OPEN(2) System Calls Manual OPEN(2)

open, openat — open or create a file for reading, writing or executing

Standard C Library (libc, -lc)

#include <fcntl.h>

int
open(const char *path, int flags, ...);

int
openat(int fd, const char *path, int flags, ...);

The file name specified by path is opened for either execution or reading and/or writing as specified by the argument flags and the file descriptor returned to the calling process. The flags argument may indicate the file is to be created if it does not exist (by specifying the O_CREAT flag). In this case () and openat() require an additional argument mode_t mode, and the file is created with mode mode as described in chmod(2) and modified by the process' umask value (see umask(2)).

The () function is equivalent to the open() function except in the case where the path is relative. In that case, it is looked up from a directory whose file descriptor was passed as fd. Search permission is required on this directory except if fd was opened with the O_SEARCH flag. fd can be set to AT_FDCWD in order to specify the current directory.

The flags are specified by 'ing the values listed below. Applications must specify exactly one of these four values (file access methods):

Open for reading only.
Open for writing only.
Open for execute only.
Open for reading and writing.

Any combination of the following may be used as well:

Do not block on open or for data to become available.
Append to the file on each write.
Create the file if it does not exist. The third argument of type mode_t is used to compute the mode bits of the file as described in chmod(2) and modified by the process' umask value (see umask(2)).
Truncate size to 0.
Error if O_CREAT and the file already exists.
Atomically obtain a shared lock.
Atomically obtain an exclusive lock.
If last path element is a symlink, don't follow it. This option is provided for compatibility with other operating systems, but its security value is questionable.
Set the close(2) on exec(3) flag.
Set the close(2) on fork(2) flag.
Return EPIPE instead of raising SIGPIPE.
If set, write operations will be performed according to synchronized I/O data integrity completion: each write will wait for the file data to be committed to stable storage.
If set, write operations will be performed according to synchronized I/O file integrity completion: each write will wait for both the file data and file status to be committed to stable storage.
If set, read operations will complete at the same level of integrity which is in effect for write operations: if specified together with O_SYNC, each read will wait for the file status to be committed to stable storage.

Combining O_RSYNC with O_DSYNC only, or specifying it without any other synchronized I/O integrity completion flag set, has no further effect.

Alternate I/O semantics will be used for read and write operations on the file descriptor. Alternate semantics are defined by the underlying layers and will not have any alternate effect in most cases.
If the file is a terminal device, the opened device is not made the controlling terminal for the session. This flag has no effect on NetBSD, since the system defaults to the abovementioned behaviour. The flag is present only for standards conformance.
If set on a regular file, data I/O operations will not buffer the data being transferred in the kernel's cache, but rather transfer the data directly between user memory and the underlying device driver if possible. This flag is advisory; the request may be performed in the normal buffered fashion if certain conditions are not met, e.g. if the request is not sufficiently aligned or if the file is mapped.

To meet the alignment requirements for direct I/O, the file offset, the length of the I/O and the address of the buffer in memory must all be multiples of DEV_BSIZE (512 bytes). If the I/O request is made using an interface that supports scatter/gather via struct iovec, each element of the request must meet the above alignment constraints.

Fail if the file is not a directory.
Fail if the path does not refer to a regular file.
Enable the SIGIO signal to be sent to the process group when I/O is possible, e.g., upon availability of data to be read.

Note: This is broken in (); it must be set explicitly with the F_SETFL command to fcntl(2).

If opening a directory, search permission checks will not be performed on subsequent usage of the file descriptor for looking up relative paths by faccessat(2), fchmodat(2), fchownat(2), fstatat(2), linkat(2), mkdirat(2), mkfifoat(2), mknodat(2), openat(2), readlinkat(2), symlinkat(2), unlinkat(2), and utimensat(2).

Opening a file with O_APPEND set causes each write on the file to be appended to the end. If O_TRUNC is specified and the file exists, the file is truncated to zero length.

If O_EXCL is set with O_CREAT and the file already exists, () returns an error. This may be used to implement a simple exclusive access locking mechanism. If O_EXCL is set and the last component of the pathname is a symbolic link, open() will fail even if the symbolic link points to a non-existent name.

If the O_NONBLOCK flag is specified, do not wait for the device or file to be ready or available. If the () call would result in the process being blocked for some reason (e.g., waiting for carrier on a dialup line), open() returns immediately. This flag also has the effect of making all subsequent I/O on the open file non-blocking.

When opening a file, a lock with flock(2) semantics can be obtained by setting O_SHLOCK for a shared lock, or O_EXLOCK for an exclusive lock. If creating a file with O_CREAT, the request for the lock will never fail (provided that the underlying file system supports locking).

If () is successful, the file pointer used to mark the current position within the file is set to the beginning of the file.

When a new file is created it is given the group of the directory which contains it.

Unless inhibited by the O_CLOEXEC flag, the new descriptor is set to remain open across execve(2) system calls; and similarly for fork(2) calls with O_CLOFORK; see close(2) and fcntl(2).

The system imposes a limit on the number of file descriptors open simultaneously by one process. Calling getdtablesize(3) returns the current system limit.

If successful, open() and openat() returns a non-negative integer, termed a file descriptor. Otherwise, a value of -1 is returned and errno is set to indicate the error.

The named file is opened unless:

[]
Search permission is denied for a component of the path prefix, the required permissions (for reading and/or writing) are denied for the given flags, or O_CREAT is specified, the file does not exist, and the directory in which it is to be created does not permit writing.
[]
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; or 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.
[]
and O_EXCL were specified and the file exists.
[]
path points outside the process's allocated address space.
[]
was specified, but the last path component is a symlink. : IEEE Std 1003.1-2008 (“POSIX.1”) specifies returning [] for this case.
[]
is specified and the last path component is not a regular file.
[]
The open() operation was interrupted by a signal.
[]
An I/O error occurred while making the directory entry or allocating the inode for O_CREAT.
[]
The named file is a directory, and the arguments specify it is to be opened for writing.
[]
Too many symbolic links were encountered in translating the pathname.
[]
The process has already reached its limit for open file descriptors.
[]
A component of a pathname exceeded {NAME_MAX} characters, or an entire path name exceeded {PATH_MAX} characters.
[]
The system file table is full.
[]
is not set and the named file does not exist, or a component of the path name that must exist does not exist.
[]
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; or 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.
[]
A component of the path prefix is not a directory; or O_DIRECTORY is specified and the last path component is not a directory.
[]
The named file is a character special or block special file, and the device associated with this special file does not exist, or the named file is a FIFO, O_NONBLOCK and O_WRONLY is set and no process has the file open for reading.
[]
or O_EXLOCK is specified but the underlying file system does not support locking; or an attempt was made to open a socket (not currently implemented).
[]
The file's flags (see chflags(2)) don't allow the file to be opened.
[]
The named file resides on a read-only file system, and the file is to be modified.
[]
The file is a pure procedure (shared text) file that is being executed and the open() call requests write access.

In addition, openat() will fail if:

[]
path does not specify an absolute path and fd is neither AT_FDCWD nor a valid file descriptor open for reading or searching.
[]
An attempt was made to open a descriptor with an illegal combination of O_RDONLY, O_WRONLY, O_RDWR, and O_EXEC.
[]
path is not an absolute path and fd is a file descriptor associated with a non-directory file.

chmod(2), close(2), dup(2), faccessat(2), fchmodat(2), fchownat(2), fcntl(2), fstatat(2), linkat(2), lseek(2), mkdirat(2), mkfifoat(2), mknodat(2), read(2), readlinkat(2), symlinkat(2), umask(2), unlinkat(2), utimensat(2), write(2), getdtablesize(3)

The open() function conforms to IEEE Std 1003.1-1990 (“POSIX.1”). openat() conforms to IEEE Std 1003.1-2008 (“POSIX.1”).

The flags values O_DSYNC, O_SYNC and O_RSYNC are extensions defined in IEEE Std 1003.1b-1993 (“POSIX.1b”).

O_SEARCH is defined in IEEE Std 1003.1-2008 (“POSIX.1”).

O_CLOFORK conforms to IEEE Std 1003.1-2024 (“POSIX.1”).

The O_SHLOCK and O_EXLOCK flags are non-standard extensions and should not be used if portability is of concern.

An open() function call appeared in Version 1 AT&T UNIX. The O_CLOFORK implementation first appeared in FreeBSD 15.0, DragonFly 6.5 and NetBSD 11.0.

O_ASYNC doesn't actually work as advertised with open; you must set O_ASYNC explicitly with the F_SETFL command to fcntl(2).

July 8, 2025 NetBSD 11.0