1 .\" Copyright (c) 1983, 1990, 1991 The Regents of the University of California.
2 .\" All rights reserved.
4 .\" %%%LICENSE_START(BSD_4_CLAUSE_UCB)
5 .\" Redistribution and use in source and binary forms, with or without
6 .\" modification, are permitted provided that the following conditions
8 .\" 1. Redistributions of source code must retain the above copyright
9 .\" notice, this list of conditions and the following disclaimer.
10 .\" 2. Redistributions in binary form must reproduce the above copyright
11 .\" notice, this list of conditions and the following disclaimer in the
12 .\" documentation and/or other materials provided with the distribution.
13 .\" 3. All advertising materials mentioning features or use of this software
14 .\" must display the following acknowledgement:
15 .\" This product includes software developed by the University of
16 .\" California, Berkeley and its contributors.
17 .\" 4. Neither the name of the University nor the names of its contributors
18 .\" may be used to endorse or promote products derived from this software
19 .\" without specific prior written permission.
21 .\" THIS SOFTWARE IS PROVIDED BY THE REGENTS AND CONTRIBUTORS ``AS IS'' AND
22 .\" ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE
23 .\" IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE
24 .\" ARE DISCLAIMED. IN NO EVENT SHALL THE REGENTS OR CONTRIBUTORS BE LIABLE
25 .\" FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL
26 .\" DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS
27 .\" OR SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION)
28 .\" HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT
29 .\" LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY
30 .\" OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF
34 .\" $Id: recv.2,v 1.3 1999/05/13 11:33:38 freitag Exp $
36 .\" Modified Sat Jul 24 00:22:20 1993 by Rik Faith <faith@cs.unc.edu>
37 .\" Modified Tue Oct 22 17:45:19 1996 by Eric S. Raymond <esr@thyrsus.com>
38 .\" Modified 1998,1999 by Andi Kleen
39 .\" 2001-06-19 corrected SO_EE_OFFENDER, bug report by James Hawtin
41 .TH RECV 2 2013-04-19 "Linux" "Linux Programmer's Manual"
43 recv, recvfrom, recvmsg \- receive a message from a socket
45 .\" .B #include <sys/uio.h>
48 .B #include <sys/types.h>
50 .B #include <sys/socket.h>
52 .BI "ssize_t recv(int " sockfd ", void *" buf ", size_t " len ", int " flags );
54 .BI "ssize_t recvfrom(int " sockfd ", void *" buf ", size_t " len ", int " flags ,
55 .BI " struct sockaddr *" src_addr ", socklen_t *" addrlen );
57 .BI "ssize_t recvmsg(int " sockfd ", struct msghdr *" msg ", int " flags );
64 calls are used to receive messages from a socket, and may be used
65 to receive data on a socket whether or not it is connection-oriented.
69 is not NULL, and the underlying protocol provides the source address,
70 this source address is filled in.
73 is NULL, nothing is filled in; in this case,
75 is not used, and should also be NULL.
76 .\" (Note: for datagram sockets in both the UNIX and Internet domains,
80 .\" is also filled in for stream sockets in the UNIX domain, but is not
81 .\" filled in for stream sockets in the Internet domain.)
82 .\" [The above notes on AF_UNIX and AF_INET sockets apply as at
83 .\" Kernel 2.4.18. (MTK, 22 Jul 02)]
86 is a value-result argument,
87 which the caller should initialize before the call to
88 the size of the buffer associated with
90 and modified on return to indicate the actual size of the source address.
91 The returned address is truncated if the buffer provided is too small;
94 will return a value greater than was supplied to the call.
98 call is normally used only on a
108 All three routines return the length of the message on successful
110 If a message is too long to fit in the supplied buffer, excess
111 bytes may be discarded depending on the type of socket the message is
114 If no messages are available at the socket, the receive calls wait for a
115 message to arrive, unless the socket is nonblocking (see
117 in which case the value \-1 is returned and the external variable
120 .BR EAGAIN " or " EWOULDBLOCK .
121 The receive calls normally return any data available, up to the requested
122 amount, rather than waiting for receipt of the full amount requested.
128 call may be used to determine when more data arrives.
134 call is formed by ORing one or more of the following values:
136 .BR MSG_CMSG_CLOEXEC " (" recvmsg "() only; since Linux 2.6.23)"
137 Set the close-on-exec flag for the file descriptor received
138 via a UNIX domain file descriptor using the
140 operation (described in
142 This flag is useful for the same reasons as the
147 .BR MSG_DONTWAIT " (since Linux 2.2)"
148 Enables nonblocking operation; if the operation would block,
149 the call fails with the error
150 .BR EAGAIN " or " EWOULDBLOCK
151 (this can also be enabled using the
157 .BR MSG_ERRQUEUE " (since Linux 2.2)"
159 specifies that queued errors should be received from the socket error queue.
160 The error is passed in
161 an ancillary message with a type dependent on the protocol (for IPv4
163 The user should supply a buffer of sufficient size.
168 for more information.
169 The payload of the original packet that caused the error
170 is passed as normal data via
172 The original destination address of the datagram that caused the error
176 For local errors, no address is passed (this can be checked with the
180 For error receives, the
184 After an error has been passed, the pending socket error
185 is regenerated based on the next queued error and will be passed
186 on the next socket operation.
188 The error is supplied in a
194 #define SO_EE_ORIGIN_NONE 0
195 #define SO_EE_ORIGIN_LOCAL 1
196 #define SO_EE_ORIGIN_ICMP 2
197 #define SO_EE_ORIGIN_ICMP6 3
199 struct sock_extended_err
201 uint32_t ee_errno; /* error number */
202 uint8_t ee_origin; /* where the error originated */
203 uint8_t ee_type; /* type */
204 uint8_t ee_code; /* code */
205 uint8_t ee_pad; /* padding */
206 uint32_t ee_info; /* additional information */
207 uint32_t ee_data; /* other data */
208 /* More data may follow */
211 struct sockaddr *SO_EE_OFFENDER(struct sock_extended_err *);
218 number of the queued error.
220 is the origin code of where the error originated.
221 The other fields are protocol-specific.
224 returns a pointer to the address of the network object
225 where the error originated from given a pointer to the ancillary message.
226 If this address is not known, the
232 and the other fields of the
235 The payload of the packet that caused the error is passed as normal data.
237 For local errors, no address is passed (this
238 can be checked with the
247 After an error has been passed, the pending socket error
248 is regenerated based on the next queued error and will be passed
249 on the next socket operation.
252 This flag requests receipt of out-of-band data that would not be received
253 in the normal data stream.
254 Some protocols place expedited data
255 at the head of the normal data queue, and thus this flag cannot
256 be used with such protocols.
259 This flag causes the receive operation to
260 return data from the beginning of the
261 receive queue without removing that data from the queue.
263 subsequent receive call will return the same data.
265 .BR MSG_TRUNC " (since Linux 2.2)"
268 Internet datagram (since Linux 2.4.27/2.6.8),
269 netlink (since Linux 2.6.22) and UNIX datagram (since Linux 3.4) sockets:
270 return the real length of the packet or datagram,
271 even when it was longer than the passed buffer.
272 Not implemented for UNIX domain
276 For use with Internet stream sockets, see
279 .BR MSG_WAITALL " (since Linux 2.2)"
280 This flag requests that the operation block until the full request is
282 However, the call may still return less data than requested if
283 a signal is caught, an error or disconnect occurs, or the next data to be
284 received is of a different type than that returned.
290 structure to minimize the number of directly supplied arguments.
291 This structure is defined as follows in
296 struct iovec { /* Scatter/gather array items */
297 void *iov_base; /* Starting address */
298 size_t iov_len; /* Number of bytes to transfer */
302 void *msg_name; /* optional address */
303 socklen_t msg_namelen; /* size of address */
304 struct iovec *msg_iov; /* scatter/gather array */
305 size_t msg_iovlen; /* # elements in msg_iov */
306 void *msg_control; /* ancillary data, see below */
307 size_t msg_controllen; /* ancillary data buffer len */
308 int msg_flags; /* flags on received message */
317 specify the source address if the socket is unconnected;
319 may be given as a NULL pointer if no names are desired or required.
324 describe scatter-gather locations, as discussed in
330 points to a buffer for other protocol control-related messages or
331 miscellaneous ancillary data.
336 should contain the length of the available buffer in
338 upon return from a successful call it will contain the length
339 of the control message sequence.
341 The messages are of the form:
346 socklen_t cmsg_len; /* data byte count, including hdr */
347 int cmsg_level; /* originating protocol */
348 int cmsg_type; /* protocol-specific type */
350 unsigned char cmsg_data[]; */
355 Ancillary data should be accessed only by the macros defined in
358 As an example, Linux uses this ancillary data mechanism to pass extended
359 errors, IP options, or file descriptors over UNIX domain sockets.
367 It can contain several flags:
370 indicates end-of-record; the data returned completed a record (generally
371 used with sockets of type
372 .BR SOCK_SEQPACKET ).
375 indicates that the trailing portion of a datagram was discarded because the
376 datagram was larger than the buffer supplied.
379 indicates that some control data were discarded due to lack of space in the
380 buffer for ancillary data.
383 is returned to indicate that expedited or out-of-band data were received.
386 indicates that no data was received but an extended error from the socket
389 These calls return the number of bytes received, or \-1
390 if an error occurred.
391 In the event of an error,
393 is set to indicate the error.
394 The return value will be 0 when the
395 peer has performed an orderly shutdown.
397 These are some standard errors generated by the socket layer.
399 may be generated and returned from the underlying protocol modules;
400 see their manual pages.
402 .BR EAGAIN " or " EWOULDBLOCK
403 .\" Actually EAGAIN on Linux
404 The socket is marked nonblocking and the receive operation
405 would block, or a receive timeout had been set and the timeout expired
406 before data was received.
407 POSIX.1-2001 allows either error to be returned for this case,
408 and does not require these constants to have the same value,
409 so a portable application should check for both possibilities.
414 is an invalid descriptor.
417 A remote host refused to allow the network connection (typically
418 because it is not running the requested service).
421 The receive buffer pointer(s) point outside the process's
425 The receive was interrupted by delivery of a signal before
426 any data were available; see
430 Invalid argument passed.
431 .\" e.g., msg_namelen < 0 for recvmsg() or addrlen < 0 for recvfrom()
434 Could not allocate memory for
438 The socket is associated with a connection-oriented protocol
439 and has not been connected (see
447 does not refer to a socket.
449 4.4BSD (these function calls first appeared in 4.2BSD),
452 POSIX.1-2001 describes only the
459 The prototypes given above follow glibc2.
460 The Single UNIX Specification agrees, except that it has return values
461 of type \fIssize_t\fP (while 4.x BSD and libc4 and libc5 all have \fIint\fP).
464 argument is \fIint\fP in 4.x BSD, but \fIunsigned int\fP in libc4 and libc5.
467 argument is \fIint\fP in 4.x BSD, but \fIsize_t\fP in libc4 and libc5.
470 argument is \fIint\ *\fP in 4.x BSD, libc4 and libc5.
471 The present \fIsocklen_t\ *\fP was invented by POSIX.
475 According to POSIX.1-2001, the
479 structure should be typed as
481 but glibc currently types it as
483 .\" glibc bug raised 12 Mar 2006
484 .\" http://sourceware.org/bugzilla/show_bug.cgi?id=2448
485 .\" The problem is an underlying kernel issue: the size of the
486 .\" __kernel_size_t type used to type this field varies
487 .\" across architectures, but socklen_t is always 32 bits.
491 for information about a Linux-specific system call
492 that can be used to receive multiple datagrams in a single call.
494 An example of the use of