select 模型

SELECT(2)                                                          Linux Programmer's Manual                                                          SELECT(2)

NAME
       select, pselect, FD_CLR, FD_ISSET, FD_SET, FD_ZERO - synchronous I/O multiplexing

SYNOPSIS
       /* According to POSIX.1-2001, POSIX.1-2008 */
       #include <sys/select.h>

       /* According to earlier standards */
       #include <sys/time.h>
       #include <sys/types.h>
       #include <unistd.h>

       int select(int nfds, fd_set *readfds, fd_set *writefds,
                  fd_set *exceptfds, struct timeval *timeout);

       void FD_CLR(int fd, fd_set *set);
       int  FD_ISSET(int fd, fd_set *set);
       void FD_SET(int fd, fd_set *set);
       void FD_ZERO(fd_set *set);

       #include <sys/select.h>

       int pselect(int nfds, fd_set *readfds, fd_set *writefds,
                   fd_set *exceptfds, const struct timespec *timeout,
                   const sigset_t *sigmask);

   Feature Test Macro Requirements for glibc (see feature_test_macros(7)):

       pselect(): _POSIX_C_SOURCE >= 200112L || _XOPEN_SOURCE >= 600

DESCRIPTION
       select()  and  pselect() allow a program to monitor multiple file descriptors, waiting until one or more of the file descriptors become "ready" for some
       class of I/O operation (e.g., input possible).  A file descriptor is considered ready if it is possible to perform a corresponding I/O operation  (e.g.,
       read(2) without blocking, or a sufficiently small write(2)).

       The operation of select() and pselect() is identical, other than these three differences:

       (i)    select()  uses  a  timeout  that  is  a  struct timeval (with seconds and microseconds), while pselect() uses a struct timespec (with seconds and
              nanoseconds).

       (ii)   select() may update the timeout argument to indicate how much time was left.  pselect() does not change this argument.

       (iii)  select() has no sigmask argument, and behaves as pselect() called with NULL sigmask.

       Three independent sets of file descriptors are watched.  Those listed in readfds will be watched to see if characters become available for reading (more
       precisely,  to see if a read will not block; in particular, a file descriptor is also ready on end-of-file), those in writefds will be watched to see if
       space is available for write (though a large write may still block), and those in exceptfds will be watched for exceptions.  On exit, the sets are modi‐
       fied  in  place  to indicate which file descriptors actually changed status.  Each of the three file descriptor sets may be specified as NULL if no file
       descriptors are to be watched for the corresponding class of events.

       Four macros are provided to manipulate the sets.  FD_ZERO() clears a set.  FD_SET() and FD_CLR() respectively add and remove  a  given  file  descriptor
       from a set.  FD_ISSET() tests to see if a file descriptor is part of the set; this is useful after select() returns.

       nfds is the highest-numbered file descriptor in any of the three sets, plus 1.

       The timeout argument specifies the interval that select() should block waiting for a file descriptor to become ready.  The call will block until either:

       *  a file descriptor becomes ready;

       *  the call is interrupted by a signal handler; or

       *  the timeout expires.

       Note  that  the  timeout  interval  will be rounded up to the system clock granularity, and kernel scheduling delays mean that the blocking interval may
       overrun by a small amount.  If both fields of the timeval structure are zero, then select() returns immediately.  (This  is  useful  for  polling.)   If
       timeout is NULL (no timeout), select() can block indefinitely.

       sigmask  is a pointer to a signal mask (see sigprocmask(2)); if it is not NULL, then pselect() first replaces the current signal mask by the one pointed
       to by sigmask, then does the "select" function, and then restores the original signal mask.

       Other than the difference in the precision of the timeout argument, the following pselect() call:

           ready = pselect(nfds, &readfds, &writefds, &exceptfds,
                           timeout, &sigmask);

       is equivalent to atomically executing the following calls:

           sigset_t origmask;

           pthread_sigmask(SIG_SETMASK, &sigmask, &origmask);
           ready = select(nfds, &readfds, &writefds, &exceptfds, timeout);
           pthread_sigmask(SIG_SETMASK, &origmask, NULL);

       The reason that pselect() is needed is that if one wants to wait for either a signal or for a file descriptor to become ready, then an  atomic  test  is
       needed  to  prevent race conditions.  (Suppose the signal handler sets a global flag and returns.  Then a test of this global flag followed by a call of
       select() could hang indefinitely if the signal arrived just after the test but just before the call.  By contrast, pselect() allows one to  first  block
       signals, handle the signals that have come in, then call pselect() with the desired sigmask, avoiding the race.)

   The timeout
       The time structures involved are defined in <sys/time.h> and look like

           struct timeval {
               long    tv_sec;         /* seconds */
               long    tv_usec;        /* microseconds */
           };

       and

           struct timespec {
               long    tv_sec;         /* seconds */
               long    tv_nsec;        /* nanoseconds */
           };

       (However, see below on the POSIX.1 versions.)

       Some code calls select() with all three sets empty, nfds zero, and a non-NULL timeout as a fairly portable way to sleep with subsecond precision.

       On  Linux, select() modifies timeout to reflect the amount of time not slept; most other implementations do not do this.  (POSIX.1 permits either behav‐
       ior.)  This causes problems both when Linux code which reads timeout is ported to other operating systems, and when code is ported to Linux that  reuses
       a struct timeval for multiple select()s in a loop without reinitializing it.  Consider timeout to be undefined after select() returns.

RETURN VALUE
       On  success,  select() and pselect() return the number of file descriptors contained in the three returned descriptor sets (that is, the total number of
       bits that are set in readfds, writefds, exceptfds) which may be zero if the timeout expires before  anything  interesting  happens.   On  error,  -1  is
       returned, and errno is set to indicate the error; the file descriptor sets are unmodified, and timeout becomes undefined.

ERRORS
       EBADF  An  invalid  file  descriptor  was  given  in  one of the sets.  (Perhaps a file descriptor that was already closed, or one on which an error has
              occurred.)

       EINTR  A signal was caught; see signal(7).

       EINVAL nfds is negative or exceeds the RLIMIT_NOFILE resource limit (see getrlimit(2)).

       EINVAL the value contained within timeout is invalid.

       ENOMEM unable to allocate memory for internal tables.

VERSIONS
       pselect() was added to Linux in kernel 2.6.16.  Prior to this, pselect() was emulated in glibc (but see BUGS).

CONFORMING TO
       select() conforms to POSIX.1-2001, POSIX.1-2008, and 4.4BSD (select() first appeared in 4.2BSD).  Generally portable to/from non-BSD systems  supporting
       clones  of the BSD socket layer (including System V variants).  However, note that the System V variant typically sets the timeout variable before exit,
       but the BSD variant does not.

       pselect() is defined in POSIX.1g, and in POSIX.1-2001 and POSIX.1-2008.

NOTES
       An fd_set is a fixed size buffer.  Executing FD_CLR() or FD_SET() with a value of fd that is negative or is equal to  or  larger  than  FD_SETSIZE  will
       result in undefined behavior.  Moreover, POSIX requires fd to be a valid file descriptor.

       Concerning  the types involved, the classical situation is that the two fields of a timeval structure are typed as long (as shown above), and the struc‐
       ture is defined in <sys/time.h>.  The POSIX.1 situation is

           struct timeval {
               time_t         tv_sec;     /* seconds */
               suseconds_t    tv_usec;    /* microseconds */
           };

       where the structure is defined in <sys/select.h> and the data types time_t and suseconds_t are defined in <sys/types.h>.

       Concerning prototypes, the classical situation is that one should include <time.h> for select().  The POSIX.1  situation  is  that  one  should  include
       <sys/select.h> for select() and pselect().

       Under  glibc  2.0,  <sys/select.h>  gives  the wrong prototype for pselect().  Under glibc 2.1 to 2.2.1, it gives pselect() when _GNU_SOURCE is defined.
       Since glibc 2.2.2, the requirements are as shown in the SYNOPSIS.

   Multithreaded applications
       If a file descriptor being monitored by select() is closed in another thread, the result is unspecified.  On some UNIX systems,  select()  unblocks  and
       returns,  with  an  indication  that  the  file  descriptor is ready (a subsequent I/O operation will likely fail with an error, unless another the file
       descriptor reopened between the time select() returned and the I/O operations was performed).  On Linux (and  some  other  systems),  closing  the  file
       descriptor  in another thread has no effect on select().  In summary, any application that relies on a particular behavior in this scenario must be con‐
       sidered buggy.

   C library/kernel differences
       The pselect() interface described in this page is implemented by glibc.  The underlying Linux system call is named pselect6().   This  system  call  has
       somewhat different behavior from the glibc wrapper function.

       The  Linux  pselect6() system call modifies its timeout argument.  However, the glibc wrapper function hides this behavior by using a local variable for
       the timeout argument that is passed to the system call.  Thus, the glibc pselect() function does not modify its timeout argument; this is  the  behavior
       required by POSIX.1-2001.

       The final argument of the pselect6() system call is not a sigset_t * pointer, but is instead a structure of the form:

           struct {
               const sigset_t *ss;     /* Pointer to signal set */
               size_t          ss_len; /* Size (in bytes) of object pointed
                                          to by 'ss' */
           };

       This allows the system call to obtain both a pointer to the signal set and its size, while allowing for the fact that most architectures support a maxi‐
       mum of 6 arguments to a system call.

BUGS
       Glibc 2.0 provided a version of pselect() that did not take a sigmask argument.

       Starting with version 2.1, glibc provided an emulation of pselect() that  was  implemented  using  sigprocmask(2)  and  select().   This  implementation
       remained  vulnerable  to  the very race condition that pselect() was designed to prevent.  Modern versions of glibc use the (race-free) pselect() system
       call on kernels where it is provided.

       On systems that lack pselect(), reliable (and more portable) signal trapping can be achieved using the self-pipe trick.  In  this  technique,  a  signal
       handler  writes  a byte to a pipe whose other end is monitored by select() in the main program.  (To avoid possibly blocking when writing to a pipe that
       may be full or reading from a pipe that may be empty, nonblocking I/O is used when reading from and writing to the pipe.)

       Under Linux, select() may report a socket file descriptor as "ready for reading", while nevertheless a subsequent read blocks.  This could  for  example
       happen  when  data has arrived but upon examination has wrong checksum and is discarded.  There may be other circumstances in which a file descriptor is
       spuriously reported as ready.  Thus it may be safer to use O_NONBLOCK on sockets that should not block.

       On Linux, select() also modifies timeout if the call is interrupted by a signal handler (i.e., the EINTR  error  return).   This  is  not  permitted  by
       POSIX.1.   The Linux pselect() system call has the same behavior, but the glibc wrapper hides this behavior by internally copying the timeout to a local
       variable and passing that variable to the system call.

EXAMPLE
       #include <stdio.h>
       #include <stdlib.h>
       #include <sys/time.h>
       #include <sys/types.h>
       #include <unistd.h>

       int
       main(void)
       {
           fd_set rfds;
           struct timeval tv;
           int retval;

           /* Watch stdin (fd 0) to see when it has input. */
           FD_ZERO(&rfds);
           FD_SET(0, &rfds);

           /* Wait up to five seconds. */
           tv.tv_sec = 5;
           tv.tv_usec = 0;

           retval = select(1, &rfds, NULL, NULL, &tv);
           /* Don't rely on the value of tv now! */

           if (retval == -1)
               perror("select()");
           else if (retval)
               printf("Data is available now.\n");
               /* FD_ISSET(0, &rfds) will be true. */
           else
               printf("No data within five seconds.\n");

           exit(EXIT_SUCCESS);
       }

SEE ALSO
       accept(2), connect(2), poll(2), read(2), recv(2), restart_syscall(2), send(2), sigprocmask(2), write(2), epoll(7), time(7)

       For a tutorial with discussion and examples, see select_tut(2).

COLOPHON
       This page is part of release 4.04 of the Linux man-pages project.  A description of the project, information about reporting bugs, and the  latest  ver‐
       sion of this page, can be found at http://www.kernel.org/doc/man-pages/.

Linux                                                                      2015-07-23                                                                 SELECT(2)

转载于:https://www.cnblogs.com/rushoutasia/p/5806733.html