man7/socket.7

   1 '\" t
   2 .\" This man page is Copyright (C) 1999 Andi Kleen <ak@muc.de>.
   3 .\" and copyright (c) 1999 Matthew Wilcox.
   4 .\"
   5 .\" %%%LICENSE_START(VERBATIM_ONE_PARA)
   6 .\" Permission is granted to distribute possibly modified copies
   7 .\" of this page provided the header is included verbatim,
   8 .\" and in case of nontrivial modification author and date
   9 .\" of the modification is added to the header.
  10 .\" %%%LICENSE_END
  11 .\"
  12 .\" 2002-10-30, Michael Kerrisk, <mtk.manpages@gmail.com>
  13 .\"     Added description of SO_ACCEPTCONN
  14 .\" 2004-05-20, aeb, added SO_RCVTIMEO/SO_SNDTIMEO text.
  15 .\" Modified, 27 May 2004, Michael Kerrisk <mtk.manpages@gmail.com>
  16 .\"     Added notes on capability requirements
  17 .\"     A few small grammar fixes
  18 .\" 2010-06-13 Jan Engelhardt <jengelh@medozas.de>
  19 .\"     Documented SO_DOMAIN and SO_PROTOCOL.
  20 .\" FIXME
  21 .\" The following are not yet documented:
  22 .\"     SO_PEERNAME (2.4?)
  23 .\"             get only
  24 .\"             Seems to do something similar to getpeernam(), but then
  25 .\"             why is it necessary / how does it differ?
  26 .\"     SO_TIMESTAMPNS (2.6.22)
  27 .\"             Documentation/networking/timestamping.txt
  28 .\"             commit 92f37fd2ee805aa77925c1e64fd56088b46094fc
  29 .\"             Author: Eric Dumazet <dada1@cosmosbay.com>
  30 .\"     SO_TIMESTAMPING (2.6.30)
  31 .\"             Documentation/networking/timestamping.txt
  32 .\"             commit cb9eff097831007afb30d64373f29d99825d0068
  33 .\"             Author: Patrick Ohly <patrick.ohly@intel.com>
  34 .\"     SO_RXQ_OVFL (2.6.33)
  35 .\"             commit 3b885787ea4112eaa80945999ea0901bf742707f
  36 .\"             Author: Neil Horman <nhorman@tuxdriver.com>
  37 .\"     SO_WIFI_STATUS (3.3)
  38 .\"             commit 6e3e939f3b1bf8534b32ad09ff199d88800835a0
  39 .\"             Author: Johannes Berg <johannes.berg@intel.com>
  40 .\"             Also: SCM_WIFI_STATUS
  41 .\"     SO_NOFCS (3.4)
  42 .\"             commit 3bdc0eba0b8b47797f4a76e377dd8360f317450f
  43 .\"             Author: Ben Greear <greearb@candelatech.com>
  44 .\"
  45 .TH SOCKET 7 2013-04-17 Linux "Linux Programmer's Manual"
  46 .SH NAME
  47 socket \- Linux socket interface
  48 .SH SYNOPSIS
  49 .B #include <sys/socket.h>
  50 .sp
  51 .IB sockfd " = socket(int " socket_family ", int " socket_type ", int " protocol );
  52 .SH DESCRIPTION
  53 This manual page describes the Linux networking socket layer user
  54 interface.
  55 The BSD compatible sockets
  56 are the uniform interface
  57 between the user process and the network protocol stacks in the kernel.
  58 The protocol modules are grouped into
  59 .I protocol families
  60 like
  61 .BR AF_INET ", " AF_IPX ", " AF_PACKET
  62 and
  63 .I socket types
  64 like
  65 .B SOCK_STREAM
  66 or
  67 .BR SOCK_DGRAM .
  68 See
  69 .BR socket (2)
  70 for more information on families and types.
  71 .SS Socket-layer functions
  72 These functions are used by the user process to send or receive packets
  73 and to do other socket operations.
  74 For more information see their respective manual pages.
  75
  76 .BR socket (2)
  77 creates a socket,
  78 .BR connect (2)
  79 connects a socket to a remote socket address,
  80 the
  81 .BR bind (2)
  82 function binds a socket to a local socket address,
  83 .BR listen (2)
  84 tells the socket that new connections shall be accepted, and
  85 .BR accept (2)
  86 is used to get a new socket with a new incoming connection.
  87 .BR socketpair (2)
  88 returns two connected anonymous sockets (implemented only for a few
  89 local families like
  90 .BR AF_UNIX )
  91 .PP
  92 .BR send (2),
  93 .BR sendto (2),
  94 and
  95 .BR sendmsg (2)
  96 send data over a socket, and
  97 .BR recv (2),
  98 .BR recvfrom (2),
  99 .BR recvmsg (2)
 100 receive data from a socket.
 101 .BR poll (2)
 102 and
 103 .BR select (2)
 104 wait for arriving data or a readiness to send data.
 105 In addition, the standard I/O operations like
 106 .BR write (2),
 107 .BR writev (2),
 108 .BR sendfile (2),
 109 .BR read (2),
 110 and
 111 .BR readv (2)
 112 can be used to read and write data.
 113 .PP
 114 .BR getsockname (2)
 115 returns the local socket address and
 116 .BR getpeername (2)
 117 returns the remote socket address.
 118 .BR getsockopt (2)
 119 and
 120 .BR setsockopt (2)
 121 are used to set or get socket layer or protocol options.
 122 .BR ioctl (2)
 123 can be used to set or read some other options.
 124 .PP
 125 .BR close (2)
 126 is used to close a socket.
 127 .BR shutdown (2)
 128 closes parts of a full-duplex socket connection.
 129 .PP
 130 Seeking, or calling
 131 .BR pread (2)
 132 or
 133 .BR pwrite (2)
 134 with a nonzero position is not supported on sockets.
 135 .PP
 136 It is possible to do nonblocking I/O on sockets by setting the
 137 .B O_NONBLOCK
 138 flag on a socket file descriptor using
 139 .BR fcntl (2).
 140 Then all operations that would block will (usually)
 141 return with
 142 .B EAGAIN
 143 (operation should be retried later);
 144 .BR connect (2)
 145 will return
 146 .B EINPROGRESS
 147 error.
 148 The user can then wait for various events via
 149 .BR poll (2)
 150 or
 151 .BR select (2).
 152 .TS
 153 tab(:) allbox;
 154 c s s
 155 l l l.
 156 I/O events
 157 Event:Poll flag:Occurrence
 158 Read:POLLIN:T{
 159 New data arrived.
 160 T}
 161 Read:POLLIN:T{
 162 A connection setup has been completed
 163 (for connection-oriented sockets)
 164 T}
 165 Read:POLLHUP:T{
 166 A disconnection request has been initiated by the other end.
 167 T}
 168 Read:POLLHUP:T{
 169 A connection is broken (only for connection-oriented protocols).
 170 When the socket is written
 171 .B SIGPIPE
 172 is also sent.
 173 T}
 174 Write:POLLOUT:T{
 175 Socket has enough send buffer space for writing new data.
 176 T}
 177 Read/Write:T{
 178 POLLIN|
 179 .br
 180 POLLOUT
 181 T}:T{
 182 An outgoing
 183 .BR connect (2)
 184 finished.
 185 T}
 186 Read/Write:POLLERR:An asynchronous error occurred.
 187 Read/Write:POLLHUP:The other end has shut down one direction.
 188 Exception:POLLPRI:T{
 189 Urgent data arrived.
 190 .B SIGURG
 191 is sent then.
 192 T}
 193 .\" FIXME . The following is not true currently:
 194 .\" It is no I/O event when the connection
 195 .\" is broken from the local end using
 196 .\" .BR shutdown (2)
 197 .\" or
 198 .\" .BR close (2).
 199 .TE
 200 .PP
 201 An alternative to
 202 .BR poll (2)
 203 and
 204 .BR select (2)
 205 is to let the kernel inform the application about events
 206 via a
 207 .B SIGIO
 208 signal.
 209 For that the
 210 .B O_ASYNC
 211 flag must be set on a socket file descriptor via
 212 .BR fcntl (2)
 213 and a valid signal handler for
 214 .B SIGIO
 215 must be installed via
 216 .BR sigaction (2).
 217 See the
 218 .I Signals
 219 discussion below.
 220 .SS Socket address structures
 221 Each socket domain has its own format for socket addresses,
 222 with a domain-specific address structure.
 223 Each of these structures begins with an
 224 integer "family" field (typed as
 225 .IR sa_family_t )
 226 that indicates the type of the address structure.
 227 This allows
 228 the various system calls (e.g.,
 229 .BR connect (2),
 230 .BR bind (2),
 231 .BR accept (2),
 232 .BR getsockname (2),
 233 .BR getpeername (2)),
 234 which are generic to all socket domains,
 235 to determine the domain of a particular socket address.
 236
 237 To allow any type of socket address to be passed to
 238 interfaces in the sockets API,
 239 the type
 240 .IR "struct sockaddr"
 241 is defined.
 242 The purpose of this type is purely to allow casting of
 243 domain-specific socket address types to a "generic" type,
 244 so as to avoid compiler warnings about type mismatches in
 245 calls to the sockets API.
 246
 247 In addition, the sockets API provides the data type
 248 .IR "struct sockaddr_storage".
 249 This type
 250 is suitable to accommodate all supported domain-specific socket
 251 address structures; it is large enough and is aligned properly.
 252 (In particular, it is large enough to hold
 253 IPv6 socket addresses.)
 254 The structure includes the following field, which can be used to identify
 255 the type of socket address actually stored in the structure:
 256
 257 .in +4n
 258 .nf
 259     sa_family_t ss_family;
 260 .fi
 261 .in
 262
 263 The
 264 .I sockaddr_storage
 265 structure is useful in programs that must handle socket addresses
 266 in a generic way
 267 (e.g., programs that must deal with both IPv4 and IPv6 socket addresses).
 268 .SS Socket options
 269 These socket options can be set by using
 270 .BR setsockopt (2)
 271 and read with
 272 .BR getsockopt (2)
 273 with the socket level set to
 274 .B SOL_SOCKET
 275 for all sockets:
 276 .\" FIXME
 277 .\" In the list below, the text used to describe argument types
 278 .\" for each socket option should be more consistent
 279 .\"
 280 .\" SO_ACCEPTCONN is in POSIX.1-2001, and its origin is explained in
 281 .\" W R Stevens, UNPv1
 282 .TP
 283 .B SO_ACCEPTCONN
 284 Returns a value indicating whether or not this socket has been marked
 285 to accept connections with
 286 .BR listen (2).
 287 The value 0 indicates that this is not a listening socket,
 288 the value 1 indicates that this is a listening socket.
 289 This socket option is read-only.
 290 .TP
 291 .B SO_BINDTODEVICE
 292 Bind this socket to a particular device like \(lqeth0\(rq,
 293 as specified in the passed interface name.
 294 If the
 295 name is an empty string or the option length is zero, the socket device
 296 binding is removed.
 297 The passed option is a variable-length null-terminated
 298 interface name string with the maximum size of
 299 .BR IFNAMSIZ .
 300 If a socket is bound to an interface,
 301 only packets received from that particular interface are processed by the
 302 socket.
 303 Note that this works only for some socket types, particularly
 304 .B AF_INET
 305 sockets.
 306 It is not supported for packet sockets (use normal
 307 .BR bind (2)
 308 there).
 309
 310 Before Linux 3.8,
 311 this socket option could be set, but could not retrieved with
 312 .BR getsockopt (2).
 313 Since Linux 3.8, it is readable.
 314 The
 315 .I optlen
 316 argument should contain the buffer size available
 317 to receive the device name and is recommended to be
 318 .BR IFNAMSZ
 319 bytes.
 320 The real device name length is reported back in the
 321 .I optlen
 322 argument.
 323 .TP
 324 .B SO_BROADCAST
 325 Set or get the broadcast flag.
 326 When enabled, datagram sockets are allowed to send
 327 packets to a broadcast address.
 328 This option has no effect on stream-oriented sockets.
 329 .TP
 330 .B SO_BSDCOMPAT
 331 Enable BSD bug-to-bug compatibility.
 332 This is used by the UDP protocol module in Linux 2.0 and 2.2.
 333 If enabled ICMP errors received for a UDP socket will not be passed
 334 to the user program.
 335 In later kernel versions, support for this option has been phased out:
 336 Linux 2.4 silently ignores it, and Linux 2.6 generates a kernel warning
 337 (printk()) if a program uses this option.
 338 Linux 2.0 also enabled BSD bug-to-bug compatibility
 339 options (random header changing, skipping of the broadcast flag) for raw
 340 sockets with this option, but that was removed in Linux 2.2.
 341 .TP
 342 .B SO_DEBUG
 343 Enable socket debugging.
 344 Only allowed for processes with the
 345 .B CAP_NET_ADMIN
 346 capability or an effective user ID of 0.
 347 .TP
 348 .BR SO_DOMAIN " (since Linux 2.6.32)"
 349 Retrieves the socket domain as an integer, returning a value such as
 350 .BR AF_INET6 .
 351 See
 352 .BR socket (2)
 353 for details.
 354 This socket option is read-only.
 355 .TP
 356 .B SO_ERROR
 357 Get and clear the pending socket error.
 358 This socket option is read-only.
 359 Expects an integer.
 360 .TP
 361 .B SO_DONTROUTE
 362 Don't send via a gateway, send only to directly connected hosts.
 363 The same effect can be achieved by setting the
 364 .B MSG_DONTROUTE
 365 flag on a socket
 366 .BR send (2)
 367 operation.
 368 Expects an integer boolean flag.
 369 .TP
 370 .B SO_KEEPALIVE
 371 Enable sending of keep-alive messages on connection-oriented sockets.
 372 Expects an integer boolean flag.
 373 .TP
 374 .B SO_LINGER
 375 Sets or gets the
 376 .B SO_LINGER
 377 option.
 378 The argument is a
 379 .I linger
 380 structure.
 381 .sp
 382 .in +4n
 383 .nf
 384 struct linger {
 385     int l_onoff;    /* linger active */
 386     int l_linger;   /* how many seconds to linger for */
 387 };
 388 .fi
 389 .in
 390 .IP
 391 When enabled, a
 392 .BR close (2)
 393 or
 394 .BR shutdown (2)
 395 will not return until all queued messages for the socket have been
 396 successfully sent or the linger timeout has been reached.
 397 Otherwise,
 398 the call returns immediately and the closing is done in the background.
 399 When the socket is closed as part of
 400 .BR exit (2),
 401 it always lingers in the background.
 402 .TP
 403 .BR SO_MARK " (since Linux 2.6.25)"
 404 .\" commit 4a19ec5800fc3bb64e2d87c4d9fdd9e636086fe0
 405 .\" and    914a9ab386a288d0f22252fc268ecbc048cdcbd5
 406 Set the mark for each packet sent through this socket
 407 (similar to the netfilter MARK target but socket-based).
 408 Changing the mark can be used for mark-based
 409 routing without netfilter or for packet filtering.
 410 Setting this option requires the
 411 .B CAP_NET_ADMIN
 412 capability.
 413 .TP
 414 .B SO_OOBINLINE
 415 If this option is enabled,
 416 out-of-band data is directly placed into the receive data stream.
 417 Otherwise out-of-band data is passed only when the
 418 .B MSG_OOB
 419 flag is set during receiving.
 420 .\" don't document it because it can do too much harm.
 421 .\".B SO_NO_CHECK
 422 .TP
 423 .B SO_PASSCRED
 424 Enable or disable the receiving of the
 425 .B SCM_CREDENTIALS
 426 control message.
 427 For more information see
 428 .BR unix (7).
 429 .\" FIXME Document SO_PASSSEC, added in 2.6.18; there is some info
 430 .\" in the 2.6.18 ChangeLog
 431 .TP
 432 .BR SO_PEEK_OFF " (since Linux 3.4)"
 433 .\" commit ef64a54f6e558155b4f149bb10666b9e914b6c54
 434 This option, which is currently supported only for
 435 .BR unix (7)
 436 sockets, sets the value of the "peek offset" for the
 437 .BR recv(2)
 438 system call when used with
 439 .BR MSG_PEEK
 440 flag.
 441
 442 When this option is set to a negative value
 443 (it is set to \-1 for all new sockets),
 444 traditional behavior is provided:
 445 .BR recv(2)
 446 with the
 447 .BR MSG_PEEK
 448 flag will peek data from the front of the queue.
 449
 450 When the option is set to a value greater than or equal to zero,
 451 then the next peek at data queued in the socket will occur at
 452 the byte offset specified by the option value.
 453 At the same time, the "peek offset" will be
 454 incremented by the number of bytes that were peeked from the queue,
 455 so that a subsequent peek will return the next data in the queue.i
 456
 457 If data is removed from the front of the queue via a call to
 458 .BR recv (2)
 459 (or similar) without the
 460 .BR MSG_PEEK
 461 flag, the "peek offset" will be decreased by the number of bytes removed.
 462 In other words, receiving data without the
 463 .B MSG_PEEK
 464 flag will cause the "peek offset" to be adjusted to maintain
 465 the correct relative position in the queued data,
 466 so that a subsequent peek will retrieve the data that would have been
 467 retrieved had the data not been removed.
 468
 469 For datagram sockets, if the "peek offset" points to the middle of a packet,
 470 the data returned will be marked with the
 471 .BR MSG_TRUNC
 472 flag.
 473
 474 The following example serves to illustrate the use of
 475 .BR SO_PEEK_OFF .
 476 Suppose a stream socket has the following queued input data:
 477
 478     aabbccddeeff
 479
 480 .IP
 481 The following sequence of
 482 .BR recv (2)
 483 calls would have the effect noted in the comments:
 484
 485 .in +4n
 486 .nf
 487 int ov = 4;                  // Set peek offset to 4
 488 setsockopt(fd, SOL_SOCKET, SO_PEEK_OFF, &ov, sizeof(ov));
 489
 490 recv(fd, buf, 2, MSG_PEEK);  // Peeks "cc"; offset set to 6
 491 recv(fd, buf, 2, MSG_PEEK);  // Peeks "dd"; offset set to 8
 492 recv(fd, buf, 2, 0);         // Reads "aa"; offset set to 6
 493 recv(fd, buf, 2, MSG_PEEK);  // Peeks "ee"; offset set to 8
 494 .fi
 495 .in
 496 .TP
 497 .B SO_PEERCRED
 498 Return the credentials of the foreign process connected to this socket.
 499 This is possible only for connected
 500 .B AF_UNIX
 501 stream sockets and
 502 .B AF_UNIX
 503 stream and datagram socket pairs created using
 504 .BR socketpair (2);
 505 see
 506 .BR unix (7).
 507 The returned credentials are those that were in effect at the time
 508 of the call to
 509 .BR connect (2)
 510 or
 511 .BR socketpair (2).
 512 The argument is a
 513 .I ucred
 514 structure; define the
 515 .B GNU_SOURCE
 516 feature test macro to obtain the definition of that structure from
 517 .IR <sys/socket.h> .
 518 This socket option is read-only.
 519 .TP
 520 .B SO_PRIORITY
 521 Set the protocol-defined priority for all packets to be sent on
 522 this socket.
 523 Linux uses this value to order the networking queues:
 524 packets with a higher priority may be processed first depending
 525 on the selected device queueing discipline.
 526 For
 527 .BR ip (7),
 528 this also sets the IP type-of-service (TOS) field for outgoing packets.
 529 Setting a priority outside the range 0 to 6 requires the
 530 .B CAP_NET_ADMIN
 531 capability.
 532 .TP
 533 .BR SO_PROTOCOL " (since Linux 2.6.32)"
 534 Retrieves the socket protocol as an integer, returning a value such as
 535 .BR IPPROTO_SCTP .
 536 See
 537 .BR socket (2)
 538 for details.
 539 This socket option is read-only.
 540 .TP
 541 .B SO_RCVBUF
 542 Sets or gets the maximum socket receive buffer in bytes.
 543 The kernel doubles this value (to allow space for bookkeeping overhead)
 544 when it is set using
 545 .\" Most (all?) other implementations do not do this -- MTK, Dec 05
 546 .BR setsockopt (2),
 547 and this doubled value is returned by
 548 .BR getsockopt (2).
 549 .\" The following thread on LMKL is quite informative:
 550 .\" getsockopt/setsockopt with SO_RCVBUF and SO_SNDBUF "non-standard" behaviour
 551 .\" 17 July 2012
 552 .\" http://thread.gmane.org/gmane.linux.kernel/1328935
 553 The default value is set by the
 554 .I /proc/sys/net/core/rmem_default
 555 file, and the maximum allowed value is set by the
 556 .I /proc/sys/net/core/rmem_max
 557 file.
 558 The minimum (doubled) value for this option is 256.
 559 .TP
 560 .BR SO_RCVBUFFORCE " (since Linux 2.6.14)"
 561 Using this socket option, a privileged
 562 .RB ( CAP_NET_ADMIN )
 563 process can perform the same task as
 564 .BR SO_RCVBUF ,
 565 but the
 566 .I rmem_max
 567 limit can be overridden.
 568 .TP
 569 .BR SO_RCVLOWAT " and " SO_SNDLOWAT
 570 Specify the minimum number of bytes in the buffer until the socket layer
 571 will pass the data to the protocol
 572 .RB ( SO_SNDLOWAT )
 573 or the user on receiving
 574 .RB ( SO_RCVLOWAT ).
 575 These two values are initialized to 1.
 576 .B SO_SNDLOWAT
 577 is not changeable on Linux
 578 .RB ( setsockopt (2)
 579 fails with the error
 580 .BR ENOPROTOOPT ).
 581 .B SO_RCVLOWAT
 582 is changeable
 583 only since Linux 2.4.
 584 The
 585 .BR select (2)
 586 and
 587 .BR poll (2)
 588 system calls currently do not respect the
 589 .B SO_RCVLOWAT
 590 setting on Linux,
 591 and mark a socket readable when even a single byte of data is available.
 592 A subsequent read from the socket will block until
 593 .B SO_RCVLOWAT
 594 bytes are available.
 595 .\" See http://marc.theaimsgroup.com/?l=linux-kernel&m=111049368106984&w=2
 596 .\" Tested on kernel 2.6.14 -- mtk, 30 Nov 05
 597 .TP
 598 .BR SO_RCVTIMEO " and " SO_SNDTIMEO
 599 .\" Not implemented in 2.0.
 600 .\" Implemented in 2.1.11 for getsockopt: always return a zero struct.
 601 .\" Implemented in 2.3.41 for setsockopt, and actually used.
 602 Specify the receiving or sending timeouts until reporting an error.
 603 The argument is a
 604 .IR "struct timeval" .
 605 If an input or output function blocks for this period of time, and
 606 data has been sent or received, the return value of that function
 607 will be the amount of data transferred; if no data has been transferred
 608 and the timeout has been reached then \-1 is returned with
 609 .I errno
 610 set to
 611 .BR EAGAIN
 612 or
 613 .BR EWOULDBLOCK ,
 614 .\" in fact to EAGAIN
 615 or
 616 .B EINPROGRESS
 617 (for
 618 .BR connect (2))
 619 just as if the socket was specified to be nonblocking.
 620 If the timeout is set to zero (the default)
 621 then the operation will never timeout.
 622 Timeouts only have effect for system calls that perform socket I/O (e.g.,
 623 .BR read (2),
 624 .BR recvmsg (2),
 625 .BR send (2),
 626 .BR sendmsg (2));
 627 timeouts have no effect for
 628 .BR select (2),
 629 .BR poll (2),
 630 .BR epoll_wait (2),
 631 and so on.
 632 .TP
 633 .B SO_REUSEADDR
 634 Indicates that the rules used in validating addresses supplied in a
 635 .BR bind (2)
 636 call should allow reuse of local addresses.
 637 For
 638 .B AF_INET
 639 sockets this
 640 means that a socket may bind, except when there
 641 is an active listening socket bound to the address.
 642 When the listening socket is bound to
 643 .B INADDR_ANY
 644 with a specific port then it is not possible
 645 to bind to this port for any local address.
 646 Argument is an integer boolean flag.
 647 .TP
 648 .B SO_SNDBUF
 649 Sets or gets the maximum socket send buffer in bytes.
 650 The kernel doubles this value (to allow space for bookkeeping overhead)
 651 when it is set using
 652 .\" Most (all?) other implementations do not do this -- MTK, Dec 05
 653 .\" See also the comment to SO_RCVBUF (17 Jul 2012 LKML mail)
 654 .BR setsockopt (2),
 655 and this doubled value is returned by
 656 .BR getsockopt (2).
 657 The default value is set by the
 658 .I /proc/sys/net/core/wmem_default
 659 file and the maximum allowed value is set by the
 660 .I /proc/sys/net/core/wmem_max
 661 file.
 662 The minimum (doubled) value for this option is 2048.
 663 .TP
 664 .BR SO_SNDBUFFORCE " (since Linux 2.6.14)"
 665 Using this socket option, a privileged
 666 .RB ( CAP_NET_ADMIN )
 667 process can perform the same task as
 668 .BR SO_SNDBUF ,
 669 but the
 670 .I wmem_max
 671 limit can be overridden.
 672 .TP
 673 .B SO_TIMESTAMP
 674 Enable or disable the receiving of the
 675 .B SO_TIMESTAMP
 676 control message.
 677 The timestamp control message is sent with level
 678 .B SOL_SOCKET
 679 and the
 680 .I cmsg_data
 681 field is a
 682 .I "struct timeval"
 683 indicating the
 684 reception time of the last packet passed to the user in this call.
 685 See
 686 .BR cmsg (3)
 687 for details on control messages.
 688 .TP
 689 .B SO_TYPE
 690 Gets the socket type as an integer (e.g.,
 691 .BR SOCK_STREAM ).
 692 This socket option is read-only.
 693 .SS Signals
 694 When writing onto a connection-oriented socket that has been shut down
 695 (by the local or the remote end)
 696 .B SIGPIPE
 697 is sent to the writing process and
 698 .B EPIPE
 699 is returned.
 700 The signal is not sent when the write call
 701 specified the
 702 .B MSG_NOSIGNAL
 703 flag.
 704 .PP
 705 When requested with the
 706 .B FIOSETOWN
 707 .BR fcntl (2)
 708 or
 709 .B SIOCSPGRP
 710 .BR ioctl (2),
 711 .B SIGIO
 712 is sent when an I/O event occurs.
 713 It is possible to use
 714 .BR poll (2)
 715 or
 716 .BR select (2)
 717 in the signal handler to find out which socket the event occurred on.
 718 An alternative (in Linux 2.2) is to set a real-time signal using the
 719 .B F_SETSIG
 720 .BR fcntl (2);
 721 the handler of the real time signal will be called with
 722 the file descriptor in the
 723 .I si_fd
 724 field of its
 725 .IR siginfo_t .
 726 See
 727 .BR fcntl (2)
 728 for more information.
 729 .PP
 730 Under some circumstances (e.g., multiple processes accessing a
 731 single socket), the condition that caused the
 732 .B SIGIO
 733 may have already disappeared when the process reacts to the signal.
 734 If this happens, the process should wait again because Linux
 735 will resend the signal later.
 736 .\" .SS Ancillary messages
 737 .SS /proc interfaces
 738 The core socket networking parameters can be accessed
 739 via files in the directory
 740 .IR /proc/sys/net/core/ .
 741 .TP
 742 .I rmem_default
 743 contains the default setting in bytes of the socket receive buffer.
 744 .TP
 745 .I rmem_max
 746 contains the maximum socket receive buffer size in bytes which a user may
 747 set by using the
 748 .B SO_RCVBUF
 749 socket option.
 750 .TP
 751 .I wmem_default
 752 contains the default setting in bytes of the socket send buffer.
 753 .TP
 754 .I wmem_max
 755 contains the maximum socket send buffer size in bytes which a user may
 756 set by using the
 757 .B SO_SNDBUF
 758 socket option.
 759 .TP
 760 .IR message_cost " and " message_burst
 761 configure the token bucket filter used to load limit warning messages
 762 caused by external network events.
 763 .TP
 764 .I netdev_max_backlog
 765 Maximum number of packets in the global input queue.
 766 .TP
 767 .I optmem_max
 768 Maximum length of ancillary data and user control data like the iovecs
 769 per socket.
 770 .\" netdev_fastroute is not documented because it is experimental
 771 .SS Ioctls
 772 These operations can be accessed using
 773 .BR ioctl (2):
 774
 775 .in +4n
 776 .nf
 777 .IB error " = ioctl(" ip_socket ", " ioctl_type ", " &value_result ");"
 778 .fi
 779 .in
 780 .TP
 781 .B SIOCGSTAMP
 782 Return a
 783 .I struct timeval
 784 with the receive timestamp of the last packet passed to the user.
 785 This is useful for accurate round trip time measurements.
 786 See
 787 .BR setitimer (2)
 788 for a description of
 789 .IR "struct timeval" .
 790 .\"
 791 This ioctl should be used only if the socket option
 792 .B SO_TIMESTAMP
 793 is not set on the socket.
 794 Otherwise, it returns the timestamp of the
 795 last packet that was received while
 796 .B SO_TIMESTAMP
 797 was not set, or it fails if no such packet has been received,
 798 (i.e.,
 799 .BR ioctl (2)
 800 returns \-1 with
 801 .I errno
 802 set to
 803 .BR ENOENT ).
 804 .TP
 805 .B SIOCSPGRP
 806 Set the process or process group to send
 807 .B SIGIO
 808 or
 809 .B SIGURG
 810 signals
 811 to when an
 812 asynchronous I/O operation has finished or urgent data is available.
 813 The argument is a pointer to a
 814 .IR pid_t .
 815 If the argument is positive, send the signals to that process.
 816 If the
 817 argument is negative, send the signals to the process group with the ID
 818 of the absolute value of the argument.
 819 The process may only choose itself or its own process group to receive
 820 signals unless it has the
 821 .B CAP_KILL
 822 capability or an effective UID of 0.
 823 .TP
 824 .B FIOASYNC
 825 Change the
 826 .B O_ASYNC
 827 flag to enable or disable asynchronous I/O mode of the socket.
 828 Asynchronous I/O mode means that the
 829 .B SIGIO
 830 signal or the signal set with
 831 .B F_SETSIG
 832 is raised when a new I/O event occurs.
 833 .IP
 834 Argument is an integer boolean flag.
 835 (This operation is synonymous with the use of
 836 .BR fcntl (2)
 837 to set the
 838 .B O_ASYNC
 839 flag.)
 840 .\"
 841 .TP
 842 .B SIOCGPGRP
 843 Get the current process or process group that receives
 844 .B SIGIO
 845 or
 846 .B SIGURG
 847 signals,
 848 or 0
 849 when none is set.
 850 .PP
 851 Valid
 852 .BR fcntl (2)
 853 operations:
 854 .TP
 855 .B FIOGETOWN
 856 The same as the
 857 .B SIOCGPGRP
 858 .BR ioctl (2).
 859 .TP
 860 .B FIOSETOWN
 861 The same as the
 862 .B SIOCSPGRP
 863 .BR ioctl (2).
 864 .SH VERSIONS
 865 .B SO_BINDTODEVICE
 866 was introduced in Linux 2.0.30.
 867 .B SO_PASSCRED
 868 is new in Linux 2.2.
 869 The
 870 .I /proc
 871 interfaces was introduced in Linux 2.2.
 872 .B SO_RCVTIMEO
 873 and
 874 .B SO_SNDTIMEO
 875 are supported since Linux 2.3.41.
 876 Earlier, timeouts were fixed to
 877 a protocol-specific setting, and could not be read or written.
 878 .SH NOTES
 879 Linux assumes that half of the send/receive buffer is used for internal
 880 kernel structures; thus the values in the corresponding
 881 .I /proc
 882 files are twice what can be observed on the wire.
 883
 884 Linux will only allow port reuse with the
 885 .B SO_REUSEADDR
 886 option
 887 when this option was set both in the previous program that performed a
 888 .BR bind (2)
 889 to the port and in the program that wants to reuse the port.
 890 This differs from some implementations (e.g., FreeBSD)
 891 where only the later program needs to set the
 892 .B SO_REUSEADDR
 893 option.
 894 Typically this difference is invisible, since, for example, a server
 895 program is designed to always set this option.
 896 .SH BUGS
 897 The
 898 .B CONFIG_FILTER
 899 socket options
 900 .B SO_ATTACH_FILTER
 901 and
 902 .B SO_DETACH_FILTER
 903 .\" FIXME Document SO_ATTACH_FILTER and SO_DETACH_FILTER
 904 are not documented.
 905 The suggested interface to use them is via the libpcap
 906 library.
 907 .\" .SH AUTHORS
 908 .\" This man page was written by Andi Kleen.
 909 .SH SEE ALSO
 910 .BR getsockopt (2),
 911 .BR connect (2),
 912 .BR setsockopt (2),
 913 .BR socket (2),
 914 .BR capabilities (7),
 915 .BR ddp (7),
 916 .BR ip (7),
 917 .BR packet (7),
 918 .BR tcp (7),
 919 .BR udp (7),
 920 .BR unix (7)