freebsd-dev/lib/libc/sys/aio_write.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

218 lines
5.9 KiB
Groff

.\" Copyright (c) 1999 Softweyr LLC.
.\" 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 Softweyr LLC 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 Softweyr LLC 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_WRITE 2
.Os
.Sh NAME
.Nm aio_write
.Nd asynchronous write to a file (REALTIME)
.Sh LIBRARY
.Lb libc
.Sh SYNOPSIS
.In aio.h
.Ft int
.Fn aio_write "struct aiocb *iocb"
.Sh DESCRIPTION
The
.Fn aio_write
system call allows the calling process to write
.Fa iocb->aio_nbytes
from the buffer pointed to by
.Fa iocb->aio_buf
to the descriptor
.Fa iocb->aio_fildes .
The call returns immediately after the write request has been enqueued
to the descriptor; the write may or may not have completed at the time
the call returns.
If the request could not be enqueued, generally due
to invalid arguments, the call returns without having enqueued the
request.
.Pp
If
.Dv O_APPEND
is set for
.Fa iocb->aio_fildes ,
.Fn aio_write
operations append to the file in the same order as the calls were
made.
If
.Dv O_APPEND
is not set for the file descriptor, the write operation will occur at
the absolute position from the beginning of the file plus
.Fa iocb->aio_offset .
.Pp
If
.Dv _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 is successfully enqueued, the value of
.Fa iocb->aio_offset
can be modified during the request as context, so this value must not
be referenced after the request is enqueued.
.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
and the buffer that the
.Fa iocb->aio_buf
member of that structure references must remain valid until the
operation has completed.
.Pp
The asynchronous I/O control buffer
.Fa iocb
should be zeroed before the
.Fn aio_write
system call to avoid passing bogus context information to the kernel.
.Pp
Modifications of the Asynchronous I/O Control Block structure or the
buffer contents are not allowed while the request is queued.
.Pp
If the file offset in
.Fa iocb->aio_offset
is past the offset maximum for
.Fa iocb->aio_fildes ,
no I/O will occur.
.Sh RETURN VALUES
.Rv -std aio_write
.Sh ERRORS
The
.Fn aio_write
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 write operations on the file descriptor
.Fa iocb->aio_fildes
are unsafe and unsafe asynchronous I/O operations are disabled.
.El
.Pp
The following conditions may be synchronously detected when the
.Fn aio_write
system call is made, or asynchronously, at any time thereafter.
If they
are detected at call time,
.Fn aio_write
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 invalid, or is not opened for writing.
.It Bq Er EINVAL
The offset
.Fa iocb->aio_offset
is not valid, the priority specified by
.Fa iocb->aio_reqprio
is not a valid priority, or the number of bytes specified by
.Fa iocb->aio_nbytes
is not valid.
.El
.Pp
If the request is successfully enqueued, but subsequently canceled
or an error occurs, the value returned by the
.Fn aio_return
system call is per the
.Xr write 2
system call, and the value returned by the
.Fn aio_error
system call is either one of the error returns from the
.Xr write 2
system call, or one of:
.Bl -tag -width Er
.It Bq Er EBADF
The
.Fa iocb->aio_fildes
argument
is invalid for writing.
.It Bq Er ECANCELED
The request was explicitly canceled via a call to
.Fn aio_cancel .
.It Bq Er EINVAL
The offset
.Fa iocb->aio_offset
would be invalid.
.El
.Sh SEE ALSO
.Xr aio_cancel 2 ,
.Xr aio_error 2 ,
.Xr aio_return 2 ,
.Xr aio_suspend 2 ,
.Xr aio_waitcomplete 2 ,
.Xr sigevent 3 ,
.Xr siginfo 3 ,
.Xr aio 4
.Sh STANDARDS
The
.Fn aio_write
system call
is expected to conform to the
.St -p1003.1
standard.
.Sh HISTORY
The
.Fn aio_write
system call first appeared in
.Fx 3.0 .
.Sh AUTHORS
This manual page was written by
.An Wes Peters Aq Mt wes@softweyr.com .
.Sh BUGS
Invalid information in
.Fa iocb->_aiocb_private
may confuse the kernel.