alert option set to this socket.
Only valid for raw sockets.
This is useful, for instance, for user-space RSVP daemons.
The tapped packets are not forwarded by the kernel; it is
the user's responsibility to send them out again.
Socket binding is ignored,
such packets are only filtered by protocol.
Expects an integer flag.
.TP
.BR IP_TOS " (since Linux 1.0)"
Set or receive the Type-Of-Service (TOS) field that is sent
with every IP packet originating from this socket.
It is used to prioritize packets on the network.
TOS is a byte.
There are some standard TOS flags defined:
.B IPTOS_LOWDELAY
to minimize delays for interactive traffic,
.B IPTOS_THROUGHPUT
to optimize throughput,
.B IPTOS_RELIABILITY
to optimize for reliability,
.B IPTOS_MINCOST
should be used for "filler data" where slow transmission doesn't matter.
At most one of these TOS values can be specified.
Other bits are invalid and shall be cleared.
Linux sends
.B IPTOS_LOWDELAY
datagrams first by default,
but the exact behavior depends on the configured queueing discipline.
.\" FIXME elaborate on this
Some high priority levels may require superuser privileges (the
.B CAP_NET_ADMIN
capability).
The priority can also be set in a protocol independent way by the
.RB ( SOL_SOCKET ", " SO_PRIORITY )
socket option (see
.BR socket (7)).
.\" Needs CAP_NET_ADMIN
.\" Boolean
.\" Since Linux 2.6.27
.\" Author: KOVACS Krisztian <hidden@sch.bme.hu>
.\" http://lwn.net/Articles/252545/
.TP
.BR IP_TRANSPARENT " (since Linux 2.6.24)"
.\" commit f5715aea4564f233767ea1d944b2637a5fd7cd2e
.\"     This patch introduces the IP_TRANSPARENT socket option: enabling that
.\"     will make the IPv4 routing omit the non-local source address check on
.\"     output. Setting IP_TRANSPARENT requires NET_ADMIN capability.
.\" http://lwn.net/Articles/252545/
Setting this boolean option enables transparent proxying on this socket.
This socket option allows
the calling application to bind to a nonlocal IP address and operate
both as a client and a server with the foreign address as the local endpoint.
NOTE: this requires that routing be set up in a way that
packets going to the foreign address are routed through the TProxy box.
Enabling this socket option requires superuser privileges
(the
.BR CAP_NET_ADMIN
capability).
.IP
TProxy redirection with the iptables TPROXY target also requires that
this option be set on the redirected socket.
.TP
.BR IP_TTL " (since Linux 1.0)"
Set or retrieve the current time-to-live field that is used in every packet
sent from this socket.
.TP
.BR IP_UNBLOCK_SOURCE " (since Linux 2.4.22 / 2.5.68)"
Unblock previously blocked multicast source.
Returns
.BR EADDRNOTAVAIL
when given source is not being blocked.
.IP
Argument is an
.I ip_mreq_source
structure as described under
.BR IP_ADD_SOURCE_MEMBERSHIP .
.SS /proc interfaces
The IP protocol
supports a set of
.I /proc
interfaces to configure some global parameters.
The parameters can be accessed by reading or writing files in the directory
.IR /proc/sys/net/ipv4/ .
.\" FIXME As at 2.6.12, 14 Jun 2005, the following are undocumented:
.\"	ip_queue_maxlen
.\"	ip_conntrack_max
Interfaces described as
.I Boolean
take an integer value, with a nonzero value ("true") meaning that
the corresponding option is enabled, and a zero value ("false")
meaning that the option is disabled.
.\"
.TP
.IR ip_always_defrag " (Boolean; since Linux 2.2.13)"
[New with kernel 2.2.13; in earlier kernel versions this feature
was controlled at compile time by the
.B CONFIG_IP_ALWAYS_DEFRAG
option; this option is not present in 2.4.x and later]

When this boolean flag is enabled (not equal 0), incoming fragments
(parts of IP packets
that arose when some host between origin and destination decided
that the packets were too large and cut them into pieces) will be
reassembled (defragmented) before being processed, even if they are
about to be forwarded.

Only enable if running either a firewall that is the sole link
to your network or a transparent proxy; never ever use it for a
normal router or host.
Otherwise fragmented communication can be disturbed
if the fragments travel over different links.
Defragmentation also has a large memory and CPU time cost.

This is automagically turned on when masquerading or transparent
proxying are configured.
.\"
.TP
.IR ip_autoconfig " (since Linux 2.2 to 2.6.17)"
.\" Precisely: since 2.1.68
.\" FIXME document ip_autoconfig
Not documented.
.\"
.TP
.IR ip_default_ttl " (integer; default: 64; since Linux 2.2)"
.\" Precisely: 2.1.15
Set the default time-to-live value of outgoing packets.
This can be changed per socket with the
.B IP_TTL
option.
.\"
.TP
.IR ip_dynaddr " (Boolean; default: disabled; since Linux 2.0.31)"
Enable dynamic socket address and masquerading entry rewriting on interface
address change.
This is useful for dialup interface with changing IP addresses.
0 means no rewriting, 1 turns it on and 2 enables verbose mode.
.\"
.TP
.IR ip_forward " (Boolean; default: disabled; since Linux 1.2)"
Enable IP forwarding with a boolean flag.
IP forwarding can be also set on a per-interface basis.
.\"
.TP
.IR ip_local_port_range " (since Linux 2.2)"
.\" Precisely: since 2.1.68
Contains two integers that define the default local port range
allocated to sockets.
Allocation starts with the first number and ends with the second number.
Note that these should not conflict with the ports used by masquerading
(although the case is handled).
Also arbitrary choices may cause problems with some firewall packet
filters that make assumptions about the local ports in use.
First number should be at least greater than 1024,
or better, greater than 4096, to avoid clashes
with well known ports and to minimize firewall problems.
.\"
.TP
.IR ip_no_pmtu_disc " (Boolean; default: disabled; since Linux 2.2)"
.\" Precisely: 2.1.15
If enabled, don't do Path MTU Discovery for TCP sockets by default.
Path MTU discovery may fail if misconfigured firewalls (that drop
all ICMP packets) or misconfigured interfaces (e.g., a point-to-point
link where the both ends don't agree on the MTU) are on the path.
It is better to fix the broken routers on the path than to turn off
Path MTU Discovery globally, because not doing it incurs a high cost
to the network.
.\"
.\" The following is from 2.6.12: Documentation/networking/ip-sysctl.txt
.TP
.IR ip_nonlocal_bind " (Boolean; default: disabled; since Linux 2.4)"
.\" Precisely: patch-2.4.0-test10
If set, allows processes to
.BR bind (2)
to nonlocal IP addresses,
which can be quite useful, but may break some applications.
.\"
.\" The following is from 2.6.12: Documentation/networking/ip-sysctl.txt
.TP
.IR ip6frag_time " (integer; default: 30)"
Time in seconds to keep an IPv6 fragment in memory.
.\"
.\" The following is from 2.6.12: Documentation/networking/ip-sysctl.txt
.TP
.IR ip6frag_secret_interval " (integer; default: 600)"
Regeneration interval (in seconds) of the hash secret (or lifetime
for the hash secret) for IPv6 fragments.
.TP
.IR ipfrag_high_thresh " (integer), " ipfrag_low_thresh " (integer)"
If the amount of queued IP fragments reaches
.IR ipfrag_high_thresh ,
the queue is pruned down to
.IR ipfrag_low_thresh .
Contains an integer with the number of bytes.
.TP
.I neigh/*
See
.BR arp (7).
.\" FIXME Document the conf/*/* interfaces
.\" FIXME Document the route/* interfaces
.\" FIXME document them all
.SS Ioctls
All ioctls described in
.BR socket (7)
apply to
.BR ip .
.\" 2006-04-02, mtk
.\" commented out the following because ipchains is obsolete
.\" .PP
.\" The ioctls to configure firewalling are documented in
.\" .BR ipfw (4)
.\" from the
.\" .B ipchains
.\" package.
.PP
Ioctls to configure generic device parameters are described in
.BR netdevice (7).
.\" FIXME Add a discussion of multicasting
.SH ERRORS
.\" FIXME document all errors.
.\"     We should really fix the kernels to give more uniform
.\"     error returns (ENOMEM vs ENOBUFS, EPERM vs EACCES etc.)
.TP
.B EACCES
The user tried to execute an operation without the necessary permissions.
These include:
sending a packet to a broadcast address without having the
.B SO_BROADCAST
flag set;
sending a packet via a
.I prohibit
route;
modifying firewall settings without superuser privileges (the
.B CAP_NET_ADMIN
capability);
binding to a privileged port without superuser privileges (the
.B CAP_NET_BIND_SERVICE
capability).
.TP
.B EADDRINUSE
Tried to bind to an address already in use.
.TP
.B EADDRNOTAVAIL
A nonexistent interface was requested or the requested source
address was not local.
.TP
.B EAGAIN
Operation on a nonblocking socket would block.
.TP
.B EALREADY
An connection operation on a nonblocking socket is already in progress.
.TP
.B ECONNABORTED
A connection was closed during an
.BR accept (2).
.TP
.B EHOSTUNREACH
No valid routing table entry matches the destination address.
This error can be caused by a ICMP message from a remote router or
for the local routing table.
.TP
.B EINVAL
Invalid argument passed.
For send operations this can be caused by sending to a
.I blackhole
route.
.TP
.B EISCONN
.BR connect (2)
was called on an already connected socket.
.TP
.B EMSGSIZE
Datagram is bigger than an MTU on the path and it cannot be fragmented.
.TP
.BR ENOBUFS ", " ENOMEM
Not enough free memory.
This often means that the memory allocation is limited by the socket
buffer limits, not by the system memory, but this is not 100% consistent.
.TP
.B ENOENT
.B SIOCGSTAMP
was called on a socket where no packet arrived.
.TP
.B ENOPKG
A kernel subsystem was not configured.
.TP
.BR ENOPROTOOPT " and " EOPNOTSUPP
Invalid socket option passed.
.TP
.B ENOTCONN
The operation is defined only on a connected socket, but the socket wasn't
connected.
.TP
.B EPERM
User doesn't have permission to set high priority, change configuration,
or send signals to the requested process or group.
.TP
.B EPIPE
The connection was unexpectedly closed or shut down by the other end.
.TP
.B ESOCKTNOSUPPORT
The socket is not configured or an unknown socket type was requested.
.PP
Other errors may be generated by the overlaying protocols; see
.BR tcp (7),
.BR raw (7),
.BR udp (7)
and
.BR socket (7).
.SH NOTES
.BR IP_FREEBIND ,
.BR IP_MSFILTER ,
.BR IP_MTU ,
.BR IP_MTU_DISCOVER ,
.BR IP_RECVORIGDSTADDR ,
.BR IP_PKTINFO ,
.BR IP_RECVERR ,
.BR IP_ROUTER_ALERT ,
and
.BR IP_TRANSPARENT
are Linux-specific.
.\" IP_PASSSEC is Linux-specific
.\" IP_XFRM_POLICY is Linux-specific
.\" IP_IPSEC_POLICY is a nonstandard extension, also present on some BSDs

Be very careful with the
.B SO_BROADCAST
option \- it is not privileged in Linux.
It is easy to overload the network
with careless broadcasts.
For new application protocols
it is better to use a multicast group instead of broadcasting.
Broadcasting is discouraged.
.PP
Some other BSD sockets implementations provide
.B IP_RCVDSTADDR
and
.B IP_RECVIF
socket options to get the destination address and the interface of
received datagrams.
Linux has the more general
.B IP_PKTINFO
for the same task.
.PP
Some BSD sockets implementations also provide an
.B IP_RECVTTL
option, but an ancillary message with type
.B IP_RECVTTL
is passed with the incoming packet.
This is different from the
.B IP_TTL
option used in Linux.
.PP
Using
.B SOL_IP
socket options level isn't portable, BSD-based stacks use
.B IPPROTO_IP
level.
.SS Compatibility
For compatibility with Linux 2.0, the obsolete
.BI "socket(AF_INET, SOCK_PACKET, " protocol )
syntax is still supported to open a
.BR packet (7)
socket.
This is deprecated and should be replaced by
.BI "socket(AF_PACKET, SOCK_RAW, " protocol )
instead.
The main difference is the new
.I sockaddr_ll
address structure for generic link layer information instead of the old
.BR sockaddr_pkt .
.SH BUGS
There are too many inconsistent error values.
.PP
The ioctls to configure IP-specific interface options and ARP tables are
not described.
.PP
Some versions of glibc forget to declare
.IR in_pktinfo .
Workaround currently is to copy it into your program from this man page.
.PP
Receiving the original destination address with
.B MSG_ERRQUEUE
in
.I msg_name
by
.BR recvmsg (2)
does not work in some 2.2 kernels.
.\" .SH AUTHORS
.\" This man page was written by Andi Kleen.
.SH SEE ALSO
.BR recvmsg (2),
.BR sendmsg (2),
.BR byteorder (3),
.BR ipfw (4),
.BR capabilities (7),
.BR icmp (7),
.BR ipv6 (7),
.BR netlink (7),
.BR raw (7),
.BR socket (7),
.BR tcp (7),
.BR udp (7)
.PP
RFC\ 791 for the original IP specification.
RFC\ 1122 for the IPv4 host requirements.
RFC\ 1812 for the IPv4 router requirements.
.\" FIXME autobind INADDR REUSEADDR