path: root/lib/libsys/open.2

.\" Copyright (c) 1980, 1991, 1993
.\"	The Regents of the University of California.  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.
.\" 3. Neither the name of the University nor the names of its contributors
.\"    may be used to endorse or promote products derived from this software
.\"    without specific prior written permission.
.\"
.\" THIS SOFTWARE IS PROVIDED BY THE REGENTS 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 REGENTS 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.
.\"
.Dd May 17, 2025
.Dt OPEN 2
.Os
.Sh NAME
.Nm open , openat
.Nd open or create a file for reading, writing or executing
.Sh LIBRARY
.Lb libc
.Sh SYNOPSIS
.In fcntl.h
.Ft int
.Fn open "const char *path" "int flags" "..."
.Ft int
.Fn openat "int fd" "const char *path" "int flags" "..."
.Sh DESCRIPTION
The file name specified by
.Fa path
is opened
for either execution or reading and/or writing as specified by the
argument
.Fa flags
and the file descriptor returned to the calling process.
The
.Fa flags
argument may indicate the file is to be
created if it does not exist (by specifying the
.Dv O_CREAT
flag).
In this case
.Fn open
and
.Fn openat
require an additional argument
.Fa "mode_t mode" ,
and the file is created with mode
.Fa mode
as described in
.Xr chmod 2
and modified by the process' umask value (see
.Xr umask 2 ) .
.Pp
The
.Fn openat
function is equivalent to the
.Fn open
function except in the case where the
.Fa path
specifies a relative path.
For
.Fn openat
and relative
.Fa path ,
the file to be opened is determined relative to the directory
associated with the file descriptor
.Fa fd
instead of the current working directory.
The
.Fa flag
parameter and the optional fourth parameter correspond exactly to
the parameters of
.Fn open .
If
.Fn openat
is passed the special value
.Dv AT_FDCWD
in the
.Fa fd
parameter, the current working directory is used
and the behavior is identical to a call to
.Fn open .
.Pp
When
.Fn openat
is called with an absolute
.Fa path ,
it ignores the
.Fa fd
argument.
.Pp
In
.Xr capsicum 4
capability mode,
.Fn open
is not permitted.
The
.Fa path
argument to
.Fn openat
must be strictly relative to a file descriptor
.Fa fd ;
that is,
.Fa path
must not be an absolute path and must not contain ".." components
which cause the path resolution to escape the directory hierarchy
starting at
.Fa fd .
Additionally, no symbolic link in
.Fa path
may target absolute path or contain escaping ".." components.
.Fa fd
must not be
.Dv AT_FDCWD .
.Pp
If the
.Dv vfs.lookup_cap_dotdot
.Xr sysctl 3
MIB is set to zero, ".." components in the paths,
used in capability mode,
are completely disabled.
If the
.Dv vfs.lookup_cap_dotdot_nonlocal
MIB is set to zero, ".." is not allowed if found on non-local filesystem.
.Pp
The
.Fa flags
are formed by
.Em or Ns 'ing
the following values:
.Pp
.Bl -tag -width O_RESOLVE_BENEATH
.It Dv O_RDONLY
open for reading only
.It Dv O_WRONLY
open for writing only
.It Dv O_RDWR
open for reading and writing
.It Dv O_EXEC
open for execute only
.It Dv O_SEARCH
open for search only
(an alias for
.Dv O_EXEC
typically used with
.Dv O_DIRECTORY )
.It Dv O_NONBLOCK
do not block on open
.It Dv O_APPEND
set file pointer to the end of the file before each write
.It Dv O_CREAT
create file if it does not exist
.It Dv O_TRUNC
truncate size to 0
.It Dv O_EXCL
fail if
.Dv O_CREAT
is set and the file exists
.It Dv O_SHLOCK
atomically obtain a shared lock
.It Dv O_EXLOCK
atomically obtain an exclusive lock
.It Dv O_DIRECT
read and write directly from the backing store
.It Dv O_FSYNC
synchronous data and metadata writes
.Pq historical synonym for Dv O_SYNC
.It Dv O_SYNC
synchronous data and metadata writes
.It Dv O_DSYNC
synchronous data writes
.It Dv O_NOFOLLOW
do not follow symlinks
.It Dv O_NOCTTY
ignored
.It Dv O_TTY_INIT
ignored
.It Dv O_DIRECTORY
error if file is not a directory
.It Dv O_CLOEXEC
automatically close file on
.Xr execve 2
.It Dv O_CLOFORK
automatically close file on any child process created with
.Fn fork 2
.It Dv O_VERIFY
verify the contents of the file with
.Xr mac_veriexec 4
.It Dv O_RESOLVE_BENEATH
.Pq Xr openat 2 only
path resolution must not cross the
.Fa fd
directory
.It Dv O_PATH
record only the target path in the opened descriptor
.It Dv O_EMPTY_PATH
.Pq Xr openat 2 only
open file referenced by
.Fa fd
if path is empty
.It Dv O_NAMEDATTR
open a named attribute or named attribute directory
.El
.Pp
Exactly one of the flags
.Dv O_RDONLY ,
.Dv O_WRONLY ,
.Dv O_RDWR ,
or
.Dv O_EXEC
must be provided.
.Pp
Opening a file with
.Dv O_APPEND
set causes each write on the resulting file descriptor
to be appended to the end of the file.
.Pp
If
.Dv O_TRUNC
is specified and the
file exists, the file is truncated to zero length.
.Pp
If
.Dv O_CREAT
is set, but file already exists,
this flag has no effect except when
.Dv O_EXCL
is set too, in this case
.Fn open
fails with
.Er EEXIST .
This may be used to
implement a simple exclusive access locking mechanism.
In all other cases, the file is created
and the access permission bits (see
.Xr chmod 2)
of the file mode
are set to the value of the third argument taken as
.Fa "mode_t mode"
and passed through the
.Xr umask 2 .
This argument does not affect whether the file is opened
for reading, writing, or for both.
The open' request for a lock on the file, created with
.Dv O_CREAT ,
will never fail
provided that the underlying file system supports locking;
see also
.Dv O_SHLOCK
and
.Dv O_EXLOCK
below.
.Pp
If
.Dv O_EXCL
is set and the last component of the pathname is
a symbolic link,
.Fn open
will fail even if the symbolic
link points to a non-existent name.
.Pp
If
.Dv O_NONBLOCK
is specified and the
.Fn open
system call would
block for some reason (for example, waiting for
carrier on a dialup line),
.Fn open
returns immediately.
The descriptor remains in non-blocking mode for subsequent operations.
.Pp
If
.Dv O_SYNC
is used in the mask, all writes will
immediately and synchronously be written to disk.
.Dv O_FSYNC
is an historical synonym for
.Dv O_SYNC .
.Pp
If
.Dv O_DSYNC
is used in the mask, all data and metadata required to read the data will be
synchronously written to disk, but changes to metadata such as file access and
modification timestamps may be written later.
.Pp
If
.Dv O_NOFOLLOW
is used in the mask and the target file passed to
.Fn open
is a symbolic link then the
.Fn open
will fail.
.Pp
When opening a file, a lock with
.Xr flock 2
semantics can be obtained by setting
.Dv O_SHLOCK
for a shared lock, or
.Dv O_EXLOCK
for an exclusive lock.
.Pp
.Dv O_DIRECT
may be used to minimize or eliminate the cache effects of reading and writing.
The system will attempt to avoid caching the data you read or write.
If it cannot avoid caching the data,
it will minimize the impact the data has on the cache.
Use of this flag can drastically reduce performance if not used with care.
The semantics of this flag are filesystem dependent,
and some filesystems may ignore it entirely.
.Pp
.Dv O_NOCTTY
may be used to ensure the OS does not assign this file as the
controlling terminal when it opens a tty device.
This is the default on
.Fx ,
but is present for
POSIX
compatibility.
The
.Fn open
system call will not assign controlling terminals on
.Fx .
.Pp
.Dv O_TTY_INIT
may be used to ensure the OS restores the terminal attributes when
initially opening a TTY.
This is the default on
.Fx ,
but is present for
POSIX
compatibility.
The initial call to
.Fn open
on a TTY will always restore default terminal attributes on
.Fx .
.Pp
.Dv O_DIRECTORY
may be used to ensure the resulting file descriptor refers to a
directory.
This flag can be used to prevent applications with elevated privileges
from opening files which are even unsafe to open with
.Dv O_RDONLY ,
such as device nodes.
.Pp
.Dv O_CLOEXEC
may be used to set
.Dv FD_CLOEXEC
flag for the newly returned file descriptor.
.Pp
.Dv O_CLOFORK
may be used to set
.Dv FD_CLOFORK
flag for the newly returned file descriptor.
The file will be closed on any child process created with
.Fn fork 2 ,
.Fn vfork 2
or
.Fn rfork 2
with the
.Dv RFFDG
flag, remaining open in the parent.
Both the
.Dv O_CLOEXEC
and
.Dv O_CLOFORK
flags can be modified with the
.Dv F_SETFD
.Fn fcntl 2
command.
.Pp
.Dv O_VERIFY
may be used to indicate to the kernel that the contents of the file should
be verified before allowing the open to proceed.
The details of what
.Dq verified
means is implementation specific.
The run-time linker (rtld) uses this flag to ensure shared objects have
been verified before operating on them.
.Pp
.Dv O_RESOLVE_BENEATH
returns
.Er ENOTCAPABLE
if any intermediate component of the specified relative path does not
reside in the directory hierarchy beneath the starting directory.
Absolute paths or even the temporal escape from beneath of the starting
directory is not allowed.
.Pp
When a directory
is opened with
.Dv O_SEARCH ,
execute permissions are checked at open time.
The returned file descriptor
may not be used for any read operations like
.Xr getdirentries 2 .
The primary use of this descriptor is as the lookup descriptor for the
.Fn *at
family of functions.
If
.Dv O_SEARCH
was not requested at open time, then the
.Fn *at
functions use the current directory permissions for the directory referenced
by the descriptor at the time of the
.Fn *at
call.
.Pp
.Dv O_PATH
returns a file descriptor that can be used as a directory file descriptor for
.Fn openat
and other system calls taking a file descriptor argument, like
.Xr fstatat 2
and others.
The other functionality of the returned file descriptor is limited to
the following descriptor-level operations:
.Pp
.Bl -tag -width __acl_aclcheck_fd -offset indent -compact
.It Xr fcntl 2
but advisory locking is not allowed
.It Xr dup 2
.It Xr close 2
.It Xr fstat 2
.It Xr fstatfs 2
.It Xr fchdir 2
.It Xr fchroot 2
.It Xr fexecve 2
.It Xr funlinkat 2
can be passed as the third argument
.It Dv SCM_RIGHTS
can be passed over a
.Xr unix 4
socket using a
.Dv SCM_RIGHTS
message
.It Xr kqueue 2
only with
.Dv EVFILT_VNODE
.It Xr __acl_get_fd 2
.It Xr __acl_aclcheck_fd 2
.It Xr extattr 2
.It Xr capsicum 4
can be passed to
.Fn cap_*_limit
and
.Fn cap_*_get
system calls (such as
.Xr cap_rights_limit 2 ) .
.El
.Pp
Other operations like
.Xr read 2 ,
.Xr ftruncate 2 ,
and any other that operate on file and not on file descriptor (except
.Xr fstat 2 ) ,
are not allowed.
.Pp
A file descriptor created with the
.Dv O_PATH
flag can be opened as a normal (operable) file descriptor by
specifying it as the
.Fa fd
argument to
.Fn openat
with an empty
.Fa path
and the
.Dv O_EMPTY_PATH
flag.
Such an open behaves as if the current path of the file referenced by
.Fa fd
is passed, except that path walk permissions are not checked.
See also the description of
.Dv AT_EMPTY_PATH
flag for
.Xr fstatat 2
and related syscalls.
.Pp
Conversely, a file descriptor
.Dv fd
referencing a filesystem file can be converted to the
.Dv O_PATH
type of descriptor by using the following call
.Dl opath_fd = openat(fd, \[dq]\[dq], O_EMPTY_PATH | O_PATH);
.Pp
If successful,
.Fn open
returns a non-negative integer, termed a file descriptor.
It returns \-1 on failure.
The file descriptor value returned is the lowest numbered descriptor
currently not in use by the process.
The file pointer used to mark the current position within the
file is set to the beginning of the file.
.Pp
If a sleeping open of a device node from
.Xr devfs 4
is interrupted by a signal, the call always fails with
.Er EINTR ,
even if the
.Dv SA_RESTART
flag is set for the signal.
A sleeping open of a fifo (see
.Xr mkfifo 2 )
is restarted as normal.
.Pp
When a new file is created, it is assigned the group of the directory
which contains it.
.Pp
Unless
.Dv O_CLOEXEC
flag was specified,
the new descriptor is set to remain open across
.Xr execve 2
system calls; see
.Xr close 2 ,
.Xr fcntl 2
and the description of the
.Dv O_CLOEXEC
flag.
.Pp
When the
.Dv O_NAMEDATTR
flag is specified for an
.Fn openat
where the
.Fa fd
argument is for a file object,
a named attribute for the file object
is opened and not the file object itself.
If the
.Dv O_CREAT
flag has been specified as well, the named attribute will be
created if it does not exist.
When the
.Dv O_NAMEDATTR
flag is specified for a
.Fn open ,
a named attribute for the current working directory is opened and
not the current working directory.
The
.Fa path
argument for this
.Fn openat
or
.Fn open
must be a single component name with no embedded
.Ql / .
If the
.Fa path
argument is
.Ql .\&
then the named attribute directory for the file object is opened.
(See
.Xr named_attribute 7
for more information.)
.Pp
The system imposes a limit on the number of file descriptors
open simultaneously by one process.
The
.Xr getdtablesize 2
system call returns the current system limit.
.Sh RETURN VALUES
If successful,
.Fn open
and
.Fn openat
return a non-negative integer, termed a file descriptor.
They return \-1 on failure, and set
.Va errno
to indicate the error.
.Sh ERRORS
The named file is opened unless:
.Bl -tag -width Er
.It Bq Er ENOTDIR
A component of the path prefix is not a directory.
.It Bq Er ENAMETOOLONG
A component of a pathname exceeded 255 characters,
or an entire path name exceeded 1023 characters.
.It Bq Er ENOENT
.Dv O_CREAT
is not set and the named file does not exist.
.It Bq Er ENOENT
A component of the path name that must exist does not exist.
.It Bq Er EACCES
Search permission is denied for a component of the path prefix.
.It Bq Er EACCES
The required permissions (for reading and/or writing)
are denied for the given flags.
.It Bq Er EACCES
.Dv O_TRUNC
is specified and write permission is denied.
.It Bq Er EACCES
.Dv O_CREAT
is specified,
the file does not exist,
and the directory in which it is to be created
does not permit writing.
.It Bq Er EPERM
.Dv O_CREAT
is specified, the file does not exist, and the directory in which it is to be
created has its immutable flag set, see the
.Xr chflags 2
manual page for more information.
.It Bq Er EPERM
The named file has its immutable flag set and the file is to be modified.
.It Bq Er EPERM
The named file has its append-only flag set, the file is to be modified, and
.Dv O_TRUNC
is specified or
.Dv O_APPEND
is not specified.
.It Bq Er ELOOP
Too many symbolic links were encountered in translating the pathname.
.It Bq Er EISDIR
The named file is a directory, and the arguments specify
it is to be modified.
.It Bq Er EISDIR
The named file is a directory, and the flags specified
.Dv O_CREAT
without
.Dv O_DIRECTORY .
.It Bq Er EROFS
The named file resides on a read-only file system,
and the file is to be modified.
.It Bq Er EROFS
.Dv O_CREAT
is specified and the named file would reside on a read-only file system.
.It Bq Er EMFILE
The process has already reached its limit for open file descriptors.
.It Bq Er ENFILE
The system file table is full.
.It Bq Er EMLINK
.Dv O_NOFOLLOW
was specified and the target is a symbolic link.
POSIX
specifies a different error for this case; see the note in
.Sx STANDARDS
below.
.It Bq Er ENXIO
The named file is a character special or block
special file, and the device associated with this special file
does not exist.
.It Bq Er ENXIO
.Dv O_NONBLOCK
is set, the named file is a fifo,
.Dv O_WRONLY
is set, and no process has the file open for reading.
.It Bq Er EINTR
The
.Fn open
operation was interrupted by a signal.
.It Bq Er EOPNOTSUPP
.Dv O_SHLOCK
or
.Dv O_EXLOCK
is specified but the underlying file system does not support locking.
.It Bq Er EOPNOTSUPP
The named file is a special file mounted through a file system that
does not support access to it (for example, NFS).
.It Bq Er EWOULDBLOCK
.Dv O_NONBLOCK
and one of
.Dv O_SHLOCK
or
.Dv O_EXLOCK
is specified and the file is locked.
.It Bq Er ENOSPC
.Dv O_CREAT
is specified,
the file does not exist,
and the directory in which the entry for the new file is being placed
cannot be extended because there is no space left on the file
system containing the directory.
.It Bq Er ENOSPC
.Dv O_CREAT
is specified,
the file does not exist,
and there are no free inodes on the file system on which the
file is being created.
.It Bq Er EDQUOT
.Dv O_CREAT
is specified,
the file does not exist,
and the directory in which the entry for the new file
is being placed cannot be extended because the
user's quota of disk blocks on the file system
containing the directory has been exhausted.
.It Bq Er EDQUOT
.Dv O_CREAT
is specified,
the file does not exist,
and the user's quota of inodes on the file system on
which the file is being created has been exhausted.
.It Bq Er EIO
An I/O error occurred while making the directory entry or
allocating the inode for
.Dv O_CREAT .
.It Bq Er EINTEGRITY
Corrupted data was detected while reading from the file system.
.It Bq Er ETXTBSY
The file is a pure procedure (shared text) file that is being
executed and the
.Fn open
system call requests write access.
.It Bq Er EFAULT
The
.Fa path
argument
points outside the process's allocated address space.
.It Bq Er EEXIST
.Dv O_CREAT
and
.Dv O_EXCL
were specified and the file exists.
.It Bq Er EOPNOTSUPP
An attempt was made to open a socket (not currently implemented).
.It Bq Er EINVAL
An attempt was made to open a descriptor with an illegal combination
of
.Dv O_RDONLY ,
.Dv O_WRONLY ,
or
.Dv O_RDWR ,
and
.Dv O_EXEC
or
.Dv O_SEARCH .
.It Bq Er EINVAL
.Dv O_CREAT
is specified,
and the last component of the
.Fa path
argument is invalid on the file system on which the file is being created.
.It Bq Er EBADF
The
.Fa path
argument does not specify an absolute path and the
.Fa fd
argument is
neither
.Dv AT_FDCWD
nor a valid file descriptor open for searching.
.It Bq Er ENOTDIR
The
.Fa path
argument is not an absolute path and
.Fa fd
is neither
.Dv AT_FDCWD
nor a file descriptor associated with a directory.
.It Bq Er ENOTDIR
.Dv O_DIRECTORY
is specified and the file is not a directory.
.It Bq Er ECAPMODE
.Dv AT_FDCWD
is specified and the process is in capability mode.
.It Bq Er ECAPMODE
.Fn open
was called and the process is in capability mode.
.It Bq Er ENOTCAPABLE
.Fa path
is an absolute path and the process is in capability mode.
.It Bq Er ENOTCAPABLE
.Fa path
is an absolute path and
.Dv O_RESOLVE_BENEATH
is specified.
.It Bq Er ENOTCAPABLE
.Fa path
contains a ".." component leading to a directory outside
of the directory hierarchy specified by
.Fa fd
and the process is in capability mode.
.It Bq Er ENOTCAPABLE
.Fa path
contains a ".." component leading to a directory outside
of the directory hierarchy specified by
.Fa fd
and
.Dv O_RESOLVE_BENEATH
is specified.
.It Bq Er ENOTCAPABLE
.Fa path
contains a ".." component, the
.Dv vfs.lookup_cap_dotdot
.Xr sysctl 3
is set, and the process is in capability mode.
.It Bq Er ENOATTR
.Dv O_NAMEDATTR
has been specified and the file object is not a named attribute
directory or named attribute.
.El
.Sh SEE ALSO
.Xr chmod 2 ,
.Xr close 2 ,
.Xr dup 2 ,
.Xr fexecve 2 ,
.Xr fhopen 2 ,
.Xr getdtablesize 2 ,
.Xr getfh 2 ,
.Xr lgetfh 2 ,
.Xr lseek 2 ,
.Xr read 2 ,
.Xr umask 2 ,
.Xr write 2 ,
.Xr fopen 3 ,
.Xr capsicum 4 ,
.Xr named_attribute 7
.Sh STANDARDS
These functions are specified by
.St -p1003.1-2008 .
.Pp
.Fx
sets
.Va errno
to
.Er EMLINK instead of
.Er ELOOP
as specified by
POSIX
when
.Dv O_NOFOLLOW
is set in flags and the final component of pathname is a symbolic link
to distinguish it from the case of too many symbolic link traversals
in one of its non-final components.
.Pp
The Open Group Extended API Set 2 specification, that introduced the
.Fn *at
API, required that the test for whether
.Fa fd
is searchable is based on whether
.Fa fd
is open for searching, not whether the underlying directory currently
permits searches.
The present implementation of the
.Fa openat
system call is believed to be compatible with
.\" .St -p1003.1-2017 ,
.\" XXX: This should be replaced in the future when an appropriate argument to
.\" the St macro is available: -p1003.1-2017
.No IEEE Std 1003.1-2008, 2017 Edition ("POSIX.1") ,
which specifies that behavior for
.Dv O_SEARCH ,
in the absence of the flag the implementation checks the current
permissions of a directory.
.Sh HISTORY
The
.Fn open
function appeared in
.At v1 .
The
.Fn openat
function was introduced in
.Fx 8.0 .
.Dv O_DSYNC
appeared in 13.0.
.Dv O_NAMEDATTR
appeared in 15.0.
.Dv O_CLOFORK
appeared in
.Fx 15.0 .
.Sh BUGS
The
.Fa mode
argument is variadic and may result in different calling conventions
than might otherwise be expected.