1 @node Pipes and FIFOs, Sockets, File System Interface, Top
2 @chapter Pipes and FIFOs
5 A @dfn{pipe} is a mechanism for interprocess communication; data written
6 to the pipe by one process can be read by another process. The data is
7 handled in a first-in, first-out (FIFO) order. The pipe has no name; it
8 is created for one use and both ends must be inherited from the single
9 process which created the pipe.
11 @cindex FIFO special file
12 A @dfn{FIFO special file} is similar to a pipe, but instead of being an
13 anonymous, temporary connection, a FIFO has a name or names like any
14 other file. Processes open the FIFO by name in order to communicate
17 A pipe or FIFO has to be open at both ends simultaneously. If you read
18 from a pipe or FIFO file that doesn't have any processes writing to it
19 (perhaps because they have all closed the file, or exited), the read
20 returns end-of-file. Writing to a pipe or FIFO that doesn't have a
21 reading process is treated as an error condition; it generates a
22 @code{SIGPIPE} signal, and fails with error code @code{EPIPE} if the
23 signal is handled or blocked.
25 Neither pipes nor FIFO special files allow file positioning. Both
26 reading and writing operations happen sequentially; reading from the
27 beginning of the file and writing at the end.
30 * Creating a Pipe:: Making a pipe with the @code{pipe} function.
31 * Pipe to a Subprocess:: Using a pipe to communicate with a
33 * FIFO Special Files:: Making a FIFO special file.
34 * Pipe Atomicity:: When pipe (or FIFO) I/O is atomic.
38 @section Creating a Pipe
39 @cindex creating a pipe
40 @cindex opening a pipe
41 @cindex interprocess communication, with pipes
43 The primitive for creating a pipe is the @code{pipe} function. This
44 creates both the reading and writing ends of the pipe. It is not very
45 useful for a single process to use a pipe to talk to itself. In typical
46 use, a process creates a pipe just before it forks one or more child
47 processes (@pxref{Creating a Process}). The pipe is then used for
48 communication either between the parent or child processes, or between
49 two sibling processes.
51 The @code{pipe} function is declared in the header file
57 @deftypefun int pipe (int @var{filedes}@t{[2]})
58 The @code{pipe} function creates a pipe and puts the file descriptors
59 for the reading and writing ends of the pipe (respectively) into
60 @code{@var{filedes}[0]} and @code{@var{filedes}[1]}.
62 An easy way to remember that the input end comes first is that file
63 descriptor @code{0} is standard input, and file descriptor @code{1} is
66 If successful, @code{pipe} returns a value of @code{0}. On failure,
67 @code{-1} is returned. The following @code{errno} error conditions are
68 defined for this function:
72 The process has too many files open.
75 There are too many open files in the entire system. @xref{Error Codes},
76 for more information about @code{ENFILE}. This error never occurs in
81 Here is an example of a simple program that creates a pipe. This program
82 uses the @code{fork} function (@pxref{Creating a Process}) to create
83 a child process. The parent process writes data to the pipe, which is
84 read by the child process.
90 @node Pipe to a Subprocess
91 @section Pipe to a Subprocess
92 @cindex creating a pipe to a subprocess
93 @cindex pipe to a subprocess
94 @cindex filtering i/o through subprocess
96 A common use of pipes is to send data to or receive data from a program
97 being run as subprocess. One way of doing this is by using a combination of
98 @code{pipe} (to create the pipe), @code{fork} (to create the subprocess),
99 @code{dup2} (to force the subprocess to use the pipe as its standard input
100 or output channel), and @code{exec} (to execute the new program). Or,
101 you can use @code{popen} and @code{pclose}.
103 The advantage of using @code{popen} and @code{pclose} is that the
104 interface is much simpler and easier to use. But it doesn't offer as
105 much flexibility as using the low-level functions directly.
108 @comment POSIX.2, SVID, BSD
109 @deftypefun {FILE *} popen (const char *@var{command}, const char *@var{mode})
110 The @code{popen} function is closely related to the @code{system}
111 function; see @ref{Running a Command}. It executes the shell command
112 @var{command} as a subprocess. However, instead of waiting for the
113 command to complete, it creates a pipe to the subprocess and returns a
114 stream that corresponds to that pipe.
116 If you specify a @var{mode} argument of @code{"r"}, you can read from the
117 stream to retrieve data from the standard output channel of the subprocess.
118 The subprocess inherits its standard input channel from the parent process.
120 Similarly, if you specify a @var{mode} argument of @code{"w"}, you can
121 write to the stream to send data to the standard input channel of the
122 subprocess. The subprocess inherits its standard output channel from
125 In the event of an error, @code{popen} returns a null pointer. This
126 might happen if the pipe or stream cannot be created, if the subprocess
127 cannot be forked, or if the program cannot be executed.
131 @comment POSIX.2, SVID, BSD
132 @deftypefun int pclose (FILE *@var{stream})
133 The @code{pclose} function is used to close a stream created by @code{popen}.
134 It waits for the child process to terminate and returns its status value,
135 as for the @code{system} function.
138 Here is an example showing how to use @code{popen} and @code{pclose} to
139 filter output through another program, in this case the paging program
143 @include popen.c.texi
146 @node FIFO Special Files
147 @section FIFO Special Files
148 @cindex creating a FIFO special file
149 @cindex interprocess communication, with FIFO
151 A FIFO special file is similar to a pipe, except that it is created in a
152 different way. Instead of being an anonymous communications channel, a
153 FIFO special file is entered into the file system by calling
156 Once you have created a FIFO special file in this way, any process can
157 open it for reading or writing, in the same way as an ordinary file.
158 However, it has to be open at both ends simultaneously before you can
159 proceed to do any input or output operations on it. Opening a FIFO for
160 reading normally blocks until some other process opens the same FIFO for
161 writing, and vice versa.
163 The @code{mkfifo} function is declared in the header file
169 @deftypefun int mkfifo (const char *@var{filename}, mode_t @var{mode})
170 The @code{mkfifo} function makes a FIFO special file with name
171 @var{filename}. The @var{mode} argument is used to set the file's
172 permissions; see @ref{Setting Permissions}.
174 The normal, successful return value from @code{mkfifo} is @code{0}. In
175 the case of an error, @code{-1} is returned. In addition to the usual
176 file name errors (@pxref{File Name Errors}), the following
177 @code{errno} error conditions are defined for this function:
181 The named file already exists.
184 The directory or file system cannot be extended.
187 The directory that would contain the file resides on a read-only file
193 @section Atomicity of Pipe I/O
195 Reading or writing pipe data is @dfn{atomic} if the size of data written
196 is less than @code{PIPE_BUF}. This means that the data transfer seems
197 to be an instantaneous unit, in that nothing else in the system can
198 observe a state in which it is partially complete. Atomic I/O may not
199 begin right away (it may need to wait for buffer space or for data), but
200 once it does begin, it finishes immediately.
202 Reading or writing a larger amount of data may not be atomic; for
203 example, output data from other processes sharing the descriptor may be
204 interspersed. Also, once @code{PIPE_BUF} characters have been written,
205 further writes will block until some characters are read.
207 @xref{Limits for Files}, for information about the @code{PIPE_BUF}