pipe.texi

@node Pipes and FIFOs, Sockets, File System Interface, Top
@c %MENU% A simple interprocess communication mechanism
@chapter Pipes and FIFOs

@cindex pipe
A @dfn{pipe} is a mechanism for interprocess communication; data written
to the pipe by one process can be read by another process.  The data is
handled in a first-in, first-out (FIFO) order.  The pipe has no name; it
is created for one use and both ends must be inherited from the single
process which created the pipe.

@cindex FIFO special file
A @dfn{FIFO special file} is similar to a pipe, but instead of being an
anonymous, temporary connection, a FIFO has a name or names like any
other file.  Processes open the FIFO by name in order to communicate
through it.

A pipe or FIFO has to be open at both ends simultaneously.  If you read
from a pipe or FIFO file that doesn't have any processes writing to it
(perhaps because they have all closed the file, or exited), the read
returns end-of-file.  Writing to a pipe or FIFO that doesn't have a
reading process is treated as an error condition; it generates a
@code{SIGPIPE} signal, and fails with error code @code{EPIPE} if the
signal is handled or blocked.

Neither pipes nor FIFO special files allow file positioning.  Both
reading and writing operations happen sequentially; reading from the
beginning of the file and writing at the end.

@menu
* Creating a Pipe::             Making a pipe with the @code{pipe} function.
* Pipe to a Subprocess::        Using a pipe to communicate with a
				 child process.
* FIFO Special Files::          Making a FIFO special file.
* Pipe Atomicity::		When pipe (or FIFO) I/O is atomic.
@end menu

@node Creating a Pipe
@section Creating a Pipe
@cindex creating a pipe
@cindex opening a pipe
@cindex interprocess communication, with pipes

The primitive for creating a pipe is the @code{pipe} function.  This
creates both the reading and writing ends of the pipe.  It is not very
useful for a single process to use a pipe to talk to itself.  In typical
use, a process creates a pipe just before it forks one or more child
processes (@pxref{Creating a Process}).  The pipe is then used for
communication either between the parent or child processes, or between
two sibling processes.

The @code{pipe} function is declared in the header file
@file{unistd.h}.
@pindex unistd.h

@comment unistd.h
@comment POSIX.1
@deftypefun int pipe (int @var{filedes}@t{[2]})
@safety{@prelim{}@mtsafe{}@assafe{}@acsafe{@acsfd{}}}
@c On Linux, syscall pipe2.  On HURD, call socketpair.
The @code{pipe} function creates a pipe and puts the file descriptors
for the reading and writing ends of the pipe (respectively) into
@code{@var{filedes}[0]} and @code{@var{filedes}[1]}.

An easy way to remember that the input end comes first is that file
descriptor @code{0} is standard input, and file descriptor @code{1} is
standard output.

If successful, @code{pipe} returns a value of @code{0}.  On failure,
@code{-1} is returned.  The following @code{errno} error conditions are
defined for this function:

@table @code
@item EMFILE
The process has too many files open.

@item ENFILE
There are too many open files in the entire system.  @xref{Error Codes},
for more information about @code{ENFILE}.  This error never occurs on
@gnuhurdsystems{}.
@end table
@end deftypefun

Here is an example of a simple program that creates a pipe.  This program
uses the @code{fork} function (@pxref{Creating a Process}) to create
a child process.  The parent process writes data to the pipe, which is
read by the child process.

@smallexample
@include pipe.c.texi
@end smallexample

@node Pipe to a Subprocess
@section Pipe to a Subprocess
@cindex creating a pipe to a subprocess
@cindex pipe to a subprocess
@cindex filtering i/o through subprocess

A common use of pipes is to send data to or receive data from a program
being run as a subprocess.  One way of doing this is by using a combination of
@code{pipe} (to create the pipe), @code{fork} (to create the subprocess),
@code{dup2} (to force the subprocess to use the pipe as its standard input
or output channel), and @code{exec} (to execute the new program).  Or,
you can use @code{popen} and @code{pclose}.

The advantage of using @code{popen} and @code{pclose} is that the
interface is much simpler and easier to use.  But it doesn't offer as
much flexibility as using the low-level functions directly.

@comment stdio.h
@comment POSIX.2, SVID, BSD
@deftypefun {FILE *} popen (const char *@var{command}, const char *@var{mode})
@safety{@prelim{}@mtsafe{}@asunsafe{@ascuheap{} @asucorrupt{}}@acunsafe{@acucorrupt{} @aculock{} @acsfd{} @acsmem{}}}
@c popen @ascuheap @asucorrupt @acucorrupt @aculock @acsfd @acsmem
@c  malloc dup @ascuheap @acsmem
@c  _IO_init ok
@c   _IO_no_init ok
@c    _IO_old_init ok
@c     _IO_lock_init ok
@c  _IO_new_file_init @asucorrupt @acucorrupt @aculock @acsfd
@c   _IO_link_in @asucorrupt @acucorrupt @aculock @acsfd
@c     the linked list is guarded by a recursive lock;
@c     it may get corrupted with async signals and cancellation
@c    _IO_lock_lock dup @aculock
@c    _IO_flockfile dup @aculock
@c    _IO_funlockfile dup @aculock
@c    _IO_lock_unlock dup @aculock
@c  _IO_new_proc_open @asucorrupt @acucorrupt @aculock @acsfd
@c    the linked list is guarded by a recursive lock;
 @c   it may get corrupted with async signals and cancellation
@c   _IO_file_is_open ok
@c   pipe2 dup @acsfd
@c   pipe dup @acsfd
@c   _IO_fork=fork @aculock
@c   _IO_close=close_not_cancel dup @acsfd
@c   fcntl dup ok
@c   _IO_lock_lock @aculock
@c   _IO_lock_unlock @aculock
@c   _IO_mask_flags ok [no @mtasurace:stream, nearly but sufficiently exclusive access]
@c  _IO_un_link @asucorrupt @acucorrupt @aculock @acsfd
@c    the linked list is guarded by a recursive lock;
@c    it may get corrupted with async signals and cancellation
@c   _IO_lock_lock dup @aculock
@c   _IO_flockfile dup @aculock
@c   _IO_funlockfile dup @aculock
@c   _IO_lock_unlock dup @aculock
@c  free dup @ascuheap @acsmem
The @code{popen} function is closely related to the @code{system}
function; see @ref{Running a Command}.  It executes the shell command
@var{command} as a subprocess.  However, instead of waiting for the
command to complete, it creates a pipe to the subprocess and returns a
stream that corresponds to that pipe.

If you specify a @var{mode} argument of @code{"r"}, you can read from the
stream to retrieve data from the standard output channel of the subprocess.
The subprocess inherits its standard input channel from the parent process.

Similarly, if you specify a @var{mode} argument of @code{"w"}, you can
write to the stream to send data to the standard input channel of the
subprocess.  The subprocess inherits its standard output channel from
the parent process.

In the event of an error @code{popen} returns a null pointer.  This
might happen if the pipe or stream cannot be created, if the subprocess
cannot be forked, or if the program cannot be executed.
@end deftypefun

@comment stdio.h
@comment POSIX.2, SVID, BSD
@deftypefun int pclose (FILE *@var{stream})
@safety{@prelim{}@mtsafe{}@asunsafe{@ascuheap{} @ascuplugin{} @asucorrupt{} @asulock{}}@acunsafe{@acucorrupt{} @aculock{} @acsfd{} @acsmem{}}}
@c Although the stream cannot be used after the call, even in case of
@c async cancellation, because the stream must not be used after pclose
@c is called, other stdio linked lists and their locks may be left in
@c corrupt states; that's where the corrupt and lock annotations come
@c from.
@c
@c pclose @ascuheap @ascuplugin @asucorrupt @asulock @acucorrupt @aculock @acsfd @acsmem
@c  _IO_new_fclose @ascuheap @ascuplugin @asucorrupt @asulock @acucorrupt @aculock @acsfd @acsmem
@c   _IO_un_link dup @asucorrupt @acucorrupt @aculock @acsfd
@c   _IO_acquire_lock dup @aculock
@c    _IO_flockfile dup @aculock
@c   _IO_file_close_it @ascuheap @ascuplugin @asucorrupt @aculock @acucorrupt @acsfd @acsmem
@c    _IO_file_is_open dup ok
@c    _IO_do_flush @asucorrupt @ascuplugin @acucorrupt
@c     _IO_do_write @asucorrupt @acucorrupt
@c      new_do_write @asucorrupt @acucorrupt
@c       _IO_SYSSEEK ok
@c        lseek64 dup ok
@c       _IO_SYSWRITE ok
@c        write_not_cancel dup ok
@c        write dup ok
@c       _IO_adjust_column ok
@c       _IO_setg dup @asucorrupt @acucorrupt [no @mtasurace:stream, locked]
@c     _IO_wdo_write @asucorrupt @ascuplugin @acucorrupt
@c      _IO_new_do_write=_IO_do_write dup @asucorrupt @acucorrupt
@c      *cc->__codecvt_do_out @ascuplugin
@c      _IO_wsetg dup @asucorrupt @acucorrupt [no @mtasurace:stream, locked]
@c    _IO_unsave_markers @ascuheap @asucorrupt @acucorrupt @acsmem
@c     _IO_have_backup dup ok
@c     _IO_free_backup_area dup @ascuheap @asucorrupt @acucorrupt @acsmem
@c    _IO_SYSCLOSE @aculock @acucorrupt @acsfd
@c     _IO_lock_lock dup @aculock
@c     _IO_close=close_not_cancel dup @acsfd
@c     _IO_lock_unlock dup @aculock
@c     _IO_waitpid=waitpid_not_cancel dup ok
@c    _IO_have_wbackup ok
@c    _IO_free_wbackup_area @ascuheap @asucorrupt @acucorrupt @acsmem
@c     _IO_in_backup dup ok
@c     _IO_switch_to_main_wget_area @asucorrupt @acucorrupt
@c     free dup @ascuheap @acsmem
@c    _IO_wsetb @asucorrupt @acucorrupt [no @mtasurace:stream, locked]
@c    _IO_wsetg @asucorrupt @acucorrupt [no @mtasurace:stream, locked]
@c    _IO_wsetp @asucorrupt @acucorrupt [no @mtasurace:stream, locked]
@c    _IO_setb @asucorrupt @acucorrupt [no @mtasurace:stream, locked]
@c    _IO_setg @asucorrupt @acucorrupt [no @mtasurace:stream, locked]
@c    _IO_setp @asucorrupt @acucorrupt [no @mtasurace:stream, locked]
@c    _IO_un_link dup @asucorrupt @acucorrupt @aculock @acsfd
@c   _IO_release_lock dup @aculock
@c    _IO_funlockfile dup @aculock
@c   _IO_FINISH @ascuheap @ascuplugin @asucorrupt @acucorrupt @aculock @acsfd @acsmem
@c    _IO_new_file_finish @ascuheap @ascuplugin @asucorrupt @acucorrupt @aculock @acsfd @acsmem
@c     _IO_file_is_open dup ok
@c     _IO_do_flush dup @ascuplugin @asucorrupt @acucorrupt
@c     _IO_SYSCLOSE dup @aculock @acucorrupt @acsfd
@c     _IO_default_finish @ascuheap @asucorrupt @acucorrupt @aculock @acsfd @acsmem
@c      FREE_BUF @acsmem
@c       munmap dup @acsmem
@c      free dup @ascuheap @acsmem
@c      _IO_un_link dup @asucorrupt @acucorrupt @aculock @acsfd
@c      _IO_lock_fini ok
@c       libc_lock_fini_recursive ok
@c   libc_lock_lock dup @asulock @aculock
@c   gconv_release_step ok
@c   libc_lock_unlock dup @asulock @aculock
@c   _IO_have_backup ok
@c   _IO_free_backup_area @ascuheap @asucorrupt @acucorrupt @acsmem
@c    _IO_in_backup ok
@c    _IO_switch_to_main_get_area @asucorrupt @acucorrupt
@c    free dup @ascuheap @acsmem
@c   free dup @ascuheap @acsmem
The @code{pclose} function is used to close a stream created by @code{popen}.
It waits for the child process to terminate and returns its status value,
as for the @code{system} function.
@end deftypefun

Here is an example showing how to use @code{popen} and @code{pclose} to
filter output through another program, in this case the paging program
@code{more}.

@smallexample
@include popen.c.texi
@end smallexample

@node FIFO Special Files
@section FIFO Special Files
@cindex creating a FIFO special file
@cindex interprocess communication, with FIFO

A FIFO special file is similar to a pipe, except that it is created in a
different way.  Instead of being an anonymous communications channel, a
FIFO special file is entered into the file system by calling
@code{mkfifo}.

Once you have created a FIFO special file in this way, any process can
open it for reading or writing, in the same way as an ordinary file.
However, it has to be open at both ends simultaneously before you can
proceed to do any input or output operations on it.  Opening a FIFO for
reading normally blocks until some other process opens the same FIFO for
writing, and vice versa.

The @code{mkfifo} function is declared in the header file
@file{sys/stat.h}.
@pindex sys/stat.h

@comment sys/stat.h
@comment POSIX.1
@deftypefun int mkfifo (const char *@var{filename}, mode_t @var{mode})
@safety{@prelim{}@mtsafe{}@assafe{}@acsafe{}}
@c On generic Posix, calls xmknod.
The @code{mkfifo} function makes a FIFO special file with name
@var{filename}.  The @var{mode} argument is used to set the file's
permissions; see @ref{Setting Permissions}.

The normal, successful return value from @code{mkfifo} is @code{0}.  In
the case of an error, @code{-1} is returned.  In addition to the usual
file name errors (@pxref{File Name Errors}), the following
@code{errno} error conditions are defined for this function:

@table @code
@item EEXIST
The named file already exists.

@item ENOSPC
The directory or file system cannot be extended.

@item EROFS
The directory that would contain the file resides on a read-only file
system.
@end table
@end deftypefun

@node Pipe Atomicity
@section Atomicity of Pipe I/O

Reading or writing pipe data is @dfn{atomic} if the size of data written
is not greater than @code{PIPE_BUF}.  This means that the data transfer
seems to be an instantaneous unit, in that nothing else in the system
can observe a state in which it is partially complete.  Atomic I/O may
not begin right away (it may need to wait for buffer space or for data),
but once it does begin it finishes immediately.

Reading or writing a larger amount of data may not be atomic; for
example, output data from other processes sharing the descriptor may be
interspersed.  Also, once @code{PIPE_BUF} characters have been written,
further writes will block until some characters are read.

@xref{Limits for Files}, for information about the @code{PIPE_BUF}
parameter.