path: root/man2/readv.2

.\" Copyright (C) 2007, 2010 Michael Kerrisk <mtk.manpages@gmail.com>
.\" and Copyright (c) 1993 by Thomas Koenig (ig25@rz.uni-karlsruhe.de)
.\"
.\" %%%LICENSE_START(VERBATIM)
.\" Permission is granted to make and distribute verbatim copies of this
.\" manual provided the copyright notice and this permission notice are
.\" preserved on all copies.
.\"
.\" Permission is granted to copy and distribute modified versions of this
.\" manual under the conditions for verbatim copying, provided that the
.\" entire resulting derived work is distributed under the terms of a
.\" permission notice identical to this one.
.\"
.\" Since the Linux kernel and libraries are constantly changing, this
.\" manual page may be incorrect or out-of-date.  The author(s) assume no
.\" responsibility for errors or omissions, or for damages resulting from
.\" the use of the information contained herein.  The author(s) may not
.\" have taken the same level of care in the production of this manual,
.\" which is licensed free of charge, as they might when working
.\" professionally.
.\"
.\" Formatted or processed versions of this manual, if unaccompanied by
.\" the source, must acknowledge the copyright and authors of this work.
.\" %%%LICENSE_END
.\"
.\" Modified Sat Jul 24 18:34:44 1993 by Rik Faith (faith@cs.unc.edu)
.\" Merged readv.[23], 2002-10-17, aeb
.\" 2007-04-30 mtk, A fairly major rewrite to fix errors and
.\"     add more details.
.\" 2010-11-16, mtk, Added documentation of preadv() and pwritev()
.\"
.TH READV 2  2020-12-21 "Linux" "Linux Programmer's Manual"
.SH NAME
readv, writev, preadv, pwritev, preadv2, pwritev2 \- read or write data into multiple buffers
.SH SYNOPSIS
.nf
.B #include <sys/uio.h>
.PP
.BI "ssize_t readv(int " fd ", const struct iovec *" iov ", int " iovcnt );
.BI "ssize_t writev(int " fd ", const struct iovec *" iov ", int " iovcnt );
.PP
.BI "ssize_t preadv(int " fd ", const struct iovec *" iov ", int " iovcnt ,
.BI "                off_t " offset );
.BI "ssize_t pwritev(int " fd ", const struct iovec *" iov ", int " iovcnt ,
.BI "                off_t " offset );
.PP
.BI "ssize_t preadv2(int " fd ", const struct iovec *" iov ", int " iovcnt ,
.BI "                off_t " offset ", int " flags );
.BI "ssize_t pwritev2(int " fd ", const struct iovec *" iov ", int " iovcnt ,
.BI "                off_t " offset ", int " flags );
.fi
.PP
.RS -4
Feature Test Macro Requirements for glibc (see
.BR feature_test_macros (7)):
.RE
.PP
.BR preadv (),
.BR pwritev ():
.nf
    Since glibc 2.19:
        _DEFAULT_SOURCE
    Glibc 2.19 and earlier:
        _BSD_SOURCE
.fi
.SH DESCRIPTION
The
.BR readv ()
system call reads
.I iovcnt
buffers from the file associated with the file descriptor
.I fd
into the buffers described by
.I iov
("scatter input").
.PP
The
.BR writev ()
system call writes
.I iovcnt
buffers of data described by
.I iov
to the file associated with the file descriptor
.I fd
("gather output").
.PP
The pointer
.I iov
points to an array of
.I iovec
structures,
defined in
.I <sys/uio.h>
as:
.PP
.in +4n
.EX
struct iovec {
    void  *iov_base;    /* Starting address */
    size_t iov_len;     /* Number of bytes to transfer */
};
.EE
.in
.PP
The
.BR readv ()
system call works just like
.BR read (2)
except that multiple buffers are filled.
.PP
The
.BR writev ()
system call works just like
.BR write (2)
except that multiple buffers are written out.
.PP
Buffers are processed in array order.
This means that
.BR readv ()
completely fills
.I iov[0]
before proceeding to
.IR iov[1] ,
and so on.
(If there is insufficient data, then not all buffers pointed to by
.I iov
may be filled.)
Similarly,
.BR writev ()
writes out the entire contents of
.I iov[0]
before proceeding to
.IR iov[1] ,
and so on.
.PP
The data transfers performed by
.BR readv ()
and
.BR writev ()
are atomic: the data written by
.\" Regarding atomicity, see https://bugzilla.kernel.org/show_bug.cgi?id=10596
.BR writev ()
is written as a single block that is not intermingled with output
from writes in other processes (but see
.BR pipe (7)
for an exception);
analogously,
.BR readv ()
is guaranteed to read a contiguous block of data from the file,
regardless of read operations performed in other threads or processes
that have file descriptors referring to the same open file description
(see
.BR open (2)).
.SS preadv() and pwritev()
The
.BR preadv ()
system call combines the functionality of
.BR readv ()
and
.BR pread (2).
It performs the same task as
.BR readv (),
but adds a fourth argument,
.IR offset ,
which specifies the file offset at which the input operation
is to be performed.
.PP
The
.BR pwritev ()
system call combines the functionality of
.BR writev ()
and
.BR pwrite (2).
It performs the same task as
.BR writev (),
but adds a fourth argument,
.IR offset ,
which specifies the file offset at which the output operation
is to be performed.
.PP
The file offset is not changed by these system calls.
The file referred to by
.I fd
must be capable of seeking.
.SS preadv2() and pwritev2()
These system calls are similar to
.BR preadv ()
and
.BR pwritev ()
calls, but add a fifth argument,
.IR flags ,
which modifies the behavior on a per-call basis.
.PP
Unlike
.BR preadv ()
and
.BR pwritev (),
if the
.I offset
argument is \-1, then the current file offset is used and updated.
.PP
The
.I flags
argument contains a bitwise OR of zero or more of the following flags:
.TP
.BR RWF_DSYNC " (since Linux 4.7)"
.\" commit e864f39569f4092c2b2bc72c773b6e486c7e3bd9
Provide a per-write equivalent of the
.B O_DSYNC
.BR open (2)
flag.
This flag is meaningful only for
.BR pwritev2 (),
and its effect applies only to the data range written by the system call.
.TP
.BR RWF_HIPRI " (since Linux 4.6)"
High priority read/write.
Allows block-based filesystems to use polling of the device,
which provides lower latency, but may use additional resources.
(Currently, this feature is usable only on a file descriptor opened using the
.BR O_DIRECT
flag.)
.TP
.BR RWF_SYNC " (since Linux 4.7)"
.\" commit e864f39569f4092c2b2bc72c773b6e486c7e3bd9
Provide a per-write equivalent of the
.B O_SYNC
.BR open (2)
flag.
This flag is meaningful only for
.BR pwritev2 (),
and its effect applies only to the data range written by the system call.
.TP
.BR RWF_NOWAIT " (since Linux 4.14)"
.\" commit 3239d834847627b6634a4139cf1dc58f6f137a46
.\" commit 91f9943e1c7b6638f27312d03fe71fcc67b23571
Do not wait for data which is not immediately available.
If this flag is specified, the
.BR preadv2 ()
system call will return instantly if it would have to read data from
the backing storage or wait for a lock.
If some data was successfully read, it will return the number of bytes read.
If no bytes were read, it will return -1 and set
.IR errno
to
.BR EAGAIN .
Currently, this flag is meaningful only for
.BR preadv2 ().
.TP
.BR RWF_APPEND " (since Linux 4.16)"
.\" commit e1fc742e14e01d84d9693c4aca4ab23da65811fb
Provide a per-write equivalent of the
.B O_APPEND
.BR open (2)
flag.
This flag is meaningful only for
.BR pwritev2 (),
and its effect applies only to the data range written by the system call.
The
.I offset
argument does not affect the write operation;
the data is always appended to the end of the file.
However, if the
.I offset
argument is \-1, the current file offset is updated.
.SH RETURN VALUE
On success,
.BR readv (),
.BR preadv (),
and
.BR preadv2 ()
return the number of bytes read;
.BR writev (),
.BR pwritev (),
and
.BR pwritev2 ()
return the number of bytes written.
.PP
Note that it is not an error for a successful call to transfer fewer bytes
than requested (see
.BR read (2)
and
.BR write (2)).
.PP
On error, \-1 is returned, and \fIerrno\fP is set appropriately.
.SH ERRORS
The errors are as given for
.BR read (2)
and
.BR write (2).
Furthermore,
.BR preadv (),
.BR preadv2 (),
.BR pwritev (),
and
.BR pwritev2 ()
can also fail for the same reasons as
.BR lseek (2).
Additionally, the following errors are defined:
.TP
.B EINVAL
The sum of the
.I iov_len
values overflows an
.I ssize_t
value.
.TP
.B EINVAL
The vector count,
.IR iovcnt ,
is less than zero or greater than the permitted maximum.
.TP
.B EOPNOTSUPP
An unknown flag is specified in \fIflags\fP.
.SH VERSIONS
.BR preadv ()
and
.BR pwritev ()
first appeared in Linux 2.6.30; library support was added in glibc 2.10.
.PP
.BR preadv2 ()
and
.BR pwritev2 ()
first appeared in Linux 4.6.
Library support was added in glibc 2.26.
.SH CONFORMING TO
.BR readv (),
.BR writev ():
POSIX.1-2001, POSIX.1-2008,
4.4BSD (these system calls first appeared in 4.2BSD).
.\" Linux libc5 used \fIsize_t\fP as the type of the \fIiovcnt\fP argument,
.\" and \fIint\fP as the return type.
.\" The readv/writev system calls were buggy before Linux 1.3.40.
.\" (Says release.libc.)
.PP
.BR preadv (),
.BR pwritev ():
nonstandard, but present also on the modern BSDs.
.PP
.BR preadv2 (),
.BR pwritev2 ():
nonstandard Linux extension.
.SH NOTES
POSIX.1 allows an implementation to place a limit on
the number of items that can be passed in
.IR iov .
An implementation can advertise its limit by defining
.B IOV_MAX
in
.I <limits.h>
or at run time via the return value from
.IR sysconf(_SC_IOV_MAX) .
On modern Linux systems, the limit is 1024.
Back in Linux 2.0 days, this limit was 16.
.\"
.\"
.SS C library/kernel differences
The raw
.BR preadv ()
and
.BR pwritev ()
system calls have call signatures that differ slightly from that of the
corresponding GNU C library wrapper functions shown in the SYNOPSIS.
The final argument,
.IR offset ,
is unpacked by the wrapper functions into two arguments in the system calls:
.PP
.BI "    unsigned long " pos_l ", unsigned long " pos
.PP
These arguments contain, respectively, the low order and high order 32 bits of
.IR offset .
.SS Historical C library/kernel differences
To deal with the fact that
.B IOV_MAX
was so low on early versions of Linux,
the glibc wrapper functions for
.BR readv ()
and
.BR writev ()
did some extra work if they detected that the underlying kernel
system call failed because this limit was exceeded.
In the case of
.BR readv (),
the wrapper function allocated a temporary buffer large enough
for all of the items specified by
.IR iov ,
passed that buffer in a call to
.BR read (2),
copied data from the buffer to the locations specified by the
.I iov_base
fields of the elements of
.IR iov ,
and then freed the buffer.
The wrapper function for
.BR writev ()
performed the analogous task using a temporary buffer and a call to
.BR write (2).
.PP
The need for this extra effort in the glibc wrapper functions
went away with Linux 2.2 and later.
However, glibc continued to provide this behavior until version 2.10.
Starting with glibc version 2.9,
the wrapper functions provide this behavior only if the library detects
that the system is running a Linux kernel older than version 2.6.18
(an arbitrarily selected kernel version).
And since glibc 2.20
(which requires a minimum Linux kernel version of 2.6.32),
the glibc wrapper functions always just directly invoke the system calls.
.SH EXAMPLES
The following code sample demonstrates the use of
.BR writev ():
.PP
.in +4n
.EX
char *str0 = "hello ";
char *str1 = "world\en";
struct iovec iov[2];
ssize_t nwritten;

iov[0].iov_base = str0;
iov[0].iov_len = strlen(str0);
iov[1].iov_base = str1;
iov[1].iov_len = strlen(str1);

nwritten = writev(STDOUT_FILENO, iov, 2);
.EE
.in
.SH SEE ALSO
.BR pread (2),
.BR read (2),
.BR write (2)