AIO_INIT(3) Linux Programmer's Manual AIO_INIT(3)NAME
aio_init - asynchronous I/O initialization
SYNOPSIS
#define _GNU_SOURCE /* See feature_test_macros(7) */
#include <aio.h>
void aio_init(const struct aioinit *init);
Link with -lrt.
DESCRIPTION
The GNU-specific aio_init() function allows the caller to provide tuning hints to the glibc POSIX AIO implementation. Use of this function
is optional, but to be effective, it must be called before employing any other functions in the POSIX AIO API.
The tuning information is provided in the buffer pointed to by the argument init. This buffer is a structure of the following form:
struct aioinit {
int aio_threads; /* Maximum number of threads */
int aio_num; /* Number of expected simultaneous
requests */
int aio_locks; /* Not used */
int aio_usedba; /* Not used */
int aio_debug; /* Not used */
int aio_numusers; /* Not used */
int aio_idle_time; /* Number of seconds before idle thread
terminates (since glibc 2.2) */
int aio_reserved;
};
The following fields are used in the aioinit structure:
aio_threads This field specifies the maximum number of worker threads that may be used by the implementation. If the number of out-
standing I/O operations exceeds this limit, then excess operations will be queued until a worker thread becomes free. If
this field is specified with a value less than 1, the value 1 is used. The default value is 20.
aio_num This field should specify the maximum number of simultaneous I/O requests that the caller expects to enqueue. If a value
less than 32 is specified for this field, it is rounded up to 32. The default value is 64.
aio_idle_time This field specifies the amount of time in seconds that a worker thread should wait for further requests before terminating,
after having completed a previous request. The default value is 1.
VERSIONS
The aio_init() function is available since glibc 2.1.
CONFORMING TO
This function is a GNU extension.
SEE ALSO aio(7)Linux 2012-04-26 AIO_INIT(3)
Check Out this Related Man Page
AIO_FSYNC(3) Linux Programmer's Manual AIO_FSYNC(3)NAME
aio_fsync - asynchronous file synchronization
SYNOPSIS
#include <aio.h>
int aio_fsync(int op, struct aiocb *aiocbp);
Link with -lrt.
DESCRIPTION
The aio_fsync() function does a sync on all outstanding asynchronous I/O operations associated with aiocbp->aio_fildes. (See aio(7) for a
description of the aiocb structure.)
More precisely, if op is O_SYNC, then all currently queued I/O operations shall be completed as if by a call of fsync(2), and if op is
O_DSYNC, this call is the asynchronous analog of fdatasync(2).
Note that this is a request only; it does not wait for I/O completion.
Apart from aio_fildes, the only field in the structure pointed to by aiocbp that is used by this call is the aio_sigevent field (a sigevent
structure, described in sigevent(7)), which indicates the desired type of asynchronous notification at completion. All other fields are
ignored.
RETURN VALUE
On success (the sync request was successfully queued) this function returns 0. On error -1 is returned, and errno is set appropriately.
ERRORS
EAGAIN Out of resources.
EBADF aio_fildes is not a valid file descriptor open for writing.
EINVAL Synchronized I/O is not supported for this file, or op is not O_SYNC or O_DSYNC.
ENOSYS aio_fsync() is not implemented.
VERSIONS
The aio_fsync() function is available since glibc 2.1.
CONFORMING TO
POSIX.1-2001, POSIX.1-2008.
SEE ALSO aio_cancel(3), aio_error(3), aio_read(3), aio_return(3), aio_suspend(3), aio_write(3), lio_listio(3), aio(7), sigevent(7)COLOPHON
This page is part of release 3.44 of the Linux man-pages project. A description of the project, and information about reporting bugs, can
be found at http://www.kernel.org/doc/man-pages/.
2012-05-08 AIO_FSYNC(3)