grfix

[thirdparty/man-pages.git] / man7 / ip.7

'\" t
.\" Don't change the line above. it tells man that tbl is needed.
.\" This man page is Copyright (C) 1999 Andi Kleen <ak@muc.de>.
.\" 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.
.\" $Id: ip.7,v 1.19 2000/12/20 18:10:31 ak Exp $
.TH IP  7 2001-06-19 "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(PF_INET, SOCK_STREAM, 0);"
.br
.IB udp_socket " = socket(PF_INET, SOCK_DGRAM, 0);"
.br
.IB raw_socket " = socket(PF_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
.BR "socket(PF_INET, socket_type, protocol)" .
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
.B 0
and
.B IPPROTO_TCP
for TCP sockets and
.B 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
.\" FIXME ip current does an autobind in listen, but I'm not sure
.\" if that should be documented.
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).
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)
or
.BR connect (2)
are called on an unbound socket, it is automatically bound to a
random free 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 */
    uint16_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 "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 only implemented 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 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 only sent or received when the
.B SO_BROADCAST
socket flag is set.
In the current implementation connection oriented sockets are only allowed
to use 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.
.\"
.\" FIXME Document IP_FREEBIND
.\"
.TP
.B IP_OPTIONS
Sets 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
.B accept_source_route
sysctl.
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
.B IP_PKTINFO
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
.B IP_RECVTOS
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
.B IP_RECVTTL
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
.B IP_RECVOPTS
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
.B IP_RETOPTS
Identical to
.B IP_RECVOPTS
but returns raw unprocessed options with timestamp and route record
options not filled in for this hop.
.TP
.B IP_TOS
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)).
.TP
.B IP_TTL
Set or retrieve the current time to live field that is used in every packet
sent from this socket.
.TP
.B IP_HDRINCL
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_RECVERR " (defined in \fI<linux/errqueue.h>\fP)"
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 a 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 SO_EE_OFFENDER address is
.B AF_UNSPEC
when the source was unknown.
When the error originated from the network, all IP options
.RI ( 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 illegal 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
.B IP_MTU_DISCOVER
Sets or receives the Path MTU Discovery setting
for a socket.
When enabled, Linux will perform Path MTU Discovery
as defined in RFC\ 1191
on this socket.
The don't fragment flag is set on all outgoing datagrams.
The system-wide default is controlled by the
.B ip_no_pmtu_disc
sysctl for
.B SOCK_STREAM
sockets, and disabled on all others.
For non
.B SOCK_STREAM
sockets 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 packets that are bigger than the known
path MTU if this flag is set (with
.B EMSGSIZE
).
.TS
tab(:);
c l
l l.
Path MTU discovery flags: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 a
.B EMSGSIZE
error occurred).
It may change over time.
For connectionless sockets with many destinations
the new also 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 .
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
.B IP_MTU
Retrieve the current known path MTU of the current socket.
Only valid when the socket has been connected.
Returns an integer.
Only valid as a
.BR getsockopt (2).
.\"
.TP
.B IP_ROUTER_ALERT
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 users responsibility to send them out again.
Socket binding is ignored,
such packets are only filtered by protocol.
Expects an integer flag.
.\"
.TP
.B IP_MULTICAST_TTL
Set or reads 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
.B IP_MULTICAST_LOOP
Sets or reads a boolean integer argument whether sent multicast
packets should be looped back to the local sockets.
.\"
.TP
.B IP_ADD_MEMBERSHIP
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 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
For compatibility, the old
.I ip_mreq
structure 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
.B IP_DROP_MEMBERSHIP
Leave a multicast group.
Argument is an
.I ip_mreqn
or
.I ip_mreq
structure similar to
.BR IP_ADD_MEMBERSHIP .
.\"
.TP
.B IP_MULTICAST_IF
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.
.SS Sysctls
The IP protocol
supports the sysctl interface to configure some global options.
The sysctls can be accessed by reading or writing the
.I /proc/sys/net/ipv4/*
files or using the
.\" FIXME As at 2.6.12, 14 Jun 2005, the following are undocumented:
.\"     ip_queue_maxlen
.\"     ip_conntrack_max
.BR sysctl (2)
interface.
Variables described as
.I Boolean
take an integer value, with a non-zero value ("true") meaning that
the corresponding option is enabled, and a zero value ("false")
meaning that the option is disabled.
.\"
.TP
.BR ip_always_defrag " (Boolean)"
[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 frag 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
.B ip_autoconfig
.\" FIXME document ip_autoconfig
Not documented.
.\"
.TP
.BR ip_default_ttl " (integer; default: 64)"
Set the default time-to-live value of outgoing packets.
This can be changed per socket with the
.B IP_TTL
option.
.\"
.TP
.BR ip_dynaddr " (Boolean; default: disabled)"
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
.BR ip_forward " (Boolean; default: disabled)"
Enable IP forwarding with a boolean flag.
IP forwarding can be also set on a per interface basis.
.\"
.TP
.B ip_local_port_range
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 >1024, better >4096 to avoid clashes
with well known ports and to minimize firewall problems.
.\"
.TP
.BR ip_no_pmtu_disc " (Boolean; default: disabled)"
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
.BR ip_nonlocal_bind " (Boolean; default: disabled)"
If set, allows processes to
.BR bind (2)
to non-local 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
.BR 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
.BR ip6frag_secret_interval " (integer; default 600)"
Regeneration interval (in seconds) of the hash secret (or lifetime
for the hash secret) for IPv6 fragments.
.TP
.BR ipfrag_high_thresh " (integer), " ipfrag_low_thresh " (integer)"
If the amount of queued IP fragments reaches
.BR ipfrag_high_thresh ,
the queue
is pruned down to
.BR ipfrag_low_thresh .
Contains an integer with the number of
bytes.
.TP
.B neigh/*
See
.BR arp (7).
.\" FIXME Document the conf/*/* sysctls
.\" FIXME Document the route/* sysctls
.\" FIXME document them all
.SS Ioctls
All ioctls described in
.BR socket (7)
apply to 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 reserved 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 non-existent interface was requested or the requested source
address was
not local.
.TP
.B EAGAIN
Operation on a non-blocking socket would block.
.TP
.B EALREADY
An connection operation on a non-blocking 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 only defined 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 VERSIONS
.BR IP_MTU ,
.BR IP_MTU_DISCOVER ,
.BR IP_PKTINFO ,
.B IP_RECVERR
and
.B IP_ROUTER_ALERT
are new options in Linux 2.2.
They are also all Linux-specific and should not be used in
programs intended to be portable.
.PP
.\" FIXME
.\" To be confirmed that IP_PMTUDISC_PROBE makes it into kernel 2.6.22
.B IP_PMTUDISC_PROBE
is new in Linux 2.6.22.
.PP
.I struct ip_mreqn
is new in Linux 2.2.
Linux 2.0 only supported
.BR ip_mreq .
.PP
The sysctls were introduced with Linux 2.2.
.SH NOTES
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(PF_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(PF_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 netlink (7),
.BR raw (7),
.BR socket (7),
.BR tcp (7),
.BR udp (7)
.PP
RFC\ 791 for the original IP specification.
.br
RFC\ 1122 for the IPv4 host requirements.
.br
RFC\ 1812 for the IPv4 router requirements.
.\" FIXME autobind INADDR REUSEADDR