freebsd-skq/lib/libc/sys/aio_fsync.2
John Baldwin 4092a25cd0 Fix various nits in the aio operation manpages.
- Avoid double use of "request" in a single sentence.  Instead, describe
  aio_sigevent as being used to request notification of the associated
  operation's completion.  This matches the language used to describe
  aio_sigevent in aio(4).
- Simplify the prohibition on modifying buffers while requests are in
  flight.
- Fix case mismatch.
- Drop note about not using stack variables. C programmers should be able
  to figure out if a stack variable is safe based on the later warning
  about the life cycle requirements of control blocks.
- Remove prohibition on modifying the I/O buffer for aio_fsync() since
  it does not use an I/O buffer.  For aio_mlock(), prohibit modifications
  to the mapping (e.g. due to mprotect, munmap, mmap, etc.) but do not
  prohibit modifications to the memory backing the buffer (stores into
  the pages backing the buffer).

Requested by:	wblock (1,2), kib (4)
Reviewed by:	kib, rpokala, wblock
MFC after:	3 days
Sponsored by:	Chelsio Communications
Differential Revision:	https://reviews.freebsd.org/D7462
2016-08-19 17:37:32 +00:00

179 lines
4.9 KiB
Groff

.\" Copyright (c) 2013 Sergey Kandaurov
.\" All rights reserved.
.\"
.\" Redistribution and use in source and binary forms, with or without
.\" modification, are permitted provided that the following conditions
.\" are met:
.\" 1. Redistributions of source code must retain the above copyright
.\" notice, this list of conditions and the following disclaimer.
.\" 2. Redistributions in binary form must reproduce the above copyright
.\" notice, this list of conditions and the following disclaimer in the
.\" documentation and/or other materials provided with the distribution.
.\"
.\" THIS SOFTWARE IS PROVIDED BY THE AUTHOR AND CONTRIBUTORS ``AS IS'' AND
.\" ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE
.\" IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE
.\" ARE DISCLAIMED. IN NO EVENT SHALL THE AUTHOR OR CONTRIBUTORS BE LIABLE
.\" FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL
.\" DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS
.\" OR SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION)
.\" HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT
.\" LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY
.\" OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF
.\" SUCH DAMAGE.
.\"
.\" $FreeBSD$
.\"
.Dd August 19, 2016
.Dt AIO_FSYNC 2
.Os
.Sh NAME
.Nm aio_fsync
.Nd asynchronous file synchronization (REALTIME)
.Sh LIBRARY
.Lb libc
.Sh SYNOPSIS
.In aio.h
.Ft int
.Fn aio_fsync "int op" "struct aiocb *iocb"
.Sh DESCRIPTION
The
.Fn aio_fsync
system call allows the calling process to move all modified data
associated with the descriptor
.Fa iocb->aio_fildes
to a permanent storage device.
The call returns immediately after the synchronization request has been
enqueued to the descriptor; the synchronization may or may not have
completed at the time the call returns.
.Pp
The
.Fa op
argument can only be set to
.Dv O_SYNC
to cause all currently queued I/O operations to be completed
as if by a call to
.Xr fsync 2 .
.Pp
If _POSIX_PRIORITIZED_IO is defined, and the descriptor supports it,
then the enqueued operation is submitted at a priority equal to that
of the calling process minus
.Fa iocb->aio_reqprio .
.Pp
The
.Fa iocb
pointer may be subsequently used as an argument to
.Fn aio_return
and
.Fn aio_error
in order to determine return or error status for the enqueued operation
while it is in progress.
.Pp
If the request could not be enqueued (generally due to invalid arguments),
the call returns without having enqueued the request.
.Pp
The
.Fa iocb->aio_sigevent
structure can be used to request notification of the operation's
completion as described in
.Xr aio 4 .
.Sh RESTRICTIONS
The Asynchronous I/O Control Block structure pointed to by
.Fa iocb
must remain valid until the
operation has completed.
.Pp
The asynchronous I/O control buffer
.Fa iocb
should be zeroed before the
.Fn aio_fsync
call to avoid passing bogus context information to the kernel.
.Pp
Modification of the Asynchronous I/O Control Block structure is not allowed
while the request is queued.
.Sh RETURN VALUES
.Rv -std aio_fsync
.Sh ERRORS
The
.Fn aio_fsync
system call will fail if:
.Bl -tag -width Er
.It Bq Er EAGAIN
The request was not queued because of system resource limitations.
.It Bq Er EINVAL
The asynchronous notification method in
.Fa iocb->aio_sigevent.sigev_notify
is invalid or not supported.
.It Bq Er EOPNOTSUPP
Asynchronous file synchronization operations on the file descriptor
.Fa iocb->aio_fildes
are unsafe and unsafe asynchronous I/O operations are disabled.
.It Bq Er EINVAL
A value of the
.Fa op
argument is not set to
.Dv O_SYNC .
.El
.Pp
The following conditions may be synchronously detected when the
.Fn aio_fsync
system call is made, or asynchronously, at any time thereafter.
If they are detected at call time,
.Fn aio_fsync
returns -1 and sets
.Va errno
appropriately; otherwise the
.Fn aio_return
system call must be called, and will return -1, and
.Fn aio_error
must be called to determine the actual value that would have been
returned in
.Va errno .
.Bl -tag -width Er
.It Bq Er EBADF
The
.Fa iocb->aio_fildes
argument
is not a valid descriptor.
.It Bq Er EINVAL
This implementation does not support synchronized I/O for this file.
.El
.Pp
If the request is successfully enqueued, but subsequently cancelled
or an error occurs, the value returned by the
.Fn aio_return
system call is per the
.Xr read 2
and
.Xr write 2
system calls, and the value returned by the
.Fn aio_error
system call is one of the error returns from the
.Xr read 2
or
.Xr write 2
system calls.
.Sh SEE ALSO
.Xr aio_cancel 2 ,
.Xr aio_error 2 ,
.Xr aio_read 2 ,
.Xr aio_return 2 ,
.Xr aio_suspend 2 ,
.Xr aio_waitcomplete 2 ,
.Xr aio_write 2 ,
.Xr fsync 2 ,
.Xr sigevent 3 ,
.Xr siginfo 3 ,
.Xr aio 4
.Sh STANDARDS
The
.Fn aio_fsync
system call is expected to conform to the
.St -p1003.1
standard.
.Sh HISTORY
The
.Fn aio_fsync
system call first appeared in
.Fx 7.0 .