1close(2) System Calls Manual close(2)
2
6 close - close a file descriptor
7
9 Standard C library (libc, -lc)
10
12 #include <unistd.h>
13 int close(int fd);
15
17 close() closes a file descriptor, so that it no longer refers to any
18 file and may be reused. Any record locks (see fcntl(2)) held on the
19 file it was associated with, and owned by the process, are removed (re‐
20 gardless of the file descriptor that was used to obtain the lock).
21 If fd is the last file descriptor referring to the underlying open file
23 description (see open(2)), the resources associated with the open file
24 description are freed; if the file descriptor was the last reference to
25 a file which has been removed using unlink(2), the file is deleted.
26
28 close() returns zero on success. On error, -1 is returned, and errno
29 is set to indicate the error.
30
32 EBADF fd isn't a valid open file descriptor.
33 EINTR The close() call was interrupted by a signal; see signal(7).
35 EIO An I/O error occurred.
37 ENOSPC, EDQUOT
39 On NFS, these errors are not normally reported against the first
40 write which exceeds the available storage space, but instead
41 against a subsequent write(2), fsync(2), or close().
42 See NOTES for a discussion of why close() should not be retried after
44 an error.
45
47 POSIX.1-2008.
48
50 POSIX.1-2001, SVr4, 4.3BSD.
51
53 A successful close does not guarantee that the data has been success‐
54 fully saved to disk, as the kernel uses the buffer cache to defer
55 writes. Typically, filesystems do not flush buffers when a file is
56 closed. If you need to be sure that the data is physically stored on
57 the underlying disk, use fsync(2). (It will depend on the disk hard‐
58 ware at this point.)
59 The close-on-exec file descriptor flag can be used to ensure that a
61 file descriptor is automatically closed upon a successful execve(2);
62 see fcntl(2) for details.
63 Multithreaded processes and close()
65 It is probably unwise to close file descriptors while they may be in
66 use by system calls in other threads in the same process. Since a file
67 descriptor may be reused, there are some obscure race conditions that
68 may cause unintended side effects.
69 When dealing with sockets, you have to be sure that there is no recv(2)
71 still blocking on it on another thread, otherwise it might block for‐
72 ever, since no more messages will be send via the socket. Be sure to
73 use shutdown(2) to shut down all parts the connection before closing
74 the socket.
75 Furthermore, consider the following scenario where two threads are per‐
77 forming operations on the same file descriptor:
78 (1) One thread is blocked in an I/O system call on the file descrip‐
80 tor. For example, it is trying to write(2) to a pipe that is al‐
81 ready full, or trying to read(2) from a stream socket which cur‐
82 rently has no available data.
83 (2) Another thread closes the file descriptor.
85 The behavior in this situation varies across systems. On some systems,
87 when the file descriptor is closed, the blocking system call returns
88 immediately with an error.
89 On Linux (and possibly some other systems), the behavior is different:
91 the blocking I/O system call holds a reference to the underlying open
92 file description, and this reference keeps the description open until
93 the I/O system call completes. (See open(2) for a discussion of open
94 file descriptions.) Thus, the blocking system call in the first thread
95 may successfully complete after the close() in the second thread.
96 Dealing with error returns from close()
98 A careful programmer will check the return value of close(), since it
99 is quite possible that errors on a previous write(2) operation are re‐
100 ported only on the final close() that releases the open file descrip‐
101 tion. Failing to check the return value when closing a file may lead
102 to silent loss of data. This can especially be observed with NFS and
103 with disk quota.
104 Note, however, that a failure return should be used only for diagnostic
106 purposes (i.e., a warning to the application that there may still be
107 I/O pending or there may have been failed I/O) or remedial purposes
108 (e.g., writing the file once more or creating a backup).
109 Retrying the close() after a failure return is the wrong thing to do,
111 since this may cause a reused file descriptor from another thread to be
112 closed. This can occur because the Linux kernel always releases the
113 file descriptor early in the close operation, freeing it for reuse; the
114 steps that may return an error, such as flushing data to the filesystem
115 or device, occur only later in the close operation.
116 Many other implementations similarly always close the file descriptor
118 (except in the case of EBADF, meaning that the file descriptor was in‐
119 valid) even if they subsequently report an error on return from
120 close(). POSIX.1 is currently silent on this point, but there are
121 plans to mandate this behavior in the next major release of the stan‐
122 dard.
123 A careful programmer who wants to know about I/O errors may precede
125 close() with a call to fsync(2).
126 The EINTR error is a somewhat special case. Regarding the EINTR error,
128 POSIX.1-2008 says:
129 If close() is interrupted by a signal that is to be caught, it
131 shall return -1 with errno set to EINTR and the state of fildes
132 is unspecified.
133 This permits the behavior that occurs on Linux and many other implemen‐
135 tations, where, as with other errors that may be reported by close(),
136 the file descriptor is guaranteed to be closed. However, it also per‐
137 mits another possibility: that the implementation returns an EINTR er‐
138 ror and keeps the file descriptor open. (According to its documenta‐
139 tion, HP-UX's close() does this.) The caller must then once more use
140 close() to close the file descriptor, to avoid file descriptor leaks.
141 This divergence in implementation behaviors provides a difficult hurdle
142 for portable applications, since on many implementations, close() must
143 not be called again after an EINTR error, and on at least one, close()
144 must be called again. There are plans to address this conundrum for
145 the next major release of the POSIX.1 standard.
146
148 close_range(2), fcntl(2), fsync(2), open(2), shutdown(2), unlink(2),
149 fclose(3)
150Linux man-pages 6.04 2023-03-30 close(2)