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 2014-02-11 "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 );
65 calls are used to receive messages from a socket.
67 to receive data on both connectionless and connection-oriented sockets.
68 This page first describes common features of all three system calls,
69 and then describes the differences between the calls.
71 All three calls return the length of the message on successful
73 If a message is too long to fit in the supplied buffer, excess
74 bytes may be discarded depending on the type of socket the message is
77 If no messages are available at the socket, the receive calls wait for a
78 message to arrive, unless the socket is nonblocking (see
80 in which case the value \-1 is returned and the external variable
83 .BR EAGAIN " or " EWOULDBLOCK .
84 The receive calls normally return any data available, up to the requested
85 amount, rather than waiting for receipt of the full amount requested.
87 An application can used
92 to determine when more data arrives on a socket.
93 .SS The flags argument
96 argument is formed by ORing one or more of the following values:
98 .BR MSG_CMSG_CLOEXEC " (" recvmsg "() only; since Linux 2.6.23)"
99 Set the close-on-exec flag for the file descriptor received
100 via a UNIX domain file descriptor using the
102 operation (described in
104 This flag is useful for the same reasons as the
109 .BR MSG_DONTWAIT " (since Linux 2.2)"
110 Enables nonblocking operation; if the operation would block,
111 the call fails with the error
112 .BR EAGAIN " or " EWOULDBLOCK
113 (this can also be enabled using the
119 .BR MSG_ERRQUEUE " (since Linux 2.2)"
121 specifies that queued errors should be received from the socket error queue.
122 The error is passed in
123 an ancillary message with a type dependent on the protocol (for IPv4
125 The user should supply a buffer of sufficient size.
130 for more information.
131 The payload of the original packet that caused the error
132 is passed as normal data via
134 The original destination address of the datagram that caused the error
138 For local errors, no address is passed (this can be checked with the
142 For error receives, the
146 After an error has been passed, the pending socket error
147 is regenerated based on the next queued error and will be passed
148 on the next socket operation.
150 The error is supplied in a
156 #define SO_EE_ORIGIN_NONE 0
157 #define SO_EE_ORIGIN_LOCAL 1
158 #define SO_EE_ORIGIN_ICMP 2
159 #define SO_EE_ORIGIN_ICMP6 3
161 struct sock_extended_err
163 uint32_t ee_errno; /* error number */
164 uint8_t ee_origin; /* where the error originated */
165 uint8_t ee_type; /* type */
166 uint8_t ee_code; /* code */
167 uint8_t ee_pad; /* padding */
168 uint32_t ee_info; /* additional information */
169 uint32_t ee_data; /* other data */
170 /* More data may follow */
173 struct sockaddr *SO_EE_OFFENDER(struct sock_extended_err *);
180 number of the queued error.
182 is the origin code of where the error originated.
183 The other fields are protocol-specific.
186 returns a pointer to the address of the network object
187 where the error originated from given a pointer to the ancillary message.
188 If this address is not known, the
194 and the other fields of the
197 The payload of the packet that caused the error is passed as normal data.
199 For local errors, no address is passed (this
200 can be checked with the
209 After an error has been passed, the pending socket error
210 is regenerated based on the next queued error and will be passed
211 on the next socket operation.
214 This flag requests receipt of out-of-band data that would not be received
215 in the normal data stream.
216 Some protocols place expedited data
217 at the head of the normal data queue, and thus this flag cannot
218 be used with such protocols.
221 This flag causes the receive operation to
222 return data from the beginning of the
223 receive queue without removing that data from the queue.
225 subsequent receive call will return the same data.
227 .BR MSG_TRUNC " (since Linux 2.2)"
230 Internet datagram (since Linux 2.4.27/2.6.8),
231 netlink (since Linux 2.6.22), and UNIX datagram (since Linux 3.4) sockets:
232 return the real length of the packet or datagram,
233 even when it was longer than the passed buffer.
235 For use with Internet stream sockets, see
238 .BR MSG_WAITALL " (since Linux 2.2)"
239 This flag requests that the operation block until the full request is
241 However, the call may still return less data than requested if
242 a signal is caught, an error or disconnect occurs, or the next data to be
243 received is of a different type than that returned.
247 places the received message into the buffer
249 The caller must specify the size of the buffer in
255 and the underlying protocol provides the source address of the message,
256 that source address is placed in the buffer pointed to by
258 .\" (Note: for datagram sockets in both the UNIX and Internet domains,
262 .\" is also filled in for stream sockets in the UNIX domain, but is not
263 .\" filled in for stream sockets in the Internet domain.)
264 .\" [The above notes on AF_UNIX and AF_INET sockets apply as at
265 .\" Kernel 2.4.18. (MTK, 22 Jul 02)]
268 is a value-result argument.
270 it should be initialized to the size of the buffer associated with
274 is updated to contain the actual size of the source address.
275 The returned address is truncated if the buffer provided is too small;
278 will return a value greater than was supplied to the call.
280 If the caller is not interested in the source address,
282 should be specified as NULL and
284 should be specified as 0.
289 call is normally used only on a
293 It is equivalent to the call:
295 recvfrom(fd, buf, len, flags, NULL, 0));
302 structure to minimize the number of directly supplied arguments.
303 This structure is defined as follows in
308 struct iovec { /* Scatter/gather array items */
309 void *iov_base; /* Starting address */
310 size_t iov_len; /* Number of bytes to transfer */
314 void *msg_name; /* optional address */
315 socklen_t msg_namelen; /* size of address */
316 struct iovec *msg_iov; /* scatter/gather array */
317 size_t msg_iovlen; /* # elements in msg_iov */
318 void *msg_control; /* ancillary data, see below */
319 size_t msg_controllen; /* ancillary data buffer len */
320 int msg_flags; /* flags on received message */
329 specify the source address if the socket is unconnected;
331 may be given as a null pointer if no names are desired or required.
336 describe scatter-gather locations, as discussed in
342 points to a buffer for other protocol control-related messages or
343 miscellaneous ancillary data.
348 should contain the length of the available buffer in
350 upon return from a successful call it will contain the length
351 of the control message sequence.
353 The messages are of the form:
358 socklen_t cmsg_len; /* data byte count, including hdr */
359 int cmsg_level; /* originating protocol */
360 int cmsg_type; /* protocol-specific type */
362 unsigned char cmsg_data[]; */
367 Ancillary data should be accessed only by the macros defined in
370 As an example, Linux uses this ancillary data mechanism to pass extended
371 errors, IP options, or file descriptors over UNIX domain sockets.
379 It can contain several flags:
382 indicates end-of-record; the data returned completed a record (generally
383 used with sockets of type
384 .BR SOCK_SEQPACKET ).
387 indicates that the trailing portion of a datagram was discarded because the
388 datagram was larger than the buffer supplied.
391 indicates that some control data were discarded due to lack of space in the
392 buffer for ancillary data.
395 is returned to indicate that expedited or out-of-band data were received.
398 indicates that no data was received but an extended error from the socket
401 These calls return the number of bytes received, or \-1
402 if an error occurred.
403 In the event of an error,
405 is set to indicate the error.
407 When a stream socket peer has performed an orderly shutdown,
408 the return value will be 0 (the traditional "end-of-file" return).
410 Datagram sockets in various domains (e.g., the UNIX and Internet domains)
411 permit zero-length datagrams.
412 When such a datagram is received, the return value is 0.
414 The value 0 may also be returned if the requested number of bytes
415 to receive from a stream socket was 0.
417 These are some standard errors generated by the socket layer.
419 may be generated and returned from the underlying protocol modules;
420 see their manual pages.
422 .BR EAGAIN " or " EWOULDBLOCK
423 .\" Actually EAGAIN on Linux
424 The socket is marked nonblocking and the receive operation
425 would block, or a receive timeout had been set and the timeout expired
426 before data was received.
427 POSIX.1-2001 allows either error to be returned for this case,
428 and does not require these constants to have the same value,
429 so a portable application should check for both possibilities.
434 is an invalid descriptor.
437 A remote host refused to allow the network connection (typically
438 because it is not running the requested service).
441 The receive buffer pointer(s) point outside the process's
445 The receive was interrupted by delivery of a signal before
446 any data were available; see
450 Invalid argument passed.
451 .\" e.g., msg_namelen < 0 for recvmsg() or addrlen < 0 for recvfrom()
454 Could not allocate memory for
458 The socket is associated with a connection-oriented protocol
459 and has not been connected (see
467 does not refer to a socket.
469 4.4BSD (these function calls first appeared in 4.2BSD),
472 POSIX.1-2001 describes only the
479 The prototypes given above follow glibc2.
480 The Single UNIX Specification agrees, except that it has return values
481 of type \fIssize_t\fP (while 4.x BSD and libc4 and libc5 all have \fIint\fP).
484 argument is \fIint\fP in 4.x BSD, but \fIunsigned int\fP in libc4 and libc5.
487 argument is \fIint\fP in 4.x BSD, but \fIsize_t\fP in libc4 and libc5.
490 argument is \fIint\ *\fP in 4.x BSD, libc4 and libc5.
491 The present \fIsocklen_t\ *\fP was invented by POSIX.
495 According to POSIX.1-2001, the
499 structure should be typed as
501 but glibc currently types it as
503 .\" glibc bug raised 12 Mar 2006
504 .\" http://sourceware.org/bugzilla/show_bug.cgi?id=2448
505 .\" The problem is an underlying kernel issue: the size of the
506 .\" __kernel_size_t type used to type this field varies
507 .\" across architectures, but socklen_t is always 32 bits.
511 for information about a Linux-specific system call
512 that can be used to receive multiple datagrams in a single call.
514 An example of the use of
530 This page is part of release 3.64 of the Linux
533 A description of the project,
534 and information about reporting bugs,
536 \%http://www.kernel.org/doc/man\-pages/.