Linux 高级I/O函数之sendfile

linux手册介绍sendfile函数:

NAME         

sendfile - transfer data between file descriptors

SYNOPSIS          

#include <sys/sendfile.h>

       ssize_t sendfile(int out_fd, int in_fd, off_t *offset, size_t count);

DESCRIPTION     

sendfile() copies data between one file descriptor and another.
       Because this copying is done within the kernel, sendfile() is more
       efficient than the combination of read(2) and write(2), which would
       require transferring data to and from user space.

       in_fd should be a file descriptor opened for reading and out_fd
       should be a descriptor opened for writing.

       If offset is not NULL, then it points to a variable holding the file
       offset from which sendfile() will start reading data from in_fd.
       When sendfile() returns, this variable will be set to the offset of
       the byte following the last byte that was read.  If offset is not
       NULL, then sendfile() does not modify the file offset of in_fd;
       otherwise the file offset is adjusted to reflect the number of bytes
       read from in_fd.

       If offset is NULL, then data will be read from in_fd starting at the
       file offset, and the file offset will be updated by the call.

       count is the number of bytes to copy between the file descriptors.

       The in_fd argument must correspond to a file which supports
       mmap(2)-like operations (i.e., it cannot be a socket).（

in_fd必须是一个支持类似mmap函数的文件描述符，即它必须指向真实的文件，不能是socket和管道

）

       In Linux kernels before 2.6.33, out_fd must refer to a socket.  Since
       Linux 2.6.33 it can be any file.  If it is a regular file, then
       sendfile() changes the file offset appropriately.(从上面说的可以看出:in_fd必须指向真实的文件，而out_fd在2.6.33可以是任何fd(不一定是socket。由此可见，sendfile几乎是专门为在网络上传输文件而设计的。)

RETURN VALUE          

If the transfer was successful, the number of bytes written to out_fd
       is returned.  Note that a successful call to sendfile() may write
       fewer bytes than requested; the caller should be prepared to retry
       the call if there were unsent bytes.  See also NOTES.

       On error, -1 is returned, and errno is set appropriately.

ERRORS        

EAGAIN Nonblocking I/O has been selected using O_NONBLOCK and the
              write would block.

       EBADF  The input file was not opened for reading or the output file
              was not opened for writing.

       EFAULT Bad address.

       EINVAL Descriptor is not valid or locked, or an mmap(2)-like
              operation is not available for in_fd, or count is negative.

       EINVAL out_fd has the O_APPEND flag set.  This is not currently
              supported by sendfile().

       EIO    Unspecified error while reading from in_fd.

       ENOMEM Insufficient memory to read from in_fd.

       EOVERFLOW
              count is too large, the operation would result in exceeding
              the maximum size of either the input file or the output file.

       ESPIPE offset is not NULL but the input file is not seek(2)-able.

VERSIONS       

sendfile() first appeared in Linux 2.2.  The include file
       <sys/sendfile.h> is present since glibc 2.1.

CONFORMING TO       

Not specified in POSIX.1-2001, nor in other standards.

       Other UNIX systems implement sendfile() with different semantics and
       prototypes.  It should not be used in portable programs.

NOTES        

sendfile() will transfer at most 0x7ffff000 (2,147,479,552) bytes,
       returning the number of bytes actually transferred.  (This is true on
       both 32-bit and 64-bit systems.)

       If you plan to use sendfile() for sending files to a TCP socket, but
       need to send some header data in front of the file contents, you will
       find it useful to employ the TCP_CORK option, described in tcp(7), to
       minimize the number of packets and to tune performance.

       In Linux 2.4 and earlier, out_fd could also refer to a regular file;
       this possibility went away in the Linux 2.6.x kernel series, but was
       restored in Linux 2.6.33.

       The original Linux sendfile() system call was not designed to handle
       large file offsets.  Consequently, Linux 2.4 added sendfile64(), with
       a wider type for the offset argument.  The glibc sendfile() wrapper
       function transparently deals with the kernel differences.

       Applications may wish to fall back to read(2)/write(2) in the case
       where sendfile() fails with EINVAL or ENOSYS.

       If out_fd refers to a socket or pipe with zero-copy support, callers
       must ensure the transferred portions of the file referred to by in_fd
       remain unmodified until the reader on the other end of out_fd has
       consumed the transferred data.

       The Linux-specific splice(2) call supports transferring data between
       arbitrary file descriptors provided one (or both) of them is a pipe.

http://man7.org/linux/man-pages/man2/sendfile.2.html

首先我们来看看传统的read/write方式进行socket的传输。
当需要对一个文件进行传输的时候，具体流程细节如下：

1：调用read函数，文件数据copy到内核缓冲区
2：read函数返回，文件数据从内核缓冲区copy到用户缓冲区
3：write函数调用，将文件数据从用户缓冲区copy到内核与socket相关的缓冲区
4：数据从socket缓冲区copy到相关协议引擎。
在这个过程中发生了四次copy操作。

硬盘->内核->用户->socket缓冲区（内核）->协议引擎。

而sendfile的工作原理呢？？

1、系统调用 sendfile() 通过 DMA 把硬盘数据拷贝到 kernel buffer，然后数据被 kernel 直接拷贝到另外一个与 socket 相关的 kernel buffer。这里没有 用户态和核心态 之间的切换，在内核中直接完成了从一个 buffer 到另一个 buffer 的拷贝。
2、DMA 把数据从 kernel buffer 直接拷贝给协议栈，没有切换，也不需要数据从用户态和核心态，因为数据就在 kernel 里。
———————————————— 

#include <sys/socket.h>
#include <netinet/in.h>
#include <arpa/inet.h>
#include <assert.h>
#include <stdio.h>
#include <errno.h>
#include <unistd.h>
#include <string.h>
#include <stdlib.h>
#include <sys/stat.h>
#include <sys/types.h>
#include <fcntl.h>
#include <sys/sendfile.h>


int main(int argc,char *argv[]) { 
    if (argc <= 3) {
        printf("usage:%s ip port filename\n",argv[0]);
        return 1;
    }
    
    const char *ip = argv[1];
    int port = atoi(argv[2]);
    const char *file_name = argv[3];
    
    int filefd = open(file_name,O_RDONLY);
    assert(filefd > 0);
    struct stat file_stat;
    //为了获取文件大小
    fstat(filefd,&file_stat);

    struct sockaddr_in address;
    bzero(&address,sizeof(address));
    address.sin_family = AF_INET;
    inet_pton(AF_INET,ip,&address.sin_addr);
    address.sin_port = htons(port);

    int sock = socket(PF_INET,SOCK_STREAM,0);
    assert(sock >= 0);
    int reuse = 1;
    setsockopt(sock,SOL_SOCKET,SO_REUSEADDR,&reuse,sizeof(reuse));
    
    int ret = bind(sock,(struct sockaddr *)&address,sizeof(address));
    assert(ret != -1);

    ret = listen(sock,5);
    assert(ret != -1);

    struct sockaddr_in client;
    socklen_t client_addrlen = sizeof(client);
    int connfd = accept(sock,(struct sockaddr *)&client,&client_addrlen);
    if (connfd < 0) {
        printf("errno is :%d\n",errno);
    }else {
        sendfile(connfd,filefd,NULL,file_stat.st_size);
        close(filefd);
        close(connfd);
    }
    close(sock);
    return 0;
}

g++ sendfile.c -o sendfile
#服务器端
[hadoop@master Linux]$ ./sendfile master 5432 makefile 
#客户端
[root@slave07]~# telnet 10.10.18.229 5432 
Trying 10.10.18.229...
Connected to master (10.10.18.229).
Escape character is ‘^]‘.
all:send rec

send:send.c
    gcc  $? -o 

rec:receve.c
    gcc $? -o 

Connection closed by foreign host.