{"id":3852,"date":"2025-09-04T10:06:15","date_gmt":"2025-09-04T09:06:15","guid":{"rendered":"https:\/\/workboot.fr\/ciela\/?page_id=3852"},"modified":"2026-06-16T10:54:52","modified_gmt":"2026-06-16T09:54:52","slug":"open-close","status":"publish","type":"page","link":"https:\/\/workboot.fr\/ciela\/open-close\/","title":{"rendered":"open\/close"},"content":{"rendered":"\n

Tout est fichiers <\/strong> : dans Unix et Linux, presque tout est consid\u00e9r\u00e9 comme un fichier.<\/p>\n\n\n\n

Ce n’est pas une simple m\u00e9taphore, c’est une conception profonde du syst\u00e8me qui le rend extr\u00eamement puissant et simple \u00e0 la fois. Voici ce que cela signifie concr\u00e8tement :<\/p>\n\n\n\n

Ces 2 fonctions sont intrins\u00e8quement li\u00e9es<\/p>\n\n\n\n

Si on ouvre un fichier , il faudra forc\u00e9ment le fermer dans votre code.<\/p>\n\n\n\n

un fichier est une s\u00e9quence de donn\u00e9es , ici plus particuli\u00e8rement des octets .<\/p>\n\n\n\n

On va utiliser aussi les fonctions write et read , qui veulent dire \u00e9crire et lire <\/p>\n\n\n\n

man open<\/strong><\/summary>\n

open(2) \u2014 Linux manual page<\/p>\n\n\n\n

NAME<\/a> | LIBRARY<\/a> | SYNOPSIS<\/a> | DESCRIPTION<\/a> | RETURN VALUE<\/a> | ERRORS<\/a> | VERSIONS<\/a> | STANDARDS<\/a> | HISTORY<\/a> | NOTES<\/a> | BUGS<\/a> | SEE ALSO<\/a> | COLOPHON<\/a><\/td><\/tr>
 <\/td><\/td><\/tr><\/tbody><\/table><\/figure>\n\n\n\n
open<\/em>(2)                    System Calls Manual                    open<\/em>(2)<\/strong>
<\/pre>\n\n\n\n
<\/a>NAME         top<\/a><\/h2>\n\n\n\n
       open, openat, creat - open and possibly create a file
<\/pre>\n\n\n\n
<\/a>LIBRARY         top<\/a><\/h2>\n\n\n\n
       Standard C library (libc<\/em>, -lc<\/em>)
<\/pre>\n\n\n\n
<\/a>SYNOPSIS         top<\/a><\/h2>\n\n\n\n
       #include <fcntl.h><\/strong>

       int open(const char *<\/strong>pathname<\/em>, int <\/strong>flags<\/em>, ...<\/strong>
                  \/* mode_t <\/strong>mode<\/em> *\/ );<\/strong>

       int creat(const char *<\/strong>pathname<\/em>, mode_t <\/strong>mode<\/em>);<\/strong>

       int openat(int <\/strong>dirfd<\/em>, const char *<\/strong>pathname<\/em>, int <\/strong>flags<\/em>, ...<\/strong>
                  \/* mode_t <\/strong>mode<\/em> *\/ );<\/strong>

       \/* Documented separately, in openat2(2)<\/a>: *\/
       int openat2(int <\/strong>dirfd<\/em>, const char *<\/strong>pathname<\/em>,<\/strong>
                  const struct open_how *<\/strong>how<\/em>, size_t <\/strong>size<\/em>);<\/strong>

   Feature Test Macro Requirements for glibc (see
   feature_test_macros(7)<\/a>):

       openat<\/strong>():
           Since glibc 2.10:
               _POSIX_C_SOURCE >= 200809L
           Before glibc 2.10:
               _ATFILE_SOURCE
<\/pre>\n\n\n\n
<\/a>DESCRIPTION         top<\/a><\/h2>\n\n\n\n
       The open<\/strong>() system call opens the file specified by pathname<\/em>.  If
       the specified file does not exist, it may optionally (if O_CREAT<\/strong>
       is specified in flags<\/em>) be created by open<\/strong>().

       The return value of open<\/strong>() is a file descriptor, a small,
       nonnegative integer that is an index to an entry in the process's
       table of open file descriptors.  The file descriptor is used in
       subsequent system calls (read(2)<\/a>, write(2)<\/a>, lseek(2)<\/a>, fcntl(2)<\/a>,
       etc.)  to refer to the open file.  The file descriptor returned by
       a successful call will be the lowest-numbered file descriptor not
       currently open for the process.

       By default, the new file descriptor is set to remain open across
       an execve(2)<\/a> (i.e., the FD_CLOEXEC <\/strong>file descriptor flag described
       in fcntl(2)<\/a> is initially disabled); the O_CLOEXEC <\/strong>flag, described
       below, can be used to change this default.  The file offset is set
       to the beginning of the file (see lseek(2)<\/a>).

       A call to open<\/strong>() creates a new open file description<\/em>, an entry in
       the system-wide table of open files.  The open file description
       records the file offset and the file status flags (see below).  A
       file descriptor is a reference to an open file description; this
       reference is unaffected if pathname<\/em> is subsequently removed or
       modified to refer to a different file.  For further details on
       open file descriptions, see NOTES.

       The argument flags<\/em> must include one of the following access modes<\/em>:
       O_RDONLY<\/strong>, O_WRONLY<\/strong>, or O_RDWR<\/strong>.  These request opening the file
       read-only, write-only, or read\/write, respectively.

       In addition, zero or more file creation flags and file status
       flags can be bitwise ORed in flags<\/em>.  The file creation flags<\/em> are
       O_CLOEXEC<\/strong>, O_CREAT<\/strong>, O_DIRECTORY<\/strong>, O_EXCL<\/strong>, O_NOCTTY<\/strong>, O_NOFOLLOW<\/strong>,
       O_TMPFILE<\/strong>, and O_TRUNC<\/strong>.  The file status flags<\/em> are all of the
       remaining flags listed below.  The distinction between these two
       groups of flags is that the file creation flags affect the
       semantics of the open operation itself, while the file status
       flags affect the semantics of subsequent I\/O operations.  The file
       status flags can be retrieved and (in some cases) modified; see
       fcntl(2)<\/a> for details.

       The full list of file creation flags and file status flags is as
       follows:

       O_APPEND<\/strong>
              The file is opened in append mode.  Before each write(2)<\/a>,
              the file offset is positioned at the end of the file, as if
              with lseek(2)<\/a>.  The modification of the file offset and the
              write operation are performed as a single atomic step.

              O_APPEND <\/strong>may lead to corrupted files on NFS filesystems if
              more than one process appends data to a file at once.  This
              is because NFS does not support appending to a file, so the
              client kernel has to simulate it, which can't be done
              without a race condition.

       O_ASYNC<\/strong>
              Enable signal-driven I\/O: generate a signal (SIGIO <\/strong>by
              default, but this can be changed via fcntl(2)<\/a>) when input
              or output becomes possible on this file descriptor.  This
              feature is available only for terminals, pseudoterminals,
              sockets, and (since Linux 2.6) pipes and FIFOs.  See
              fcntl(2)<\/a> for further details.  See also BUGS, below.

       O_CLOEXEC <\/strong>(since Linux 2.6.23)
              Enable the close-on-exec flag for the new file descriptor.
              Specifying this flag permits a program to avoid additional
              fcntl(2)<\/a> F_SETFD <\/strong>operations to set the FD_CLOEXEC <\/strong>flag.

              Note that the use of this flag is essential in some
              multithreaded programs, because using a separate fcntl(2)<\/a>
              F_SETFD <\/strong>operation to set the FD_CLOEXEC <\/strong>flag does not
              suffice to avoid race conditions where one thread opens a
              file descriptor and attempts to set its close-on-exec flag
              using fcntl(2)<\/a> at the same time as another thread does a
              fork(2)<\/a> plus execve(2)<\/a>.  Depending on the order of
              execution, the race may lead to the file descriptor
              returned by open<\/strong>() being unintentionally leaked to the
              program executed by the child process created by fork(2)<\/a>.
              (This kind of race is in principle possible for any system
              call that creates a file descriptor whose close-on-exec
              flag should be set, and various other Linux system calls
              provide an equivalent of the O_CLOEXEC <\/strong>flag to deal with
              this problem.)

       O_CREAT<\/strong>
              If pathname<\/em> does not exist, create it as a regular file.

              The owner (user ID) of the new file is set to the effective
              user ID of the process.

              The group ownership (group ID) of the new file is set
              either to the effective group ID of the process (System V
              semantics) or to the group ID of the parent directory (BSD
              semantics).  On Linux, the behavior depends on whether the
              set-group-ID mode bit is set on the parent directory: if
              that bit is set, then BSD semantics apply; otherwise,
              System V semantics apply.  For some filesystems, the
              behavior also depends on the bsdgroups<\/em> and sysvgroups<\/em> mount
              options described in mount(8)<\/a>.

              The mode<\/em> argument specifies the file mode bits to be
              applied when a new file is created.  If neither O_CREAT <\/strong>nor
              O_TMPFILE <\/strong>is specified in flags<\/em>, then mode<\/em> is ignored (and
              can thus be specified as 0, or simply omitted).  The mode<\/em>
              argument must <\/strong>be supplied if O_CREAT <\/strong>or O_TMPFILE <\/strong>is
              specified in flags<\/em>; if it is not supplied, some arbitrary
              bytes from the stack will be applied as the file mode.

              The effective mode is modified by the process's umask<\/em> in
              the usual way: in the absence of a default ACL, the mode of
              the created file is (mode & ~umask)<\/em>.

              Note that mode<\/em> applies only to future accesses of the newly
              created file; the open<\/strong>() call that creates a read-only file
              may well return a read\/write file descriptor.

              The following symbolic constants are provided for mode<\/em>:

              S_IRWXU  <\/strong>00700 user (file owner) has read, write, and
                       execute permission

              S_IRUSR  <\/strong>00400 user has read permission

              S_IWUSR  <\/strong>00200 user has write permission

              S_IXUSR  <\/strong>00100 user has execute permission

              S_IRWXG  <\/strong>00070 group has read, write, and execute
                       permission

              S_IRGRP  <\/strong>00040 group has read permission

              S_IWGRP  <\/strong>00020 group has write permission

              S_IXGRP  <\/strong>00010 group has execute permission

              S_IRWXO  <\/strong>00007 others have read, write, and execute
                       permission

              S_IROTH  <\/strong>00004 others have read permission

              S_IWOTH  <\/strong>00002 others have write permission

              S_IXOTH  <\/strong>00001 others have execute permission

              According to POSIX, the effect when other bits are set in
              mode<\/em> is unspecified.  On Linux, the following bits are also
              honored in mode<\/em>:

              S_ISUID  <\/strong>0004000 set-user-ID bit

              S_ISGID  <\/strong>0002000 set-group-ID bit (see inode(7)<\/a>).

              S_ISVTX  <\/strong>0001000 sticky bit (see inode(7)<\/a>).

       O_DIRECT <\/strong>(since Linux 2.4.10)
              Try to minimize cache effects of the I\/O to and from this
              file.  In general this will degrade performance, but it is
              useful in special situations, such as when applications do
              their own caching.  File I\/O is done directly to\/from user-
              space buffers.  The O_DIRECT <\/strong>flag on its own makes an
              effort to transfer data synchronously, but does not give
              the guarantees of the O_SYNC <\/strong>flag that data and necessary
              metadata are transferred.  To guarantee synchronous I\/O,
              O_SYNC <\/strong>must be used in addition to O_DIRECT<\/strong>.  See NOTES
              below for further discussion.

              A semantically similar (but deprecated) interface for block
              devices is described in raw<\/strong>(8).

       O_DIRECTORY<\/strong>
              If pathname<\/em> is not a directory, cause the open to fail.
              This flag was added in Linux 2.1.126, to avoid denial-of-
              service problems if opendir(3)<\/a> is called on a FIFO or tape
              device.

       O_DSYNC<\/strong>
              Write operations on the file will complete according to the
              requirements of synchronized I\/O data<\/em> integrity completion.

              By the time write(2)<\/a> (and similar) return, the output data
              has been transferred to the underlying hardware, along with
              any file metadata that would be required to retrieve that
              data (i.e., as though each write(2)<\/a> was followed by a call
              to fdatasync(2)<\/a>).  See VERSIONS.

       O_EXCL <\/strong>Ensure that this call creates the file: if this flag is
              specified in conjunction with O_CREAT<\/strong>, and pathname<\/em> already
              exists, then open<\/strong>() fails with the error EEXIST<\/strong>.

              When these two flags are specified, symbolic links are not
              followed: if pathname<\/em> is a symbolic link, then open<\/strong>() fails
              regardless of where the symbolic link points.

              In general, the behavior of O_EXCL <\/strong>is undefined if it is
              used without O_CREAT<\/strong>.  There is one exception: on Linux 2.6
              and later, O_EXCL <\/strong>can be used without O_CREAT <\/strong>if pathname<\/em>
              refers to a block device.  If the block device is in use by
              the system (e.g., mounted), open<\/strong>() fails with the error
              EBUSY<\/strong>.

              On NFS, O_EXCL <\/strong>is supported only when using NFSv3 or later
              on kernel 2.6 or later.  In NFS environments where O_EXCL<\/strong>
              support is not provided, programs that rely on it for
              performing locking tasks will contain a race condition.
              Portable programs that want to perform atomic file locking
              using a lockfile, and need to avoid reliance on NFS support
              for O_EXCL<\/strong>, can create a unique file on the same filesystem
              (e.g., incorporating hostname and PID), and use link(2)<\/a> to
              make a link to the lockfile.  If link(2)<\/a> returns 0, the
              lock is successful.  Otherwise, use stat(2)<\/a> on the unique
              file to check if its link count has increased to 2, in
              which case the lock is also successful.

       O_LARGEFILE<\/strong>
              (LFS) Allow files whose sizes cannot be represented in an
              off_t<\/em> (but can be represented in an off64_t<\/em>) to be opened.
              The _LARGEFILE64_SOURCE <\/strong>macro must be defined (before
              including any<\/em> header files) in order to obtain this
              definition.  Setting the _FILE_OFFSET_BITS <\/strong>feature test
              macro to 64 (rather than using O_LARGEFILE<\/strong>) is the
              preferred method of accessing large files on 32-bit systems
              (see feature_test_macros(7)<\/a>).

       O_NOATIME <\/strong>(since Linux 2.6.8)
              Do not update the file last access time (st_atime<\/em> in the
              inode) when the file is read(2)<\/a>.

              This flag can be employed only if one of the following
              conditions is true:

              \u2022  The effective UID of the process matches the owner UID
                 of the file.

              \u2022  The calling process has the CAP_FOWNER <\/strong>capability in its
                 user namespace and the owner UID of the file has a
                 mapping in the namespace.

              This flag is intended for use by indexing or backup
              programs, where its use can significantly reduce the amount
              of disk activity.  This flag may not be effective on all
              filesystems.  One example is NFS, where the server
              maintains the access time.

       O_NOCTTY<\/strong>
              If pathname<\/em> refers to a terminal device\u2014see tty(4)<\/a>\u2014it will
              not become the process's controlling terminal even if the
              process does not have one.

       O_NOFOLLOW<\/strong>
              If the trailing component (i.e., basename) of pathname<\/em> is a
              symbolic link, then the open fails, with the error ELOOP<\/strong>.
              Symbolic links in earlier components of the pathname will
              still be followed.  (Note that the ELOOP <\/strong>error that can
              occur in this case is indistinguishable from the case where
              an open fails because there are too many symbolic links
              found while resolving components in the prefix part of the
              pathname.)

              This flag is a FreeBSD extension, which was added in Linux
              2.1.126, and has subsequently been standardized in
              POSIX.1-2008.

              See also O_PATH <\/strong>below.

       O_NONBLOCK <\/strong>or O_NDELAY<\/strong>
              When possible, the file is opened in nonblocking mode.
              Neither the open<\/strong>() nor any subsequent I\/O operations on the
              file descriptor which is returned will cause the calling
              process to wait.

              Note that the setting of this flag has no effect on the
              operation of poll(2)<\/a>, select(2)<\/a>, epoll(7)<\/a>, and similar,
              since those interfaces merely inform the caller about
              whether a file descriptor is \"ready\", meaning that an I\/O
              operation performed on the file descriptor with the
              O_NONBLOCK <\/strong>flag clear<\/em> would not block.

              Note that this flag has no effect for regular files and
              block devices; that is, I\/O operations will (briefly) block
              when device activity is required, regardless of whether
              O_NONBLOCK <\/strong>is set.  Since O_NONBLOCK <\/strong>semantics might
              eventually be implemented, applications should not depend
              upon blocking behavior when specifying this flag for
              regular files and block devices.

              For the handling of FIFOs (named pipes), see also fifo(7)<\/a>.
              For a discussion of the effect of O_NONBLOCK <\/strong>in conjunction
              with mandatory file locks and with file leases, see
              fcntl(2)<\/a>.

       O_PATH <\/strong>(since Linux 2.6.39)
              Obtain a file descriptor that can be used for two purposes:
              to indicate a location in the filesystem tree and to
              perform operations that act purely at the file descriptor
              level.  The file itself is not opened, and other file
              operations (e.g., read(2)<\/a>, write(2)<\/a>, fchmod(2)<\/a>, fchown(2)<\/a>,
              fgetxattr(2)<\/a>, ioctl(2)<\/a>, mmap(2)<\/a>) fail with the error EBADF<\/strong>.

              The following operations can<\/em> be performed on the resulting
              file descriptor:

              \u2022  close(2)<\/a>.

              \u2022  fchdir(2)<\/a>, if the file descriptor refers to a directory
                 (since Linux 3.5).

              \u2022  fstat(2)<\/a> (since Linux 3.6).

              \u2022  fstatfs(2)<\/a> (since Linux 3.12).

              \u2022  Duplicating the file descriptor (dup(2)<\/a>, fcntl(2)<\/a>
                 F_DUPFD<\/strong>, etc.).

              \u2022  Getting and setting file descriptor flags (fcntl(2)<\/a>
                 F_GETFD <\/strong>and F_SETFD<\/strong>).

              \u2022  Retrieving open file status flags using the fcntl(2)<\/a>
                 F_GETFL <\/strong>operation: the returned flags will include the
                 bit O_PATH<\/strong>.

              \u2022  Passing the file descriptor as the dirfd<\/em> argument of
                 openat<\/strong>() and the other \"*at()\" system calls.  This
                 includes linkat(2)<\/a> with AT_EMPTY_PATH <\/strong>(or via procfs
                 using AT_SYMLINK_FOLLOW<\/strong>) even if the file is not a
                 directory.

              \u2022  Passing the file descriptor to another process via a
                 UNIX domain socket (see SCM_RIGHTS <\/strong>in unix(7)<\/a>).

              When O_PATH <\/strong>is specified in flags<\/em>, flag bits other than
              O_CLOEXEC<\/strong>, O_DIRECTORY<\/strong>, and O_NOFOLLOW <\/strong>are ignored.

              Opening a file or directory with the O_PATH <\/strong>flag requires
              no permissions on the object itself (but does require
              execute permission on the directories in the path prefix).
              Depending on the subsequent operation, a check for suitable
              file permissions may be performed (e.g., fchdir(2)<\/a> requires
              execute permission on the directory referred to by its file
              descriptor argument).  By contrast, obtaining a reference
              to a filesystem object by opening it with the O_RDONLY <\/strong>flag
              requires that the caller have read permission on the
              object, even when the subsequent operation (e.g.,
              fchdir(2)<\/a>, fstat(2)<\/a>) does not require read permission on
              the object.

              If pathname<\/em> is a symbolic link and the O_NOFOLLOW <\/strong>flag is
              also specified, then the call returns a file descriptor
              referring to the symbolic link.  This file descriptor can
              be used as the dirfd<\/em> argument in calls to fchownat(2)<\/a>,
              fstatat(2)<\/a>, linkat(2)<\/a>, and readlinkat(2)<\/a> with an empty
              pathname to have the calls operate on the symbolic link.

              If pathname<\/em> refers to an automount point that has not yet
              been triggered, so no other filesystem is mounted on it,
              then the call returns a file descriptor referring to the
              automount directory without triggering a mount.  fstatfs(2)<\/a>
              can then be used to determine if it is, in fact, an
              untriggered automount point (.f_type ==<\/strong>
              AUTOFS_SUPER_MAGIC<\/strong>).

              One use of O_PATH <\/strong>for regular files is to provide the
              equivalent of POSIX.1's O_EXEC <\/strong>functionality.  This permits
              us to open a file for which we have execute permission but
              not read permission, and then execute that file, with steps
              something like the following:

                  char buf[PATH_MAX];
                  fd = open(\"some_prog\", O_PATH);
                  snprintf(buf, PATH_MAX, \"\/proc\/self\/fd\/%d\", fd);
                  execl(buf, \"some_prog\", (char *) NULL);

              An O_PATH <\/strong>file descriptor can also be passed as the
              argument of fexecve(3)<\/a>.

       O_SYNC <\/strong>Write operations on the file will complete according to the
              requirements of synchronized I\/O file<\/em> integrity completion
              (by contrast with the synchronized I\/O data<\/em> integrity
              completion provided by O_DSYNC<\/strong>.)

              By the time write(2)<\/a> (or similar) returns, the output data
              and associated file metadata have been transferred to the
              underlying hardware (i.e., as though each write(2)<\/a> was
              followed by a call to fsync(2)<\/a>).  See VERSIONS.

       O_TMPFILE <\/strong>(since Linux 3.11)
              Create an unnamed temporary regular file.  The pathname<\/em>
              argument specifies a directory; an unnamed inode will be
              created in that directory's filesystem.  Anything written
              to the resulting file will be lost when the last file
              descriptor is closed, unless the file is given a name.

              O_TMPFILE <\/strong>must be specified with one of O_RDWR <\/strong>or O_WRONLY<\/strong>
              and, optionally, O_EXCL<\/strong>.  If O_EXCL <\/strong>is not specified, then
              linkat(2)<\/a> can be used to link the temporary file into the
              filesystem, making it permanent, using code like the
              following:

                  char path[PATH_MAX];
                  fd = open(\"\/path\/to\/dir\", O_TMPFILE | O_RDWR,
                                          S_IRUSR | S_IWUSR);

                  \/* File I\/O on 'fd'... *\/

                  linkat(fd, \"\", AT_FDCWD, \"\/path\/for\/file\", AT_EMPTY_PATH);

                  \/* If the caller doesn't have the CAP_DAC_READ_SEARCH
                     capability (needed to use AT_EMPTY_PATH with linkat(2)),
                     and there is a proc(5) filesystem mounted, then the
                     linkat(2) call above can be replaced with:

                  snprintf(path, PATH_MAX,  \"\/proc\/self\/fd\/%d\", fd);
                  linkat(AT_FDCWD, path, AT_FDCWD, \"\/path\/for\/file\",
                                          AT_SYMLINK_FOLLOW);
                  *\/

              In this case, the open<\/strong>() mode<\/em> argument determines the file
              permission mode, as with O_CREAT<\/strong>.

              Specifying O_EXCL <\/strong>in conjunction with O_TMPFILE <\/strong>prevents a
              temporary file from being linked into the filesystem in the
              above manner.  (Note that the meaning of O_EXCL <\/strong>in this
              case is different from the meaning of O_EXCL <\/strong>otherwise.)

              There are two main use cases for O_TMPFILE<\/strong>:

              \u2022  Improved tmpfile(3)<\/a> functionality: race-free creation of
                 temporary files that (1) are automatically deleted when
                 closed; (2) can never be reached via any pathname; (3)
                 are not subject to symlink attacks; and (4) do not
                 require the caller to devise unique names.

              \u2022  Creating a file that is initially invisible, which is
                 then populated with data and adjusted to have
                 appropriate filesystem attributes (fchown(2)<\/a>, fchmod(2)<\/a>,
                 fsetxattr(2)<\/a>, etc.)  before being atomically linked into
                 the filesystem in a fully formed state (using linkat(2)<\/a>
                 as described above).

              O_TMPFILE <\/strong>requires support by the underlying filesystem;
              only a subset of Linux filesystems provide that support.
              In the initial implementation, support was provided in the
              ext2, ext3, ext4, UDF, Minix, and tmpfs filesystems.
              Support for other filesystems has subsequently been added
              as follows: XFS (Linux 3.15); Btrfs (Linux 3.16); F2FS
              (Linux 3.16); and ubifs (Linux 4.9)

       O_TRUNC<\/strong>
              If the file already exists and is a regular file and the
              access mode allows writing (i.e., is O_RDWR <\/strong>or O_WRONLY<\/strong>) it
              will be truncated to length 0.  If the file is a FIFO or
              terminal device file, the O_TRUNC <\/strong>flag is ignored.
              Otherwise, the effect of O_TRUNC <\/strong>is unspecified.

   creat()<\/strong>
       A call to creat<\/strong>() is equivalent to calling open<\/strong>() with flags<\/em> equal
       to O_CREAT|O_WRONLY|O_TRUNC<\/strong>.

   openat()<\/strong>
       The openat<\/strong>() system call operates in exactly the same way as
       open<\/strong>(), except for the differences described here.

       The dirfd<\/em> argument is used in conjunction with the pathname<\/em>
       argument as follows:

       \u2022  If the pathname given in pathname<\/em> is absolute, then dirfd<\/em> is
          ignored.

       \u2022  If the pathname given in pathname<\/em> is relative and dirfd<\/em> is the
          special value AT_FDCWD<\/strong>, then pathname<\/em> is interpreted relative
          to the current working directory of the calling process (like
          open<\/strong>()).

       \u2022  If the pathname given in pathname<\/em> is relative, then it is
          interpreted relative to the directory referred to by the file
          descriptor dirfd<\/em> (rather than relative to the current working
          directory of the calling process, as is done by open<\/strong>() for a
          relative pathname).  In this case, dirfd<\/em> must be a directory
          that was opened for reading (O_RDONLY<\/strong>) or using the O_PATH<\/strong>
          flag.

       If the pathname given in pathname<\/em> is relative, and dirfd<\/em> is not a
       valid file descriptor, an error (EBADF<\/strong>) results.  (Specifying an
       invalid file descriptor number in dirfd<\/em> can be used as a means to
       ensure that pathname<\/em> is absolute.)

   openat2(2)<\/strong>
       The openat2(2)<\/a> system call is an extension of openat<\/strong>(), and
       provides a superset of the features of openat<\/strong>().  It is documented
       separately, in openat2(2)<\/a>.
<\/pre>\n\n\n\n
<\/a>RETURN VALUE         top<\/a><\/h2>\n\n\n\n
       On success, open<\/strong>(), openat<\/strong>(), and creat<\/strong>() return the new file
       descriptor (a nonnegative integer).  On error, -1 is returned and
       errno<\/a><\/em> is set to indicate the error.
<\/pre>\n\n\n\n
<\/a>ERRORS         top<\/a><\/h2>\n\n\n\n
       open<\/strong>(), openat<\/strong>(), and creat<\/strong>() can fail with the following errors:

       EACCES <\/strong>The requested access to the file is not allowed, or search
              permission is denied for one of the directories in the path
              prefix of pathname<\/em>, or the file did not exist yet and write
              access to the parent directory is not allowed.  (See also
              path_resolution(7)<\/a>.)

       EACCES <\/strong>Where O_CREAT <\/strong>is specified, the protected_fifos<\/em> or
              protected_regular<\/em> sysctl is enabled, the file already
              exists and is a FIFO or regular file, the owner of the file
              is neither the current user nor the owner of the containing
              directory, and the containing directory is both world- or
              group-writable and sticky.  For details, see the
              descriptions of \/proc\/sys\/fs\/protected_fifos<\/em> and
              \/proc\/sys\/fs\/protected_regular<\/em> in proc_sys_fs(5)<\/a>.

       EBADF  <\/strong>(openat<\/strong>()) pathname<\/em> is relative but dirfd<\/em> is neither
              AT_FDCWD <\/strong>nor a valid file descriptor.

       EBUSY  O_EXCL <\/strong>was specified in flags<\/em> and pathname<\/em> refers to a
              block device that is in use by the system (e.g., it is
              mounted).

       EDQUOT <\/strong>Where O_CREAT <\/strong>is specified, the file does not exist, and
              the user's quota of disk blocks or inodes on the filesystem
              has been exhausted.

       EEXIST <\/strong>pathname<\/em> already exists and O_CREAT <\/strong>and O_EXCL <\/strong>were used.

       EFAULT <\/strong>pathname<\/em> points outside your accessible address space.

       EFBIG  <\/strong>See EOVERFLOW<\/strong>.

       EINTR  <\/strong>While blocked waiting to complete an open of a slow device
              (e.g., a FIFO; see fifo(7)<\/a>), the call was interrupted by a
              signal handler; see signal(7)<\/a>.

       EINVAL <\/strong>The filesystem does not support the O_DIRECT <\/strong>flag.  See
              NOTES <\/strong>for more information.

       EINVAL <\/strong>Invalid value in flags<\/em>.

       EINVAL O_TMPFILE <\/strong>was specified in flags<\/em>, but neither O_WRONLY <\/strong>nor
              O_RDWR <\/strong>was specified.

       EINVAL O_CREAT <\/strong>was specified in flags<\/em> and the final component
              (\"basename\") of the new file's pathname<\/em> is invalid (e.g.,
              it contains characters not permitted by the underlying
              filesystem).

       EINVAL <\/strong>The final component (\"basename\") of pathname<\/em> is invalid
              (e.g., it contains characters not permitted by the
              underlying filesystem).

       EISDIR <\/strong>pathname<\/em> refers to a directory and the access requested
              involved writing (that is, O_WRONLY <\/strong>or O_RDWR <\/strong>is set).

       EISDIR <\/strong>pathname<\/em> refers to an existing directory, O_TMPFILE <\/strong>and one
              of O_WRONLY <\/strong>or O_RDWR <\/strong>were specified in flags<\/em>, but this
              kernel version does not provide the O_TMPFILE<\/strong>
              functionality.

       ELOOP  <\/strong>Too many symbolic links were encountered in resolving
              pathname<\/em>.

       ELOOP  <\/strong>pathname<\/em> was a symbolic link, and flags<\/em> specified
              O_NOFOLLOW <\/strong>but not O_PATH<\/strong>.

       EMFILE <\/strong>The per-process limit on the number of open file
              descriptors has been reached (see the description of
              RLIMIT_NOFILE <\/strong>in getrlimit(2)<\/a>).

       ENAMETOOLONG<\/strong>
              pathname<\/em> was too long.

       ENFILE <\/strong>The system-wide limit on the total number of open files has
              been reached.

       ENODEV <\/strong>pathname<\/em> refers to a device special file and no
              corresponding device exists.  (This is a Linux kernel bug;
              in this situation ENXIO <\/strong>must be returned.)

       ENOENT O_CREAT <\/strong>is not set and the named file does not exist.

       ENOENT <\/strong>A directory component in pathname<\/em> does not exist or is a
              dangling symbolic link.

       ENOENT <\/strong>pathname<\/em> refers to a nonexistent directory, O_TMPFILE <\/strong>and
              one of O_WRONLY <\/strong>or O_RDWR <\/strong>were specified in flags<\/em>, but this
              kernel version does not provide the O_TMPFILE<\/strong>
              functionality.

       ENOMEM <\/strong>The named file is a FIFO, but memory for the FIFO buffer
              can't be allocated because the per-user hard limit on
              memory allocation for pipes has been reached and the caller
              is not privileged; see pipe(7)<\/a>.

       ENOMEM <\/strong>Insufficient kernel memory was available.

       ENOSPC <\/strong>pathname<\/em> was to be created but the device containing
              pathname<\/em> has no room for the new file.

       ENOTDIR<\/strong>
              A component used as a directory in pathname<\/em> is not, in
              fact, a directory, or O_DIRECTORY <\/strong>was specified and
              pathname<\/em> was not a directory.

       ENOTDIR<\/strong>
              (openat<\/strong>()) pathname<\/em> is a relative pathname and dirfd<\/em> is a
              file descriptor referring to a file other than a directory.

       ENXIO  O_NONBLOCK <\/strong>| O_WRONLY <\/strong>is set, the named file is a FIFO, and
              no process has the FIFO open for reading.

       ENXIO  <\/strong>The file is a device special file and no corresponding
              device exists.

       ENXIO  <\/strong>The file is a UNIX domain socket.

       EOPNOTSUPP<\/strong>
              The filesystem containing pathname<\/em> does not support
              O_TMPFILE<\/strong>.

       EOVERFLOW<\/strong>
              pathname<\/em> refers to a regular file that is too large to be
              opened.  The usual scenario here is that an application
              compiled on a 32-bit platform without
              -D_FILE_OFFSET_BITS=64<\/em> tried to open a file whose size
              exceeds (1<<31)-1<\/em> bytes; see also O_LARGEFILE <\/strong>above.  This
              is the error specified by POSIX.1; before Linux 2.6.24,
              Linux gave the error EFBIG <\/strong>for this case.

       EPERM  <\/strong>The O_NOATIME <\/strong>flag was specified, but the effective user ID
              of the caller did not match the owner of the file and the
              caller was not privileged.

       EPERM  <\/strong>The operation was prevented by a file seal; see fcntl(2)<\/a>.

       EROFS  <\/strong>pathname<\/em> refers to a file on a read-only filesystem and
              write access was requested.

       ETXTBSY<\/strong>
              pathname<\/em> refers to an executable image which is currently
              being executed and write access was requested.

       ETXTBSY<\/strong>
              pathname<\/em> refers to a file that is currently in use as a
              swap file, and the O_TRUNC <\/strong>flag was specified.

       ETXTBSY<\/strong>
              pathname<\/em> refers to a file that is currently being read by
              the kernel (e.g., for module\/firmware loading), and write
              access was requested.

       EWOULDBLOCK<\/strong>
              The O_NONBLOCK <\/strong>flag was specified, and an incompatible
              lease was held on the file (see fcntl(2)<\/a>).
<\/pre>\n\n\n\n
<\/a>VERSIONS         top<\/a><\/h2>\n\n\n\n
       The (undefined) effect of O_RDONLY | O_TRUNC <\/strong>varies among
       implementations.  On many systems the file is actually truncated.

   Synchronized I\/O<\/strong>
       The POSIX.1-2008 \"synchronized I\/O\" option specifies different
       variants of synchronized I\/O, and specifies the open<\/strong>() flags
       O_SYNC<\/strong>, O_DSYNC<\/strong>, and O_RSYNC <\/strong>for controlling the behavior.
       Regardless of whether an implementation supports this option, it
       must at least support the use of O_SYNC <\/strong>for regular files.

       Linux implements O_SYNC <\/strong>and O_DSYNC<\/strong>, but not O_RSYNC<\/strong>.  Somewhat
       incorrectly, glibc defines O_RSYNC <\/strong>to have the same value as
       O_SYNC<\/strong>.  (O_RSYNC <\/strong>is defined in the Linux header file
       <asm\/fcntl.h><\/em> on HP PA-RISC, but it is not used.)

       O_SYNC <\/strong>provides synchronized I\/O file<\/em> integrity completion,
       meaning write operations will flush data and all associated
       metadata to the underlying hardware.  O_DSYNC <\/strong>provides
       synchronized I\/O data<\/em> integrity completion, meaning write
       operations will flush data to the underlying hardware, but will
       only flush metadata updates that are required to allow a
       subsequent read operation to complete successfully.  Data
       integrity completion can reduce the number of disk operations that
       are required for applications that don't need the guarantees of
       file integrity completion.

       To understand the difference between the two types of completion,
       consider two pieces of file metadata: the file last modification
       timestamp (st_mtime<\/em>) and the file length.  All write operations
       will update the last file modification timestamp, but only writes
       that add data to the end of the file will change the file length.
       The last modification timestamp is not needed to ensure that a
       read completes successfully, but the file length is.  Thus,
       O_DSYNC <\/strong>would only guarantee to flush updates to the file length
       metadata (whereas O_SYNC <\/strong>would also always flush the last
       modification timestamp metadata).

       Before Linux 2.6.33, Linux implemented only the O_SYNC <\/strong>flag for
       open<\/strong>().  However, when that flag was specified, most filesystems
       actually provided the equivalent of synchronized I\/O data<\/em>
       integrity completion (i.e., O_SYNC <\/strong>was actually implemented as the
       equivalent of O_DSYNC<\/strong>).

       Since Linux 2.6.33, proper O_SYNC <\/strong>support is provided.  However,
       to ensure backward binary compatibility, O_DSYNC <\/strong>was defined with
       the same value as the historical O_SYNC<\/strong>, and O_SYNC <\/strong>was defined as
       a new (two-bit) flag value that includes the O_DSYNC <\/strong>flag value.
       This ensures that applications compiled against new headers get at
       least O_DSYNC <\/strong>semantics before Linux 2.6.33.

   C library\/kernel differences<\/strong>
       Since glibc 2.26, the glibc wrapper function for open<\/strong>() employs
       the openat<\/strong>() system call, rather than the kernel's open<\/strong>() system
       call.  For certain architectures, this is also true before glibc
       2.26.
<\/pre>\n\n\n\n
<\/a>STANDARDS         top<\/a><\/h2>\n\n\n\n
       open<\/strong>()
       creat<\/strong>()
       openat<\/strong>()
              POSIX.1-2008.

       openat2(2)<\/a> Linux.

       The O_DIRECT<\/strong>, O_NOATIME<\/strong>, O_PATH<\/strong>, and O_TMPFILE <\/strong>flags are Linux-
       specific.  One must define _GNU_SOURCE <\/strong>to obtain their
       definitions.

       The O_CLOEXEC<\/strong>, O_DIRECTORY<\/strong>, and O_NOFOLLOW <\/strong>flags are not specified
       in POSIX.1-2001, but are specified in POSIX.1-2008.  Since glibc
       2.12, one can obtain their definitions by defining either
       _POSIX_C_SOURCE <\/strong>with a value greater than or equal to 200809L or
       _XOPEN_SOURCE <\/strong>with a value greater than or equal to 700.  In glibc
       2.11 and earlier, one obtains the definitions by defining
       _GNU_SOURCE<\/strong>.
<\/pre>\n\n\n\n
<\/a>HISTORY         top<\/a><\/h2>\n\n\n\n
       open<\/strong>()
       creat<\/strong>()
              SVr4, 4.3BSD, POSIX.1-2001.

       openat<\/strong>()
              POSIX.1-2008.  Linux 2.6.16, glibc 2.4.
<\/pre>\n\n\n\n
<\/a>NOTES         top<\/a><\/h2>\n\n\n\n
       Under Linux, the O_NONBLOCK <\/strong>flag is sometimes used in cases where
       one wants to open but does not necessarily have the intention to
       read or write.  For example, this may be used to open a device in
       order to get a file descriptor for use with ioctl(2)<\/a>.

       Note that open<\/strong>() can open device special files, but creat<\/strong>() cannot
       create them; use mknod(2)<\/a> instead.

       If the file is newly created, its st_atime<\/em>, st_ctime<\/em>, st_mtime<\/em>
       fields (respectively, time of last access, time of last status
       change, and time of last modification; see stat(2)<\/a>) are set to the
       current time, and so are the st_ctime<\/em> and st_mtime<\/em> fields of the
       parent directory.  Otherwise, if the file is modified because of
       the O_TRUNC <\/strong>flag, its st_ctime<\/em> and st_mtime<\/em> fields are set to the
       current time.

       The files in the \/proc\/<\/em>pid\/fd<\/em> directory show the open file
       descriptors of the process with the PID pid<\/em>.  The files in the
       \/proc\/<\/em>pid\/fdinfo<\/em> directory show even more information about these
       file descriptors.  See proc(5)<\/a> for further details of both of
       these directories.

       The Linux header file <asm\/fcntl.h> <\/strong>doesn't define O_ASYNC<\/strong>; the
       (BSD-derived) FASYNC <\/strong>synonym is defined instead.

   Open file descriptions<\/strong>
       The term open file description is the one used by POSIX to refer
       to the entries in the system-wide table of open files.  In other
       contexts, this object is variously also called an \"open file
       object\", a \"file handle\", an \"open file table entry\", or\u2014in
       kernel-developer parlance\u2014a struct file<\/em>.

       When a file descriptor is duplicated (using dup(2)<\/a> or similar),
       the duplicate refers to the same open file description as the
       original file descriptor, and the two file descriptors
       consequently share the file offset and file status flags.  Such
       sharing can also occur between processes: a child process created
       via fork(2)<\/a> inherits duplicates of its parent's file descriptors,
       and those duplicates refer to the same open file descriptions.

       Each open<\/strong>() of a file creates a new open file description; thus,
       there may be multiple open file descriptions corresponding to a
       file inode.

       On Linux, one can use the kcmp(2)<\/a> KCMP_FILE <\/strong>operation to test
       whether two file descriptors (in the same process or in two
       different processes) refer to the same open file description.

   NFS<\/strong>
       There are many infelicities in the protocol underlying NFS,
       affecting amongst others O_SYNC <\/strong>and O_NDELAY<\/strong>.

       On NFS filesystems with UID mapping enabled, open<\/strong>() may return a
       file descriptor but, for example, read(2)<\/a> requests are denied with
       EACCES<\/strong>.  This is because the client performs open<\/strong>() by checking
       the permissions, but UID mapping is performed by the server upon
       read and write requests.

   FIFOs<\/strong>
       Opening the read or write end of a FIFO blocks until the other end
       is also opened (by another process or thread).  See fifo(7)<\/a> for
       further details.

   File access mode<\/strong>
       Unlike the other values that can be specified in flags<\/em>, the access<\/em>
       mode<\/em> values O_RDONLY<\/strong>, O_WRONLY<\/strong>, and O_RDWR <\/strong>do not specify
       individual bits.  Rather, they define the low order two bits of
       flags<\/em>, and are defined respectively as 0, 1, and 2.  In other
       words, the combination O_RDONLY | O_WRONLY <\/strong>is a logical error, and
       certainly does not have the same meaning as O_RDWR<\/strong>.

       Linux reserves the special, nonstandard access mode 3 (binary 11)
       in flags<\/em> to mean: check for read and write permission on the file
       and return a file descriptor that can't be used for reading or
       writing.  This nonstandard access mode is used by some Linux
       drivers to return a file descriptor that is to be used only for
       device-specific ioctl(2)<\/a> operations.

   Rationale for openat() and other directory file descriptor APIs<\/strong>
       openat<\/strong>() and the other system calls and library functions that
       take a directory file descriptor argument (i.e., execveat(2)<\/a>,
       faccessat(2)<\/a>, fanotify_mark(2)<\/a>, fchmodat(2)<\/a>, fchownat(2)<\/a>,
       fspick<\/strong>(2), fstatat(2)<\/a>, futimesat(2)<\/a>, linkat(2)<\/a>, mkdirat(2)<\/a>,
       mknodat(2)<\/a>, mount_setattr(2)<\/a>, move_mount<\/strong>(2), name_to_handle_at(2)<\/a>,
       open_tree<\/strong>(2), openat2(2)<\/a>, readlinkat(2)<\/a>, renameat(2)<\/a>,
       renameat2(2)<\/a>, statx(2)<\/a>, symlinkat(2)<\/a>, unlinkat(2)<\/a>, utimensat(2)<\/a>,
       mkfifoat(3)<\/a>, and scandirat(3)<\/a>) address two problems with the older
       interfaces that preceded them.  Here, the explanation is in terms
       of the openat<\/strong>() call, but the rationale is analogous for the other
       interfaces.

       First, openat<\/strong>() allows an application to avoid race conditions
       that could occur when using open<\/strong>() to open files in directories
       other than the current working directory.  These race conditions
       result from the fact that some component of the directory prefix
       given to open<\/strong>() could be changed in parallel with the call to
       open<\/strong>().  Suppose, for example, that we wish to create the file
       dir1\/dir2\/xxx.dep<\/em> if the file dir1\/dir2\/xxx<\/em> exists.  The problem
       is that between the existence check and the file-creation step,
       dir1<\/em> or dir2<\/em> (which might be symbolic links) could be modified to
       point to a different location.  Such races can be avoided by
       opening a file descriptor for the target directory, and then
       specifying that file descriptor as the dirfd<\/em> argument of (say)
       fstatat(2)<\/a> and openat<\/strong>().  The use of the dirfd<\/em> file descriptor
       also has other benefits:

       \u2022  the file descriptor is a stable reference to the directory,
          even if the directory is renamed; and

       \u2022  the open file descriptor prevents the underlying filesystem
          from being dismounted, just as when a process has a current
          working directory on a filesystem.

       Second, openat<\/strong>() allows the implementation of a per-thread
       \"current working directory\", via file descriptor(s) maintained by
       the application.  (This functionality can also be obtained by
       tricks based on the use of \/proc\/self\/fd\/<\/em>dirfd, but less
       efficiently.)

       The dirfd<\/em> argument for these APIs can be obtained by using open<\/strong>()
       or openat<\/strong>() to open a directory (with either the O_RDONLY <\/strong>or the
       O_PATH <\/strong>flag).  Alternatively, such a file descriptor can be
       obtained by applying dirfd(3)<\/a> to a directory stream created using
       opendir(3)<\/a>.

       When these APIs are given a dirfd<\/em> argument of AT_FDCWD <\/strong>or the
       specified pathname is absolute, then they handle their pathname
       argument in the same way as the corresponding conventional APIs.
       However, in this case, several of the APIs have a flags<\/em> argument
       that provides access to functionality that is not available with
       the corresponding conventional APIs.

   O_DIRECT<\/strong>
       The O_DIRECT <\/strong>flag may impose alignment restrictions on the length
       and address of user-space buffers and the file offset of I\/Os.  In
       Linux alignment restrictions vary by filesystem and kernel version
       and might be absent entirely.  The handling of misaligned O_DIRECT<\/strong>
       I\/Os also varies; they can either fail with EINVAL <\/strong>or fall back to
       buffered I\/O.

       Since Linux 6.1, O_DIRECT <\/strong>support and alignment restrictions for a
       file can be queried using statx(2)<\/a>, using the STATX_DIOALIGN <\/strong>flag.
       Support for STATX_DIOALIGN <\/strong>varies by filesystem; see statx(2)<\/a>.

       Some filesystems provide their own interfaces for querying
       O_DIRECT <\/strong>alignment restrictions, for example the XFS_IOC_DIOINFO<\/strong>
       operation in xfsctl(3)<\/a>.  STATX_DIOALIGN <\/strong>should be used instead
       when it is available.

       If none of the above is available, then direct I\/O support and
       alignment restrictions can only be assumed from known
       characteristics of the filesystem, the individual file, the
       underlying storage device(s), and the kernel version.  In Linux
       2.4, most filesystems based on block devices require that the file
       offset and the length and memory address of all I\/O segments be
       multiples of the filesystem block size (typically 4096 bytes).  In
       Linux 2.6.0, this was relaxed to the logical block size of the
       block device (typically 512 bytes).  A block device's logical
       block size can be determined using the ioctl(2)<\/a> BLKSSZGET<\/strong>
       operation or from the shell using the command:

           blockdev --getss

       O_DIRECT <\/strong>I\/Os should never be run concurrently with the fork(2)<\/a>
       system call, if the memory buffer is a private mapping (i.e., any
       mapping created with the mmap(2)<\/a> MAP_PRIVATE <\/strong>flag; this includes
       memory allocated on the heap and statically allocated buffers).
       Any such I\/Os, whether submitted via an asynchronous I\/O interface
       or from another thread in the process, should be completed before
       fork(2)<\/a> is called.  Failure to do so can result in data corruption
       and undefined behavior in parent and child processes.  This
       restriction does not apply when the memory buffer for the O_DIRECT<\/strong>
       I\/Os was created using shmat(2)<\/a> or mmap(2)<\/a> with the MAP_SHARED<\/strong>
       flag.  Nor does this restriction apply when the memory buffer has
       been advised as MADV_DONTFORK <\/strong>with madvise(2)<\/a>, ensuring that it
       will not be available to the child after fork(2)<\/a>.

       The O_DIRECT <\/strong>flag was introduced in SGI IRIX, where it has
       alignment restrictions similar to those of Linux 2.4.  IRIX has
       also a fcntl(2)<\/a> call to query appropriate alignments, and sizes.
       FreeBSD 4.x introduced a flag of the same name, but without
       alignment restrictions.

       O_DIRECT <\/strong>support was added in Linux 2.4.10.  Older Linux kernels
       simply ignore this flag.  Some filesystems may not implement the
       flag, in which case open<\/strong>() fails with the error EINVAL <\/strong>if it is
       used.

       Applications should avoid mixing O_DIRECT <\/strong>and normal I\/O to the
       same file, and especially to overlapping byte regions in the same
       file.  Even when the filesystem correctly handles the coherency
       issues in this situation, overall I\/O throughput is likely to be
       slower than using either mode alone.  Likewise, applications
       should avoid mixing mmap(2)<\/a> of files with direct I\/O to the same
       files.

       The behavior of O_DIRECT <\/strong>with NFS will differ from local
       filesystems.  Older kernels, or kernels configured in certain
       ways, may not support this combination.  The NFS protocol does not
       support passing the flag to the server, so O_DIRECT <\/strong>I\/O will
       bypass the page cache only on the client; the server may still
       cache the I\/O.  The client asks the server to make the I\/O
       synchronous to preserve the synchronous semantics of O_DIRECT<\/strong>.
       Some servers will perform poorly under these circumstances,
       especially if the I\/O size is small.  Some servers may also be
       configured to lie to clients about the I\/O having reached stable
       storage; this will avoid the performance penalty at some risk to
       data integrity in the event of server power failure.  The Linux
       NFS client places no alignment restrictions on O_DIRECT <\/strong>I\/O.

       In summary, O_DIRECT <\/strong>is a potentially powerful tool that should be
       used with caution.  It is recommended that applications treat use
       of O_DIRECT <\/strong>as a performance option which is disabled by default.
<\/pre>\n\n\n\n
<\/a>BUGS         top<\/a><\/h2>\n\n\n\n
       Currently, it is not possible to enable signal-driven I\/O by
       specifying O_ASYNC <\/strong>when calling open<\/strong>(); use fcntl(2)<\/a> to enable
       this flag.

       One must check for two different error codes, EISDIR <\/strong>and ENOENT<\/strong>,
       when trying to determine whether the kernel supports O_TMPFILE<\/strong>
       functionality.

       When both O_CREAT <\/strong>and O_DIRECTORY <\/strong>are specified in flags<\/em> and the
       file specified by pathname<\/em> does not exist, open<\/strong>() will create a
       regular file (i.e., O_DIRECTORY <\/strong>is ignored).
<\/pre>\n\n\n\n
<\/a>SEE ALSO         top<\/a><\/h2>\n\n\n\n
       chmod(2)<\/a>, chown(2)<\/a>, close(2)<\/a>, dup(2)<\/a>, fcntl(2)<\/a>, link(2)<\/a>, lseek(2)<\/a>,
       mknod(2)<\/a>, mmap(2)<\/a>, mount(2)<\/a>, open_by_handle_at(2)<\/a>, openat2(2)<\/a>,
       read(2)<\/a>, socket(2)<\/a>, stat(2)<\/a>, umask(2)<\/a>, unlink(2)<\/a>, write(2)<\/a>,
       fopen(3)<\/a>, acl(5)<\/a>, fifo(7)<\/a>, inode(7)<\/a>, path_resolution(7)<\/a>,
       symlink(7)<\/a><\/p>\n<\/details>\n\n\n\n
Ecriture : <\/h1>\n\n\n\n
Exemple simple de cr\u00e9ation d’un fichier texte en lecture \u00e9criture uniquement par le propri\u00e9taire<\/p>\n\n\n\n
\/* exemple simple avec write *\/\n\n#include <stdlib.h>\n#include <stdio.h>\n#include <sys\/stat.h>\n#include <fcntl.h>\n#include <unistd.h>\n\nint main (int argc, char ** argv)\n{\nint fd; \/* file desciptor *\/\nfd = open ( \"fichier.txt\",O_CREAT | O_RDWR \n                         ,S_IRUSR | S_IWUSR );  \/* droits Read USER Write USER...*\/\n\/* aller voir man 3 open *\/\nif (fd==-1)     \n              {\n              printf(\"Erreur d'ouverture du fichier \\n\");\n              return EXIT_FAILURE;\n              }\nwrite (fd, \"Bonjour le monde\\n\",17);  \/* 17 etant le nombre de caractere de la chaine *\/\nclose(fd);\nreturn EXIT_SUCCESS;\n\n}<\/pre>\n\n\n\n
Ecrire un fichier en compliquant un peu <\/p>\n\n\n\n
exemple 2 d\u2019\u00e9criture <\/p>\n\n\n\n
\/* exemple 2 avec une chaine *\/\n#include <stdio.h>\n#include <stdlib.h>\n#include <termios.h>\n#include <unistd.h>\n#include <sys\/types.h>\n#include <sys\/stat.h>\n#include <fcntl.h>\n\nint main()\n{\n   int fd;\n   char chaine[100]=\"Bonjour le monda !\";\n   fd=open (\"\/tmp\/bonjour.txt\",O_CREAT | O_WRONLY,\n                        S_IRUSR | S_IWUSR | S_IRGRP | S_IROTH );\n   if (fd==-1)         {\n                       perror(\"Pas possible d'ouvrir ce fichier \");\n                       return EXIT_FAILURE;\n                       }\n   write(fd,chaine,16);\n   printf(\"%s\",chaine);\n   close (fd);\n   return EXIT_SUCCESS;\n}<\/pre>\n\n\n\n
Lire :<\/h2>\n\n\n\n
\/*  Lecture de fichier *\/\n#include <stdio.h>\n#include <stdlib.h>\n#include <termios.h>\n#include <unistd.h>\n#include <sys\/types.h>\n#include <sys\/stat.h>\n#include <fcntl.h>\n\nint main()\n{\n   int fd;\n   char chaine[100];\n   fd=open (\"fichier.txt\",O_CREAT | O_RDONLY,\n                        S_IRUSR | S_IWUSR | S_IRGRP | S_IROTH );\n   if (fd==-1)         {\n                       perror(\"Pas possible d'ouvrir ce fichier \");\n                       return EXIT_FAILURE;\n                       }\n   read(fd,chaine,16);\n   printf(\"%s\",chaine);\n   close (fd);\n   printf(\"Hello World!\\n\");\n   return EXIT_SUCCESS;\n}<\/pre>\n\n\n\n
Les flags!<\/h2>\n\n\n\n
Voici le d\u00e9tail des constantes de droits (issues de <sys\/stat.h><\/code>) :<\/p>\n\n\n\n
Constante<\/th>Valeur octale<\/th>Signification<\/th><\/tr><\/thead>
S_IRUSR<\/code><\/strong><\/td>0400<\/td>Lecture autoris\u00e9e pour le propri\u00e9taire<\/strong> (User)<\/td><\/tr>
S_IWUSR<\/code><\/strong><\/td>0200<\/td>\u00c9criture autoris\u00e9e pour le propri\u00e9taire<\/strong> (User)<\/td><\/tr>
S_IXUSR<\/code><\/strong><\/td>0100<\/td>Ex\u00e9cution autoris\u00e9e pour le propri\u00e9taire<\/td><\/tr>
S_IRGRP<\/code><\/td>0040<\/td>Lecture pour le groupe<\/strong><\/td><\/tr>
S_IWGRP<\/code><\/td>0020<\/td>\u00c9criture pour le groupe<\/td><\/tr>
S_IROTH<\/code><\/td>0004<\/td>Lecture pour les autres<\/strong><\/td><\/tr>
S_IWOTH<\/code><\/td>0002<\/td>\u00c9criture pour les autres<\/td><\/tr><\/tbody><\/table><\/figure>\n\n\n\n
Combinaison courante<\/th>Signification<\/th><\/tr><\/thead>
O_RDONLY<\/code><\/td>Lecture seule (fichier doit exister)<\/td><\/tr>
O_WRONLY | O_CREAT | O_TRUNC<\/code><\/td>\u00c9criture seule, cr\u00e9e ou \u00e9crase le fichier.<\/td><\/tr>
O_RDWR | O_CREAT | O_APPEND<\/code><\/td>Lecture\/\u00c9criture, cr\u00e9e si absent, \u00e9crit \u00e0 la fin.<\/td><\/tr>
O_RDWR | O_CREAT | O_EXCL<\/code><\/td>Cr\u00e9ation atomique, \u00e9choue si le fichier existe (utile pour les verrous).<\/td><\/tr><\/tbody><\/table><\/figure>\n","protected":false},"excerpt":{"rendered":"
Tout est fichiers : dans Unix et Linux, presque tout est consid\u00e9r\u00e9 comme un fichier. Ce n’est pas une simple m\u00e9taphore, c’est une conception profonde du syst\u00e8me qui le rend extr\u00eamement puissant et simple \u00e0 la fois. Voici ce que cela signifie concr\u00e8tement : Ces 2 fonctions sont intrins\u00e8quement li\u00e9es Si on ouvre un fichier […]<\/p>\n","protected":false},"author":1,"featured_media":0,"parent":0,"menu_order":0,"comment_status":"closed","ping_status":"closed","template":"","meta":{"_uag_custom_page_level_css":"","footnotes":""},"class_list":["post-3852","page","type-page","status-publish","hentry"],"uagb_featured_image_src":{"full":false,"thumbnail":false,"medium":false,"medium_large":false,"large":false,"1536x1536":false,"2048x2048":false},"uagb_author_info":{"display_name":"admin","author_link":"https:\/\/workboot.fr\/ciela\/author\/admin\/"},"uagb_comment_info":0,"uagb_excerpt":"Tout est fichiers : dans Unix et Linux, presque tout est consid\u00e9r\u00e9 comme un fichier. Ce n’est pas une simple m\u00e9taphore, c’est une conception profonde du syst\u00e8me qui le rend extr\u00eamement puissant et simple \u00e0 la fois. Voici ce que cela signifie concr\u00e8tement : Ces 2 fonctions sont intrins\u00e8quement li\u00e9es Si on ouvre un fichier\u2026","_links":{"self":[{"href":"https:\/\/workboot.fr\/ciela\/wp-json\/wp\/v2\/pages\/3852","targetHints":{"allow":["GET"]}}],"collection":[{"href":"https:\/\/workboot.fr\/ciela\/wp-json\/wp\/v2\/pages"}],"about":[{"href":"https:\/\/workboot.fr\/ciela\/wp-json\/wp\/v2\/types\/page"}],"author":[{"embeddable":true,"href":"https:\/\/workboot.fr\/ciela\/wp-json\/wp\/v2\/users\/1"}],"replies":[{"embeddable":true,"href":"https:\/\/workboot.fr\/ciela\/wp-json\/wp\/v2\/comments?post=3852"}],"version-history":[{"count":14,"href":"https:\/\/workboot.fr\/ciela\/wp-json\/wp\/v2\/pages\/3852\/revisions"}],"predecessor-version":[{"id":8311,"href":"https:\/\/workboot.fr\/ciela\/wp-json\/wp\/v2\/pages\/3852\/revisions\/8311"}],"wp:attachment":[{"href":"https:\/\/workboot.fr\/ciela\/wp-json\/wp\/v2\/media?parent=3852"}],"curies":[{"name":"wp","href":"https:\/\/api.w.org\/{rel}","templated":true}]}}