tcp — Internet
Transmission Control Protocol
The TCP protocol provides reliable, flow-controlled, two-way
transmission of data. It is a byte-stream protocol used to support the
SOCK_STREAM abstraction. TCP uses the standard
Internet address format and, in addition, provides a per-host collection of
“port addresses”. Thus, each address is composed of an
Internet address specifying the host and network, with a specific TCP port
on the host identifying the peer entity.
Sockets utilizing the TCP protocol are either
“active” or “passive”. Active sockets initiate
connections to passive sockets. By default, TCP sockets are created active;
to create a passive socket, the
listen(2) system call must be used
after binding the socket with the bind(2)
system call. Only passive sockets may use the
accept(2) call to accept incoming
connections. Only active sockets may use the
connectx(2) call to initiate
Passive sockets may “underspecify” their location to
match incoming connection requests from multiple networks. This technique,
termed “wildcard addressing”, allows a single server to
provide service to clients on multiple networks. To create a socket which
listens on all networks, the Internet address
INADDR_ANY must be bound. The TCP port may still be
specified at this time; if the port is not specified, the system will assign
one. Once a connection has been established, the socket's address is fixed
by the peer entity's location. The address assigned to the socket is the
address associated with the network interface through which packets are
being transmitted and received. Normally, this address corresponds to the
peer entity's network.
TCP supports a number of socket options which can be set with
setsockopt(2) and tested with
- Under most circumstances, TCP sends data when it is presented; when
outstanding data has not yet been acknowledged, it gathers small amounts
of output to be sent in a single packet once an acknowledgement is
received. For a small number of clients, such as window systems that send
a stream of mouse events which receive no replies, this packetization may
cause significant delays. The boolean option
TCP_NODELAY defeats this algorithm.
- By default, a sender- and receiver-TCP will
negotiate among themselves to determine the maximum segment size to be
used for each connection. The
allows the user to determine the result of this negotiation, and to reduce
it if desired.
- TCP usually sends a number of options in each packet, corresponding to
various TCP extensions which are provided in this implementation. The
TCP_NOOPT is provided to disable
TCP option use on a per-connection basis.
- By convention, the sender-TCP will set the
“push” bit, and begin transmission immediately (if
permitted) at the end of every user call to
writev(2). When this option is set to
a non-zero value, TCP will delay sending any data at all until either the
socket is closed, or the internal send buffer is filled.
TCP_KEEPALIVE options enable to specify the
amount of time, in seconds, that the connection must be idle before
keepalive probes (if enabled) are sent. The default value is specified by
the MIB variable net.inet.tcp.keepidle.
TCP_CONNECTIONTIMEOUT option allows to specify
the timeout, in seconds, for new, non established TCP connections. This
option can be useful for both active and passive TCP connections. The
default value is specified by the MIB variable
- When keepalive probes are enabled, this option will set the amount of time
in seconds between successive keepalives sent to probe an unresponsive
- When keepalive probes are enabled, this option will set the number of
times a keepalive probe should be repeated if the peer is not responding.
After this many probes, the connection will be closed.
- When a stream of TCP data packets are received, OS X uses an algorithm to
reduce the number of acknowlegements by generating a TCP acknowlegement
for 8 data packets instead of acknowledging every other data packet. When
this socket option is enabled, the connection will always send a TCP
acknowledgement for every other data packet.
- Using Explicit Congestion Notification (ECN) on TCP allows bi-directional
end-to-end notification of congestion without dropping packets.
Conventionally TCP/IP networks signal congestion by dropping packets. When
ECN is successfully negotiated, an ECN-aware router may set a mark in the
IP header instead of dropping a packet in order to signal impending
congestion. The TCP receiver of the packet echoes congestion indication to
the TCP sender, which reduces it's transmission rate as if it detected a
dropped packet. This will avoid unnecessary retransmissions and will
improve latency by saving the time required for recovering a lost
- The send socket buffer of a TCP sender has unsent and unacknowledged data.
This option allows a TCP sender to control the amount of unsent data kept
in the send socket buffer. The value of the option should be the maximum
amount of unsent data in bytes. Kevent, poll and select will generate a
write notification when the unsent data falls below the amount given by
this option. This will allow an application to generate just-in-time fresh
updates for real-time communication.
- The TCP listener can set this option to use TCP Fast Open feature. After
setting this option, an accept(2) may
return a socket that is in SYN_RECEIVED state but is readable and
- This socket option can be used to obtain TCP connection level statistics.
The "struct tcp_connection_info" defined in
<netinet/tcp_var.h> is copied to the user buffer.
The option level for the
setsockopt(2) call is the protocol
number for TCP, available from
IPPROTO_TCP. All options are declared in
Options at the IP transport level may be used with TCP; see
ip(4). Incoming connection requests that
are source-routed are noted, and the reverse source route is used in
When a TCP socket is set non-blocking, and the connection cannot
be established immediately, connect(2)
or connectx(2) returns with the error
EINPROGRESS, and the connection is established
When the asynchronous connection completes successfully,
kqueue(2) will indicate the file
descriptor is ready for writing. If the connection encounters an error, the
file descriptor is marked ready for both reading and writing, and the
pending error can be retrieved via the socket option
Note that even if the socket is non-blocking, it is possible for
the connection to be established immediately. In that case
connectx(2) does not return with
A socket operation may fail with one of the following errors
- when trying to establish a connection on a socket which already has
- when the system runs out of memory for an internal data structure;
- when a connection was dropped due to excessive retransmissions;
- when the remote peer forces the connection to be closed;
- when the remote peer actively refuses connection establishment (usually
because no process is listening to the port);
- when an attempt is made to create a socket with a port which has already
- when an attempt is made to create a socket with a network address for
which no network interface exists;
- when an attempt is made to bind or connect a socket to a multicast
- returned by connect(2) or
connectx(2) when the socket is set
nonblocking, and the connection cannot be immediately established;
- returned by connect(2) or
connectx(2) when connection request
is already in progress for the specified socket.
- returned by recv(2) or
send(2) in case a connection is
experiencing a data-stall (probably due to a middlebox issue). It is
advised that the current connection gets closed by the application and a
new attempt is being made.
The TCP protocol appeared in 4.2BSD.
The socket option
first appeared in Mac OS X 10.6.