tcp协议系列文章(6):send



NAME

send, sendto, sendmsg - send a message on a socket

SYNOPSIS

       #include <sys/types.h>
       #include <sys/socket.h>

       ssize_t send(int sockfd, const void *buf, size_t len, int flags);

       ssize_t sendto(int sockfd, const void *buf, size_t len, int flags,
                      const struct sockaddr *dest_addr, socklen_t addrlen);

       ssize_t sendmsg(int sockfd, const struct msghdr *msg, int flags);


DESCRIPTION

       The system calls send(), sendto(), and sendmsg() are used to transmit a message to another socket.

       The  send()  call  may  be  used  only  when the socket is in a connected state (so that the intended recipient is known).  The only difference between send() and write(2) is the presence of flags.  With a zero flags argument, send() is equivalent to write(2).  Also, the following call

           send(sockfd, buf, len, flags);

       is equivalent to

           sendto(sockfd, buf, len, flags, NULL, 0);

       The argument sockfd is the file descriptor of the sending socket.

       send()系统调用通常只用与一个处于connected状态的socket(接收方是提前已知的)。send()和write(2)的唯一区别在是否有参数flags。当flags传0时,send()等价于write(2)。另外,下面两个调用也相互等价:
       send(sockfd, buf, len, flags);
       sendto(sockfd, buf, len, flags, NULL, 0);
       参数sockfd是发送socket的fd。

       If sendto() is used on a connection-mode (SOCK_STREAM, SOCK_SEQPACKET) socket, the arguments dest_addr and addrlen are ignored  (and  the  error  EISCONN  may  be returned  when  they  are not NULL and 0), and the error ENOTCONN is returned when the socket was not actually connected.  Otherwise, the address of the target is given by dest_addr with addrlen specifying its size.  For sendmsg(), the address of the target is given by msg.msg_name, with msg.msg_namelen specifying its size.

       如果sendto()用于基于连接的socket(SOCK_STREAM,SOCK_SEQPACKET),参数dest_addr和addrlen将被忽略(并且当它们不为NULL或0时,错误EISCONN将会返回),当socket没有连接时会返回ENOTCONN错误。否则,目标地址需要通过dest_addr给出,addrlen指出dest_addr的长度。对于sendmsg(),目标地址由msg.msg_name给出,msg.msg_namelen指示其长度。


       For send() and sendto(), the message is found in buf and has length len.  For sendmsg(), the message is pointed to by the elements of the array msg.msg_iov.   The sendmsg() call also allows sending ancillary data (also known as control information).
       
       对于send() 和sendto(),消息在buf中存放,消息长度为len。对于sendmsg(),消息用数组msg.msg_iov指向。sendmsg()系统调用允许发送辅助数据(即所谓的控制信息)。


       If the message is too long to pass atomically through the underlying protocol, the error EMSGSIZE is returned, and the message is not transmitted.

       
       如果消息太长而无法原子性传给下层协议,会返回EMSGSIZE错误,并且消息不回传递。

       No indication of failure to deliver is implicit in a send().  Locally detected errors are indicated by a return value of -1.
       
       send()没有明显的失败迹象。本地代码通过-1返回值检测错误。


       When the message does not fit into the send buffer of the socket, send() normally blocks, unless the socket has been placed in nonblocking I/O mode.  In nonblocking mode it would fail with the error EAGAIN or EWOULDBLOCK in this case.  The select(2) call may be used to determine when it is possible to send more data.
       
       当消息长度比socket的发送缓冲区大时,send()通常会阻塞,除非这个socket被设置为非阻塞I/O模式。在非阻塞模式,这种情况通常会返回EAGAIN或EWOULDBLOCK错误。select(2)可以用于检测何时可以发送更多数据。(笔者注:send()的行为是否是这样,可能依赖于send()的版本。我用的linux系统是3.10版本。调用send时,当需要发送的数据大小超过sokcet可用发送缓冲区大小时,send的行为是它会拷贝一部分字节到发送缓冲区、并返回拷贝的数量。而不是如man send中所说的返回EAGAIN的错误码。)


       The flags argument is the bitwise OR of zero or more of the following flags.
       参数flags是基于bit的或(OR)的关系,可以赋值为0或下面标识的多个值。

       MSG_CONFIRM (Since Linux 2.3.15)
              Tell the link layer that forward progress happened: you got a successful reply from the other side.  If the link layer doesn't get this it  will  regularly reprobe  the  neighbor  (e.g.,  via  a  unicast ARP).  Only valid on SOCK_DGRAM and SOCK_RAW sockets and currently implemented only for IPv4 and IPv6.  See arp(7) for details.
              告诉链路层,向前的进展发生了:你从对端得到了成功的回复。如果链路层没收到这个消息,它将定期重新探测邻居(比如,使用单播ARP)。只在SOCK_DGRAM或SOCK_RAW的socket上有效,现在只在IPv4和IPv6上实现了。查看arp(7)获取更多细节。


       MSG_DONTROUTE
              Don't use a gateway to send out the packet, send to hosts only on directly connected networks.  This is usually used only by  diagnostic  or  routing  programs.  This is defined only for protocol families that route; packet sockets don't.
              不使用网关发送数据包,在直连的网络上向主机发送。这通常只用于诊断或路由程序。这仅为路由的协议簇定义,包socket没有。

       MSG_DONTWAIT (since Linux 2.2)
              Enables nonblocking operation; if the operation would block, EAGAIN or EWOULDBLOCK is returned (this can also be enabled using the O_NONBLOCK flag with the F_SETFL fcntl(2)).
              使能非阻塞操作。如果操作会阻塞,EAGAIN或EWOULDBLOCK会返回(也可以通过F_SETFL fcntl(2)的O_NONBLOCK标识来设置生效)。


       MSG_MORE (Since Linux 2.4.4)
              The caller has more data to send.  This flag is used with TCP sockets to obtain the same effect as the TCP_CORK socket option (see tcp(7)), with  the  difference that this flag can be set on a per-call basis.

              Since  Linux  2.6,  this  flag is also supported for UDP sockets, and informs the kernel to package all of the data sent in calls with this flag set into a single datagram which is transmitted only when a call is performed that does not specify this flag.  (See also the  UDP_CORK  socket  option  described  in udp(7).)
              调用者有更多数据要发送。这个标志用于tcp socket以取得跟TCP_CORK选项相同的效果(见tcp(7)),它们的区别是这个标志可以每次调用时设置。
              自linux2.6,该标志也在UDP上支持,当一次(没有指明该标志的)调用被执行时,通知内核将所有(发送时设置了该标志的)数据打包成一个单独的数据报来发送。
              笔者注:用于udp时,sendto会调用N次,其中前N-1次的flags设置了MSG_MORE,最后的第N次调用没有设置MSG_MORE,那么内核会将这所有的N次送入的buffer打包成一个udp包发送出去。


       MSG_NOSIGNAL (since Linux 2.2)
              Requests not to send SIGPIPE on errors on stream oriented sockets when the other end breaks the connection.  The EPIPE error is still returned.
              当流式socket的对端关闭连接时,请求在发生错误时不要发送SIGPIPE。但是EPIPE仍然会返回。

       MSG_OOB
              Sends out-of-band data on sockets that support this notion (e.g., of type SOCK_STREAM); the underlying protocol must also support out-of-band data.
              在支持OOB概念的socket(如SOCK_STREAM类型)上发送带外数据,下层协议栈也必须支持带外数据概念。


       The definition of the msghdr structure follows.  See recv(2) and below for an exact description of its fields.
       msghdr结构体如下定义,获取对其域的准确描述,查看recv(2)和下面的内容。

           struct msghdr {
               void         *msg_name;       /* optional address */
               socklen_t     msg_namelen;    /* size of address */
               struct iovec *msg_iov;        /* scatter/gather array */
               size_t        msg_iovlen;     /* # elements in msg_iov */
               void         *msg_control;    /* ancillary data, see below */
               size_t        msg_controllen; /* ancillary data buffer len */
               int           msg_flags;      /* flags on received message */
           };

       You may send control information using the msg_control and msg_controllen members.  The maximum control buffer length the kernel can process is limited per socket by the value in /proc/sys/net/core/optmem_max; see socket(7).
       你可以借助msg_control和msg_controllen发送控制消息。内核能处理的控制消息的最大长度,对每个socket都有限制,在文件/proc/sys/net/core/optmem_max中限制该值。详见socket(7)。


RETURN VALUE

       On success, these calls return the number of characters sent.  On error, -1 is returned, and errno is set appropriately.
       成功返回发送的字节数,失败返回-1,并设置相应的错误码到errno。

ERRORS

        These are some standard errors generated by the socket layer.  Additional errors may be generated and returned from the underlying  protocol  modules;  see  their respective manual pages.
        socket层会生成一些标准错误。更多错误会由下层协议模块生成并返回,查看它们各自的手册以获取更多信息。
        
        EACCES (For  UNIX  domain sockets, which are identified by pathname) Write permission is denied on the destination socket file, or search permission is denied for one of the directories the path prefix.  (See path_resolution(7).)
               (For UDP sockets) An attempt was made to send to a network/broadcast address as though it was a unicast address.
        EACCES(用于UNIX域socket,使用pathname来标识)到目的的socket文件的写权限被拒绝,或路径前缀下的任何一个目录的搜索权限被拒绝。(查看path_resolution(7))

        (对于UDPsocket)尝试发送一个网络/广播地址,就好像它是一个单播的地址。


        EAGAIN or EWOULDBLOCK
               The socket is marked nonblocking and the requested operation would block.  POSIX.1-2001 allows either error to be returned for  this  case,  and  does  not require these constants to have the same value, so a portable application should check for both possibilities.
        EAGAIN or EWOULDBLOCK 标记为非阻塞的socket,在执行请求的操作时会阻塞。POSIX.1-2001允许在这种情形返回任何错误,且不允许这两个常量有相同的值,所以,一个可移植的程序需要检查这两个值。

        EBADF  An invalid descriptor was specified.

        
        ECONNRESET
               Connection reset by peer.
        
        EDESTADDRREQ
               The socket is not connection-mode, and no peer address is set.

        EFAULT An invalid user space address was specified for an argument.

        
        EINTR  A signal occurred before any data was transmitted; see signal(7).
        
        EINVAL Invalid argument passed.

        EISCONN
               The  connection-mode  socket  was  connected  already but a recipient was specified.  (Now either this error is returned, or the recipient specification is ignored.)
               socket已经处于连接状态,但是又指定了一个接收者。(现在,要么返回这个错误,要么新指定的接收者被忽略。)
        
        EMSGSIZE
               The socket type requires that message be sent atomically, and the size of the message to be sent made this impossible.
               这种socket类型要求消息被原子发送,但是消息的长度却使的原子性发送变得不可能。

        ENOBUFS
               The output queue for a network interface was full.  This generally indicates that the interface has stopped sending, but may be caused by transient congestion.  (Normally, this does not occur in Linux.  Packets are just silently dropped when a device queue overflows.)
               网卡(网络接口)的输出队列已满。这通常表明网卡已经停止发送,有可能是拥塞引起。(通常,这在linux不会发生。当设备的队列溢出时数据包会被静默丢弃。)
        
        ENOMEM No memory available.
        
        ENOTCONN
               The socket is not connected, and no target has been given.
        
        ENOTSOCK
               The argument sockfd is not a socket.

        EOPNOTSUPP
               Some bit in the flags argument is inappropriate for the socket type.
               flags中的某些位对这种socket类型来讲不合法。
        
        EPIPE  The local end has been shut down on a connection oriented socket.  In this case the process will also receive a SIGPIPE unless MSG_NOSIGNAL is set.
        本端关闭了一个基于连接的socket。在这种情况下,进程会收到SIGPIPE信号除非设置了MSG_NOSIGNAL。

CONFORMING TO 遵循

       4.4BSD, SVr4, POSIX.1-2001.  These function calls appeared in 4.2BSD.

       POSIX.1-2001 describes only the MSG_OOB and MSG_EOR flags.  POSIX.1-2008 adds a specification of MSG_NOSIGNAL.  The MSG_CONFIRM flag is a Linux extension.
       
       遵循4.4BSD, SVr4, POSIX.1-2001。这心函数出现于4.2BSD。
       POSIX.1-2001只描述了MSG_OOB和MSG_EOR。POSIX.1-2008增加MSG_NOSIGNAL的描述。MSG_CONFIRM是一个linux扩展。

NOTES

       The  prototypes given above follow the Single UNIX Specification, as glibc2 also does; the flags argument was int in 4.x BSD, but unsigned int in libc4 and libc5; the len argument was int in 4.x BSD and libc4, but size_t in libc5; the addrlen argument was int in 4.x BSD and libc4 and libc5.  See also accept(2).

       According to POSIX.1-2001, the msg_controllen field of the msghdr structure should be typed as socklen_t, but glibc currently types it as size_t.

       See sendmmsg(2) for information about a Linux-specific system call that can be used to transmit multiple datagrams in a single call.

       上面给出的原型遵循Single UNIX Specification,就如同glibc2那样。参数flags在4.x BSD中是int类型,但是在libc4和libc5中是unsigned int类型。参数len在4.x BSD和libc4中是int类型,但在libc5中是size_t类型。参数addrlen在4.x BSD、libc4和libc5中都是int类型。另见accept(2)。
       
       根据POSIX.1-2001,msghdr结构体的msg_controllen域必须是socklen_t类型,但是在glibc中现在确实size_t类型。
       
       参考sendmsg(2),获取关于在特定Linux系统上,在单次系统调用时传送多个数据报(datagrams)的信息。

BUGS

       Linux may return EPIPE instead of ENOTCONN.

EXAMPLE

       An example of the use of sendto() is shown in getaddrinfo(3).

SEE ALSO

       fcntl(2), getsockopt(2), recv(2), select(2), sendfile(2), sendmmsg(2), shutdown(2), socket(2), write(2), cmsg(3), ip(7), socket(7), tcp(7), udp(7)

COLOPHON

       This page is part of release 3.53 of the Linux man-pages project.  A description  of  the  project,  and  information  about  reporting  bugs,  can  be  found  at
       http://www.kernel.org/doc/man-pages/.


  • 1
    点赞
  • 0
    收藏
    觉得还不错? 一键收藏
  • 0
    评论

“相关推荐”对你有帮助么?

  • 非常没帮助
  • 没帮助
  • 一般
  • 有帮助
  • 非常有帮助
提交
评论
添加红包

请填写红包祝福语或标题

红包个数最小为10个

红包金额最低5元

当前余额3.43前往充值 >
需支付:10.00
成就一亿技术人!
领取后你会自动成为博主和红包主的粉丝 规则
hope_wisdom
发出的红包
实付
使用余额支付
点击重新获取
扫码支付
钱包余额 0

抵扣说明:

1.余额是钱包充值的虚拟货币,按照1:1的比例进行支付金额的抵扣。
2.余额无法直接购买下载,可以购买VIP、付费专栏及课程。

余额充值