1AIO_FSYNC(3P) POSIX Programmer's Manual AIO_FSYNC(3P)
2
3
4
6 This manual page is part of the POSIX Programmer's Manual. The Linux
7 implementation of this interface may differ (consult the corresponding
8 Linux manual page for details of Linux behavior), or the interface may
9 not be implemented on Linux.
10
12 aio_fsync — asynchronous file synchronization
13
15 #include <aio.h>
16
17 int aio_fsync(int op, struct aiocb *aiocbp);
18
20 The aio_fsync() function shall asynchronously perform a file synchro‐
21 nization operation, as specified by the op argument, for I/O operations
22 associated with the file indicated by the file descriptor aio_fildes
23 member of the aiocb structure referenced by the aiocbp argument and
24 queued at the time of the call to aio_fsync(). The function call shall
25 return when the synchronization request has been initiated or queued to
26 the file or device (even when the data cannot be synchronized immedi‐
27 ately).
28
29 If op is O_DSYNC, all currently queued I/O operations shall be com‐
30 pleted as if by a call to fdatasync(); that is, as defined for synchro‐
31 nized I/O data integrity completion.
32
33 If op is O_SYNC, all currently queued I/O operations shall be completed
34 as if by a call to fsync(); that is, as defined for synchronized I/O
35 file integrity completion. If the aio_fsync() function fails, or if
36 the operation queued by aio_fsync() fails, then outstanding I/O opera‐
37 tions are not guaranteed to have been completed.
38
39 If aio_fsync() succeeds, then it is only the I/O that was queued at the
40 time of the call to aio_fsync() that is guaranteed to be forced to the
41 relevant completion state. The completion of subsequent I/O on the file
42 descriptor is not guaranteed to be completed in a synchronized fashion.
43
44 The aiocbp argument refers to an asynchronous I/O control block. The
45 aiocbp value may be used as an argument to aio_error() and aio_return()
46 in order to determine the error status and return status, respectively,
47 of the asynchronous operation while it is proceeding. When the request
48 is queued, the error status for the operation is [EINPROGRESS]. When
49 all data has been successfully transferred, the error status shall be
50 reset to reflect the success or failure of the operation. If the opera‐
51 tion does not complete successfully, the error status for the operation
52 shall be set to indicate the error. The aio_sigevent member determines
53 the asynchronous notification to occur as specified in Section 2.4.1,
54 Signal Generation and Delivery when all operations have achieved syn‐
55 chronized I/O completion. All other members of the structure referenced
56 by aiocbp are ignored. If the control block referenced by aiocbp
57 becomes an illegal address prior to asynchronous I/O completion, then
58 the behavior is undefined.
59
60 If the aio_fsync() function fails or aiocbp indicates an error condi‐
61 tion, data is not guaranteed to have been successfully transferred.
62
64 The aio_fsync() function shall return the value 0 if the I/O operation
65 is successfully queued; otherwise, the function shall return the value
66 -1 and set errno to indicate the error.
67
69 The aio_fsync() function shall fail if:
70
71 EAGAIN The requested asynchronous operation was not queued due to tem‐
72 porary resource limitations.
73
74 EBADF The aio_fildes member of the aiocb structure referenced by the
75 aiocbp argument is not a valid file descriptor.
76
77 EINVAL This implementation does not support synchronized I/O for this
78 file.
79
80 EINVAL The aio_fildes member of the aiocb structure refers to a file on
81 which an fsync() operation is not possible.
82
83 EINVAL A value of op other than O_DSYNC or O_SYNC was specified, or
84 O_DSYNC was specified and the implementation does not provide
85 runtime support for the Synchronized Input and Output option, or
86 O_SYNC was specified and the implementation does not provide
87 runtime support for the File Synchronization option.
88
89 In the event that any of the queued I/O operations fail, aio_fsync()
90 shall return the error condition defined for read() and write(). The
91 error is returned in the error status for the asynchronous operation,
92 which can be retrieved using aio_error().
93
94 The following sections are informative.
95
97 None.
98
100 Note that even if the file descriptor is not open for writing, if there
101 are any pending write requests on the underlying file, then that I/O
102 will be completed prior to the return of a call to aio_error() or
103 aio_return() indicating that the operation has completed.
104
106 None.
107
109 None.
110
112 aio_error(), aio_return(), fcntl(), fdatasync(), fsync(), open(),
113 read(), write()
114
115 The Base Definitions volume of POSIX.1‐2017, <aio.h>
116
118 Portions of this text are reprinted and reproduced in electronic form
119 from IEEE Std 1003.1-2017, Standard for Information Technology -- Por‐
120 table Operating System Interface (POSIX), The Open Group Base Specifi‐
121 cations Issue 7, 2018 Edition, Copyright (C) 2018 by the Institute of
122 Electrical and Electronics Engineers, Inc and The Open Group. In the
123 event of any discrepancy between this version and the original IEEE and
124 The Open Group Standard, the original IEEE and The Open Group Standard
125 is the referee document. The original Standard can be obtained online
126 at http://www.opengroup.org/unix/online.html .
127
128 Any typographical or formatting errors that appear in this page are
129 most likely to have been introduced during the conversion of the source
130 files to man page format. To report such errors, see https://www.ker‐
131 nel.org/doc/man-pages/reporting_bugs.html .
132
133
134
135IEEE/The Open Group 2017 AIO_FSYNC(3P)