"fcntl" --- "fcntl" および "ioctl" システムコール
*************************************************

======================================================================

This module performs file and I/O control on file descriptors. It is
an interface to the "fcntl()" and "ioctl()" Unix routines. See the
*fcntl(2)* and *ioctl(2)* Unix manual pages for full details.

Availability: Unix, not WASI.

このモジュール内の全ての関数はファイル記述子 *fd* を最初の引数に取りま
す。 この値は "sys.stdin.fileno()" が返すような整数のファイル記述子で
も、 "sys.stdin" 自体のような、純粋にファイル記述子だけを返す
"fileno()" メソッドを提供している "io.IOBase"  オブジェクトでもかまい
ません。

バージョン 3.3 で変更: 以前は "IOError" を送出していたこのモジュールの
操作が、 "OSError" を送出するようになりました。

バージョン 3.8 で変更: The "fcntl" module now contains "F_ADD_SEALS",
"F_GET_SEALS", and "F_SEAL_*" constants for sealing of
"os.memfd_create()" file descriptors.

バージョン 3.9 で変更: On macOS, the "fcntl" module exposes the
"F_GETPATH" constant, which obtains the path of a file from a file
descriptor. On Linux(>=3.15), the "fcntl" module exposes the
"F_OFD_GETLK", "F_OFD_SETLK" and "F_OFD_SETLKW" constants, which are
used when working with open file description locks.

バージョン 3.10 で変更: On Linux >= 2.6.11, the "fcntl" module exposes
the "F_GETPIPE_SZ" and "F_SETPIPE_SZ" constants, which allow to check
and modify a pipe's size respectively.

バージョン 3.11 で変更: On FreeBSD, the "fcntl" module exposes the
"F_DUP2FD" and "F_DUP2FD_CLOEXEC" constants, which allow to duplicate
a file descriptor, the latter setting "FD_CLOEXEC" flag in addition.

バージョン 3.12 で変更: On Linux >= 4.5, the "fcntl" module exposes
the "FICLONE" and "FICLONERANGE" constants, which allow to share some
data of one file with another file by reflinking on some filesystems
(e.g., btrfs, OCFS2, and XFS). This behavior is commonly referred to
as "copy-on-write".

バージョン 3.13 で変更: On Linux >= 2.6.32, the "fcntl" module exposes
the "F_GETOWN_EX", "F_SETOWN_EX", "F_OWNER_TID", "F_OWNER_PID",
"F_OWNER_PGRP" constants, which allow to direct I/O availability
signals to a specific thread, process, or process group. On Linux >=
4.13, the "fcntl" module exposes the "F_GET_RW_HINT", "F_SET_RW_HINT",
"F_GET_FILE_RW_HINT", "F_SET_FILE_RW_HINT", and "RWH_WRITE_LIFE_*"
constants, which allow to inform the kernel about the relative
expected lifetime of writes on a given inode or via a particular open
file description. On Linux >= 5.1 and NetBSD, the "fcntl" module
exposes the "F_SEAL_FUTURE_WRITE" constant for use with "F_ADD_SEALS"
and "F_GET_SEALS" operations. On FreeBSD, the "fcntl" module exposes
the "F_READAHEAD", "F_ISUNIONSTACK", and "F_KINFO" constants. On macOS
and FreeBSD, the "fcntl" module exposes the "F_RDAHEAD" constant. On
NetBSD and AIX, the "fcntl" module exposes the "F_CLOSEM" constant. On
NetBSD, the "fcntl" module exposes the "F_MAXFD" constant. On macOS
and NetBSD, the "fcntl" module exposes the "F_GETNOSIGPIPE" and
"F_SETNOSIGPIPE" constant.

バージョン 3.14 で変更: On Linux >= 6.1, the "fcntl" module exposes
the "F_DUPFD_QUERY" to query a file descriptor pointing to the same
file.

このモジュールには、以下の関数が定義されています:

fcntl.fcntl(fd, cmd, arg=0, /)

   Perform the operation *cmd* on file descriptor *fd* (file objects
   providing a "fileno()" method are accepted as well).  The values
   used for *cmd* are operating system dependent, and are available as
   constants in the "fcntl" module, using the same names as used in
   the relevant C header files. The argument *arg* can either be an
   integer value, a *bytes-like object*, or a string. The type and
   size of *arg* must match the type and size of the argument of the
   operation as specified in the relevant C documentation.

   When *arg* is an integer, the function returns the integer return
   value of the C "fcntl()" call.

   When the argument is bytes-like object, it represents a binary
   structure, for example, created by "struct.pack()". A string value
   is encoded to binary using the UTF-8 encoding. The binary data is
   copied to a buffer whose address is passed to the C "fcntl()" call.
   The return value after a successful call is the contents of the
   buffer, converted to a "bytes" object. The length of the returned
   object will be the same as the length of the *arg* argument. This
   is limited to 1024 bytes.

   If the "fcntl()" call fails, an "OSError" is raised.

   注釈:

     If the type or the size of *arg* does not match the type or size
     of the argument of the operation (for example, if an integer is
     passed when a pointer is expected, or the information returned in
     the buffer by the operating system is larger than 1024 bytes),
     this is most likely to result in a segmentation violation or a
     more subtle data corruption.

   引数 "fd", "cmd", "arg" を指定して 監査イベント "fcntl.fcntl" を送
   出します。

   バージョン 3.14 で変更: Add support of arbitrary *bytes-like
   objects*, not only "bytes".

fcntl.ioctl(fd, request, arg=0, mutate_flag=True, /)

   この関数は "fcntl()" 関数と同じですが、引数の扱いがより複雑であると
   ころが異なります。

   The *request* parameter is limited to values that can fit in
   32-bits or 64-bits, depending on the platform. Additional constants
   of interest for use as the *request* argument can be found in the
   "termios" module, under the same names as used in the relevant C
   header files.

   The parameter *arg* can be an integer, a *bytes-like object*, or a
   string. The type and size of *arg* must match the type and size of
   the argument of the operation as specified in the relevant C
   documentation.

   If *arg* does not support the read-write buffer interface or the
   *mutate_flag* is false, behavior is as for the "fcntl()" function.

   If *arg* supports the read-write buffer interface (like
   "bytearray") and *mutate_flag* is true (the default), then the
   buffer is (in effect) passed to the underlying "ioctl()" system
   call, the latter's return code is passed back to the calling
   Python, and the buffer's new contents reflect the action of the
   "ioctl()".  This is a slight simplification, because if the
   supplied buffer is less than 1024 bytes long it is first copied
   into a static buffer 1024 bytes long which is then passed to
   "ioctl()" and copied back into the supplied buffer.

   If the "ioctl()" call fails, an "OSError" exception is raised.

   注釈:

     If the type or size of *arg* does not match the type or size of
     the operation's argument (for example, if an integer is passed
     when a pointer is expected, or the information returned in the
     buffer by the operating system is larger than 1024 bytes, or the
     size of the mutable bytes-like object is too small), this is most
     likely to result in a segmentation violation or a more subtle
     data corruption.

   以下に例を示します:

      >>> import array, fcntl, struct, termios, os
      >>> os.getpgrp()
      13341
      >>> struct.unpack('h', fcntl.ioctl(0, termios.TIOCGPGRP, "  "))[0]
      13341
      >>> buf = array.array('h', [0])
      >>> fcntl.ioctl(0, termios.TIOCGPGRP, buf, 1)
      0
      >>> buf
      array('h', [13341])

   引数 "fd", "request", "arg" を指定して 監査イベント "fcntl.ioctl"
   を送出します。

   バージョン 3.14 で変更: The GIL is always released during a system
   call. System calls failing with EINTR are automatically retried.

fcntl.flock(fd, operation, /)

   ファイル記述子 *fd* ("fileno()" メソッドを提供しているファイルオブ
   ジェクトも含む) に対してロック操作 *operation* を実行します。 詳細
   は Unix マニュアルの *flock(2)* を参照してください (システムによっ
   ては、この関数は "fcntl()" を使ってエミュレーションされています)。

   If the "flock()" call fails, an "OSError" exception is raised.

   引数 "fd", "operation" を指定して 監査イベント "fcntl.flock" を送出
   します。

fcntl.lockf(fd, cmd, len=0, start=0, whence=0, /)

   本質的に "fcntl()" によるロッキングの呼び出しをラップしたものです。
   *fd* はロックまたはアンロックするファイルのファイル記述子
   ("fileno()" メソッドを提供するファイルオブジェクトも受け付けられま
   す) で、 *cmd* は以下の値のうちいずれかになります:

   fcntl.LOCK_UN

      Release an existing lock.

   fcntl.LOCK_SH

      Acquire a shared lock.

   fcntl.LOCK_EX

      Acquire an exclusive lock.

   fcntl.LOCK_NB

      Bitwise OR with any of the other three "LOCK_*" constants to
      make the request non-blocking.

   If "LOCK_NB" is used and the lock cannot be acquired, an "OSError"
   will be raised and the exception will have an *errno* attribute set
   to "EACCES" or "EAGAIN" (depending on the operating system; for
   portability, check for both values).  On at least some systems,
   "LOCK_EX" can only be used if the file descriptor refers to a file
   opened for writing.

   *len* はロックを行いたいバイト数、 *start* はロック領域先頭の
   *whence* からの相対的なバイトオフセット、 *whence* は
   "io.IOBase.seek()" と同じで、具体的には:

   * "0" -- relative to the start of the file ("os.SEEK_SET")

   * "1" -- relative to the current buffer position ("os.SEEK_CUR")

   * "2" -- relative to the end of the file ("os.SEEK_END")

   *start* の標準の値は 0 で、ファイルの先頭から開始することを意味しま
   す。*len* の標準の値は 0 で、ファイルの終了までロックすることを表し
   ます。*whence* の標準の値も 0 です。

   引数 "fd", "cmd", "len", "start", "whence" を指定して 監査イベント
   "fcntl.lockf" を送出します。

以下に (全ての SVR4 互換システムでの) 例を示します:

   import struct, fcntl, os

   f = open(...)
   rv = fcntl.fcntl(f, fcntl.F_SETFL, os.O_NDELAY)

   lockdata = struct.pack('hhllhh', fcntl.F_WRLCK, 0, 0, 0, 0, 0)
   rv = fcntl.fcntl(f, fcntl.F_SETLKW, lockdata)

最初の例では、戻り値 *rv* は整数値を保持しています; 二つ目の例では
"bytes" オブジェクトを保持しています。 *lockdata* 変数の構造体レイアウ
トはシステム依存です --- 従って "flock()" を呼ぶ方が良いでしょう。

参考:

  "os" モジュール
     If the locking flags "O_SHLOCK" and "O_EXLOCK" are present in the
     "os" module (on BSD only), the "os.open()" function provides an
     alternative to the "lockf()" and "flock()" functions.
