.\" Modified Thu Mar 3 09:49:35 2005 by Michael Haardt <michael@moria.de>
.\" 2007-03-25, mtk, added various text to DESCRIPTION.
.\"
-.TH RENAME 2 2014-02-21 "Linux" "Linux Programmer's Manual"
+.TH RENAME 2 2019-03-06 "Linux" "Linux Programmer's Manual"
.SH NAME
-rename, renameat \- change the name or location of a file
+rename, renameat, renameat2 \- change the name or location of a file
.SH SYNOPSIS
.nf
.B #include <stdio.h>
-.sp
+.PP
.BI "int rename(const char *" oldpath ", const char *" newpath );
-.sp
+
.BR "#include <fcntl.h> " "/* Definition of AT_* constants */"
.B #include <stdio.h>
-.sp
+.PP
.BI "int renameat(int " olddirfd ", const char *" oldpath ,
.BI " int " newdirfd ", const char *" newpath );
+.PP
+.BI "int renameat2(int " olddirfd ", const char *" oldpath ,
+.BI " int " newdirfd ", const char *" newpath \
+", unsigned int " flags );
.fi
-.sp
+.PP
.in -4n
Feature Test Macro Requirements for glibc (see
.BR feature_test_macros (7)):
.in
-.sp
+.PP
.BR renameat ():
.PD 0
.ad l
.RS 4
.TP 4
Since glibc 2.10:
-_XOPEN_SOURCE\ >=\ 700 || _POSIX_C_SOURCE\ >=\ 200809L
+_POSIX_C_SOURCE\ >=\ 200809L
.TP
Before glibc 2.10:
_ATFILE_SOURCE
.RE
+.PP
+.BR renameat2 ():
+.RS 4
+.TP
+_GNU_SOURCE
+.RE
.ad
.PD
.SH DESCRIPTION
Open file descriptors for
.I oldpath
are also unaffected.
-
+.PP
+Various restrictions determine whether or not the rename operation succeeds:
+see ERRORS below.
+.PP
If
.I newpath
-already exists, it will be atomically replaced (subject to
-a few conditions; see ERRORS below), so that there is
+already exists, it will be atomically replaced, so that there is
no point at which another process attempting to access
.I newpath
will find it missing.
-
+However, there will probably be a window in which both
+.I oldpath
+and
+.I newpath
+refer to the file being renamed.
+.PP
If
.I oldpath
and
are existing hard links referring to the same file, then
.BR rename ()
does nothing, and returns a success status.
-
+.PP
If
.I newpath
exists but the operation fails for some reason,
guarantees to leave an instance of
.I newpath
in place.
-
+.PP
.I oldpath
can specify a directory.
In this case,
.I newpath
must either not exist, or it must specify an empty directory.
-
-However, when overwriting there will probably be a window in which
-both
-.I oldpath
-and
-.I newpath
-refer to the file being renamed.
-
+.PP
If
.I oldpath
refers to a symbolic link, the link is renamed; if
.I newpath
refers to a symbolic link, the link will be overwritten.
-.SS renameat ()
+.SS renameat()
The
.BR renameat ()
system call operates in exactly the same way as
-.BR rename (2),
+.BR rename (),
except for the differences described here.
-
+.PP
If the pathname given in
.I oldpath
is relative, then it is interpreted relative to the directory
.I olddirfd
(rather than relative to the current working directory of
the calling process, as is done by
-.BR rename (2)
+.BR rename ()
for a relative pathname).
-
+.PP
If
.I oldpath
is relative and
.I oldpath
is interpreted relative to the current working
directory of the calling process (like
-.BR rename (2)).
-
+.BR rename ()).
+.PP
If
.I oldpath
is absolute, then
.I olddirfd
is ignored.
-
+.PP
The interpretation of
.I newpath
is as for
except that a relative pathname is interpreted relative
to the directory referred to by the file descriptor
.IR newdirfd .
-
+.PP
See
.BR openat (2)
for an explanation of the need for
.BR renameat ().
+.SS renameat2()
+.BR renameat2 ()
+has an additional
+.I flags
+argument.
+A
+.BR renameat2 ()
+call with a zero
+.I flags
+argument is equivalent to
+.BR renameat ().
+.PP
+The
+.I flags
+argument is a bit mask consisting of zero or more of the following flags:
+.TP
+.B RENAME_EXCHANGE
+Atomically exchange
+.IR oldpath
+and
+.IR newpath .
+Both pathnames must exist
+but may be of different types (e.g., one could be a non-empty directory
+and the other a symbolic link).
+.TP
+.B RENAME_NOREPLACE
+Don't overwrite
+.IR newpath
+of the rename.
+Return an error if
+.IR newpath
+already exists.
+.IP
+.B RENAME_NOREPLACE
+can't be employed together with
+.BR RENAME_EXCHANGE .
+.IP
+.B RENAME_NOREPLACE
+requires support from the underlying filesystem.
+Support for various filesystems was added as follows:
+.RS
+.IP * 3
+ext4 (Linux 3.15);
+.\" ext4: commit 0a7c3937a1f23f8cb5fc77ae01661e9968a51d0c
+.IP *
+btrfs, shmem, and cifs (Linux 3.17);
+.IP *
+xfs (Linux 4.0);
+.\" btrfs: commit 80ace85c915d0f41016f82917218997b72431258
+.\" shmem: commit 3b69ff51d087d265aa4af3a532fc4f20bf33e718
+.\" cifs: commit 7c33d5972ce382bcc506d16235f1e9b7d22cbef8
+.\"
+.\" gfs2 in 4.2?
+.IP *
+Support for many other filesystems was added in Linux 4.9, including
+ext2, minix, reiserfs, jfs, vfat, and bpf.
+.\" Also affs, bfs, exofs, hfs, hfsplus, jffs2, logfs, msdos,
+.\" nilfs2, omfs, sysvfs, ubifs, udf, ufs
+.\" hugetlbfs, ramfs
+.\" local filesystems: commit f03b8ad8d38634d13e802165cc15917481b47835
+.\" libfs: commit e0e0be8a835520e2f7c89f214dfda570922a1b90
+.RE
+.TP
+.BR RENAME_WHITEOUT " (since Linux 3.18)"
+.\" commit 0d7a855526dd672e114aff2ac22b60fc6f155b08
+.\" commit 787fb6bc9682ec7c05fb5d9561b57100fbc1cc41
+This operation makes sense only for overlay/union
+filesystem implementations.
+.IP
+Specifying
+.B RENAME_WHITEOUT
+creates a "whiteout" object at the source of
+the rename at the same time as performing the rename.
+The whole operation is atomic,
+so that if the rename succeeds then the whiteout will also have been created.
+.IP
+A "whiteout" is an object that has special meaning in union/overlay
+filesystem constructs.
+In these constructs,
+multiple layers exist and only the top one is ever modified.
+A whiteout on an upper layer will effectively hide a
+matching file in the lower layer,
+making it appear as if the file didn't exist.
+.IP
+When a file that exists on the lower layer is renamed,
+the file is first copied up (if not already on the upper layer)
+and then renamed on the upper, read-write layer.
+At the same time, the source file needs to be "whiteouted"
+(so that the version of the source file in the lower layer
+is rendered invisible).
+The whole operation needs to be done atomically.
+.IP
+When not part of a union/overlay,
+the whiteout appears as a character device with a {0,0} device number.
+.\" https://www.freebsd.org/cgi/man.cgi?query=mount_unionfs&manpath=FreeBSD+11.0-RELEASE
+(Note that other union/overlay implementations may employ different methods
+for storing whiteout entries; specifically, BSD union mount employs
+a separate inode type,
+.BR DT_WHT ,
+which, while supported by some filesystems available in Linux,
+such as CODA and XFS, is ignored by the kernel's whiteout support code,
+as of Linux 4.19, at least.)
+.IP
+.B RENAME_WHITEOUT
+requires the same privileges as creating a device node (i.e., the
+.BR CAP_MKNOD
+capability).
+.IP
+.B RENAME_WHITEOUT
+can't be employed together with
+.BR RENAME_EXCHANGE .
+.IP
+.B RENAME_WHITEOUT
+requires support from the underlying filesystem.
+Among the filesystems that provide that support are
+tmpfs (since Linux 3.18),
+.\" tmpfs: commit 46fdb794e3f52ef18b859ebc92f0a9d7db21c5df
+ext4 (since Linux 3.18),
+.\" ext4: commit cd808deced431b66b5fa4e5c193cb7ec0059eaff
+XFS (since Linux 4.1),
+.\" XFS: commit 7dcf5c3e4527cfa2807567b00387cf2ed5e07f00
+f2fs (since Linux 4.2),
+.\" f2fs: commit 7e01e7ad746bc8198a8b46163ddc73a1c7d22339
+btrfs (since Linux 4.7),
+.\" btrfs: commit cdd1fedf8261cd7a73c0596298902ff4f0f04492
+and ubifs (since Linux 4.9).
+.\" ubifs: commit 9e0a1fff8db56eaaebb74b4a3ef65f86811c4798
.SH RETURN VALUE
On success, zero is returned.
On error, \-1 is returned, and
even if the same filesystem is mounted on both.)
.PP
The following additional errors can occur for
-.BR renameat ():
+.BR renameat ()
+and
+.BR renameat2 ():
.TP
.B EBADF
.I olddirfd
.I newpath
and
.I newdirfd
+.PP
+The following additional errors can occur for
+.BR renameat2 ():
+.TP
+.B EEXIST
+.I flags
+contains
+.B RENAME_NOREPLACE
+and
+.I newpath
+already exists.
+.TP
+.B EINVAL
+An invalid flag was specified in
+.IR flags .
+.TP
+.B EINVAL
+Both
+.B RENAME_NOREPLACE
+and
+.B RENAME_EXCHANGE
+were specified in
+.IR flags .
+.TP
+.B EINVAL
+Both
+.B RENAME_WHITEOUT
+and
+.B RENAME_EXCHANGE
+were specified in
+.IR flags .
+.TP
+.B EINVAL
+The filesystem does not support one of the flags in
+.IR flags .
+.TP
+.B ENOENT
+.I flags
+contains
+.B RENAME_EXCHANGE
+and
+.IR newpath
+does not exist.
+.TP
+.B EPERM
+.B RENAME_WHITEOUT
+was specified in
+.IR flags ,
+but the caller does not have the
+.B CAP_MKNOD
+capability.
.SH VERSIONS
.BR renameat ()
was added to Linux in kernel 2.6.16;
library support was added to glibc in version 2.4.
+.PP
+.BR renameat2 ()
+was added to Linux in kernel 3.15; library support was added in glibc 2.28.
.SH CONFORMING TO
.BR rename ():
4.3BSD, C89, C99, POSIX.1-2001, POSIX.1-2008.
-
+.PP
.BR renameat ():
POSIX.1-2008.
+.PP
+.BR renameat2 ()
+is Linux-specific.
+.SH NOTES
+.\"
+.SS Glibc notes
+On older kernels where
+.BR renameat ()
+is unavailable, the glibc wrapper function falls back to the use of
+.BR rename ().
+When
+.I oldpath
+and
+.I newpath
+are relative pathnames,
+glibc constructs pathnames based on the symbolic links in
+.IR /proc/self/fd
+that correspond to the
+.I olddirfd
+and
+.IR newdirfd
+arguments.
.SH BUGS
On NFS filesystems, you can not assume that if the operation
failed, the file was not renamed.
projects / thirdparty / man-pages.git / blobdiff