ARK-OS/arkos/kernel/glibc/manual/filesys.texi

@node File System Interface, Pipes and FIFOs, Low-Level I/O, Top
@c %MENU% Functions for manipulating files
@chapter File System Interface

This chapter describes @theglibc{}'s functions for manipulating
files.  Unlike the input and output functions (@pxref{I/O on Streams};
@pxref{Low-Level I/O}), these functions are concerned with operating
on the files themselves rather than on their contents.

Among the facilities described in this chapter are functions for
examining or modifying directories, functions for renaming and deleting
files, and functions for examining and setting file attributes such as
access permissions and modification times.

@menu
* Working Directory::           This is used to resolve relative
				 file names.
* Descriptor-Relative Access::  Ways to control file name lookup.
* Accessing Directories::       Finding out what files a directory
				 contains.
* Working with Directory Trees:: Apply actions to all files or a selectable
                                 subset of a directory hierarchy.
* Hard Links::                  Adding alternate names to a file.
* Symbolic Links::              A file that ``points to'' a file name.
* Deleting Files::              How to delete a file, and what that means.
* Renaming Files::              Changing a file's name.
* Creating Directories::        A system call just for creating a directory.
* File Attributes::             Attributes of individual files.
* Making Special Files::        How to create special files.
* Temporary Files::             Naming and creating temporary files.
@end menu

@node Working Directory
@section Working Directory

@cindex current working directory
@cindex working directory
@cindex change working directory
Each process has associated with it a directory, called its @dfn{current
working directory} or simply @dfn{working directory}, that is used in
the resolution of relative file names (@pxref{File Name Resolution}).

When you log in and begin a new session, your working directory is
initially set to the home directory associated with your login account
in the system user database.  You can find any user's home directory
using the @code{getpwuid} or @code{getpwnam} functions; see @ref{User
Database}.

Users can change the working directory using shell commands like
@code{cd}.  The functions described in this section are the primitives
used by those commands and by other programs for examining and changing
the working directory.
@pindex cd

Prototypes for these functions are declared in the header file
@file{unistd.h}.
@pindex unistd.h

@deftypefun {char *} getcwd (char *@var{buffer}, size_t @var{size})
@standards{POSIX.1, unistd.h}
@safety{@prelim{}@mtsafe{}@asunsafe{@ascuheap{}}@acunsafe{@acsmem{} @acsfd{}}}
@c If buffer is NULL, this function calls malloc and realloc, and, in
@c case of error, free.  Linux offers a getcwd syscall that we use on
@c GNU/Linux systems, but it may fail if the pathname is too long.  As a
@c fallback, and on other systems, the generic implementation opens each
@c parent directory with opendir, which allocates memory for the
@c directory stream with malloc.  If a fstatat64 syscall is not
@c available, very deep directory trees may also have to malloc to build
@c longer sequences of ../../../... than those supported by a global
@c const read-only string.

@c linux/__getcwd
@c  posix/__getcwd
@c   malloc/realloc/free if buffer is NULL, or if dir is too deep
@c   lstat64 -> see its own entry
@c   fstatat64
@c     direct syscall if possible, alloca+snprintf+*stat64 otherwise
@c   openat64_not_cancel_3, close_not_cancel_no_status
@c   __fdopendir, __opendir, __readdir, rewinddir
The @code{getcwd} function returns an absolute file name representing
the current working directory, storing it in the character array
@var{buffer} that you provide.  The @var{size} argument is how you tell
the system the allocation size of @var{buffer}.

The @glibcadj{} version of this function also permits you to specify a
null pointer for the @var{buffer} argument.  Then @code{getcwd}
allocates a buffer automatically, as with @code{malloc}
(@pxref{Unconstrained Allocation}).  If the @var{size} is greater than
zero, then the buffer is that large; otherwise, the buffer is as large
as necessary to hold the result.

The return value is @var{buffer} on success and a null pointer on failure.
The following @code{errno} error conditions are defined for this function:

@table @code
@item EINVAL
The @var{size} argument is zero and @var{buffer} is not a null pointer.

@item ERANGE
The @var{size} argument is less than the length of the working directory
name.  You need to allocate a bigger array and try again.

@item EACCES
Permission to read or search a component of the file name was denied.
@end table
@end deftypefun

You could implement the behavior of GNU's @w{@code{getcwd (NULL, 0)}}
using only the standard behavior of @code{getcwd}:

@smallexample
char *
gnu_getcwd ()
@{
  size_t size = 100;

  while (1)
    @{
      char *buffer = (char *) xmalloc (size);
      if (getcwd (buffer, size) == buffer)
        return buffer;
      free (buffer);
      if (errno != ERANGE)
        return 0;
      size *= 2;
    @}
@}
@end smallexample

@noindent
@xref{Malloc Examples}, for information about @code{xmalloc}, which is
not a library function but is a customary name used in most GNU
software.

@deftypefn {Deprecated Function} {char *} getwd (char *@var{buffer})
@standards{BSD, unistd.h}
@safety{@prelim{}@mtsafe{}@asunsafe{@ascuheap{} @ascuintl{}}@acunsafe{@acsmem{} @acsfd{}}}
@c Besides the getcwd safety issues, it calls strerror_r on error, which
@c brings in all of the i18n issues.
This is similar to @code{getcwd}, but has no way to specify the size of
the buffer.  @Theglibc{} provides @code{getwd} only
for backwards compatibility with BSD.

The @var{buffer} argument should be a pointer to an array at least
@code{PATH_MAX} bytes long (@pxref{Limits for Files}).  On @gnuhurdsystems{}
there is no limit to the size of a file name, so this is not
necessarily enough space to contain the directory name.  That is why
this function is deprecated.
@end deftypefn

@vindex PWD
@deftypefun {char *} get_current_dir_name (void)
@standards{GNU, unistd.h}
@safety{@prelim{}@mtsafe{@mtsenv{}}@asunsafe{@ascuheap{}}@acunsafe{@acsmem{} @acsfd{}}}
@c Besides getcwd, which this function calls as a fallback, it calls
@c getenv, with the potential thread-safety issues that brings about.
The @code{get_current_dir_name} function is basically equivalent to
@w{@code{getcwd (NULL, 0)}}, except the value of the @env{PWD}
environment variable is first examined, and if it does in fact
correspond to the current directory, that value is returned.  This is
a subtle difference which is visible if the path described by the
value in @env{PWD} is using one or more symbolic links, in which case
the value returned by @code{getcwd} would resolve the symbolic links
and therefore yield a different result.

This function is a GNU extension.
@end deftypefun

@deftypefun int chdir (const char *@var{filename})
@standards{POSIX.1, unistd.h}
@safety{@prelim{}@mtsafe{}@assafe{}@acsafe{}}
This function is used to set the process's working directory to
@var{filename}.

The normal, successful return value from @code{chdir} is @code{0}.  A
value of @code{-1} is returned to indicate an error.  The @code{errno}
error conditions defined for this function are the usual file name
syntax errors (@pxref{File Name Errors}), plus @code{ENOTDIR} if the
file @var{filename} is not a directory.
@end deftypefun

@deftypefun int fchdir (int @var{filedes})
@standards{XPG, unistd.h}
@safety{@prelim{}@mtsafe{}@assafe{}@acsafe{}}
This function is used to set the process's working directory to
directory associated with the file descriptor @var{filedes}.

The normal, successful return value from @code{fchdir} is @code{0}.  A
value of @code{-1} is returned to indicate an error.  The following
@code{errno} error conditions are defined for this function:

@table @code
@item EACCES
Read permission is denied for the directory named by @code{dirname}.

@item EBADF
The @var{filedes} argument is not a valid file descriptor.

@item ENOTDIR
The file descriptor @var{filedes} is not associated with a directory.

@item EINTR
The function call was interrupt by a signal.

@item EIO
An I/O error occurred.
@end table
@end deftypefun

@node Descriptor-Relative Access
@section Descriptor-Relative Access
@cindex file name resolution based on descriptors
@cindex descriptor-based file name resolution
@cindex @code{@dots{}at} functions

Many functions that accept file names have @code{@dots{}at} variants
which accept a file descriptor and a file name argument instead of just
a file name argument.  For example, @code{fstatat} is the
descriptor-based variant of the @code{fstat} function.  Most such
functions also accept an additional flags argument which changes the
behavior of the file name lookup based on the passed @code{AT_@dots{}}
flags.

There are several reasons to use descriptor-relative access:

@itemize @bullet
@item
The working directory is a process-wide resource, so individual threads
cannot change it without affecting other threads in the process.
Explicitly specifying the directory against which relative paths are
resolved can be a thread-safe alternative to changing the working
directory.

@item
If a program wishes to access a directory tree which is being modified
concurrently, perhaps even by a different user on the system, the
program must avoid looking up file names with multiple components, in
order to detect symbolic links, using the @code{O_NOFOLLOW} flag
(@pxref{Open-time Flags}) or the @code{AT_SYMLINK_FOLLOW} flag
(described below).  Without directory-relative access, it is necessary
to use the @code{fchdir} function to change the working directory
(@pxref{Working Directory}), which is not thread-safe.

@item
Listing directory contents using the @code{readdir} or @code{readdir64}
functions (@pxref{Reading/Closing Directory}) does not provide full file
name paths.  Using @code{@dots{}at} functions, it is possible to use the
file names directly, without having to construct such full paths.

@item
Additional flags available with some of the @code{@dots{}at} functions
provide access to functionality which is not available otherwise.
@end itemize

The file descriptor used by these @code{@dots{}at} functions has the
following uses:

@itemize @bullet
@item
It can be a file descriptor referring to a directory.  Such a descriptor
can be created explicitly using the @code{open} function and the
@code{O_RDONLY} file access mode, with or without the @code{O_DIRECTORY}
flag.  @xref{Opening and Closing Files}.  Or it can be created
implicitly by @code{opendir} and retrieved using the @code{dirfd}
function.  @xref{Opening a Directory}.

If a directory descriptor is used with one of the @code{@dots{}at}
functions, a relative file name argument is resolved relative to
directory referred to by the file descriptor, just as if that directory
were the current working directory.  Absolute file name arguments
(starting with @samp{/}) are resolved against the file system root, and
the descriptor argument is effectively ignored.

This means that file name lookup is not constrained to the directory of
the descriptor.  For example, it is possible to access a file
@file{example} in the descriptor's parent directory using a file name
argument @code{"../example"}, or in the root directory using
@code{"/example"}.

If the file descriptor refers to a directory, the empty string @code{""}
is not a valid file name argument.  It is possible to use @code{"."} to
refer to the directory itself.  Also see @code{AT_EMPTY_PATH} below.

@item
@vindex @code{AT_FDCWD}
The special value @code{AT_FDCWD}.  This means that the current working
directory is used for the lookup if the file name is a relative.  For
@code{@dots{}at} functions with an @code{AT_@dots{}} flags argument,
this provides a shortcut to use those flags with regular (not
descriptor-based) file name lookups.

If @code{AT_FDCWD} is used, the empty string @code{""} is not a valid
file name argument.

@item
An arbitrary file descriptor, along with an empty string @code{""} as
the file name argument, and the @code{AT_EMPTY_PATH} flag.  In this
case, the operation uses the file descriptor directly, without further
file name resolution.  On Linux, this allows operations on descriptors
opened with the @code{O_PATH} flag.  For regular descriptors (opened
without @code{O_PATH}), the same functionality is also available through
the plain descriptor-based functions (for example, @code{fstat} instead
of @code{fstatat}).

This is a GNU extension.
@end itemize

@cindex file name resolution flags
@cindex @code{AT_*} file name resolution flags
The flags argument in @code{@dots{}at} functions can be a combination of
the following flags, defined in @file{fcntl.h}.  Not all such functions
support all flags, and some (such as @code{openat}) do not accept a
flags argument at all.  Although the flags specific to each function have
distinct values from each other, some flags (relevant to different
functions) might share the same value and therefore are not guaranteed to
have unique values.

A non-exhaustive list of common flags and their descriptions follows.  Flags
specific to a function are described alongside the function itself.  In
these flag descriptions, the @dfn{effective final path component} refers to
the final component (basename) of the full path constructed from the
descriptor and file name arguments, using file name lookup, as described
above.

@vtable @code
@item AT_EMPTY_PATH
This flag is used with an empty file name @code{""} and a descriptor
which does not necessarily refer to a directory.  It is most useful with
@code{O_PATH} descriptors, as described above.  This flag is a GNU
extension.

@item AT_NO_AUTOMOUNT
If the effective final path component refers to a potential file system
mount point controlled by an auto-mounting service, the operation does
not trigger auto-mounting and refers to the unmounted mount point
instead.  @xref{Mount-Unmount-Remount}.  If a file system has already
been mounted at the effective final path component, the operation
applies to the file or directory in the mounted file system, not the
underlying file system that was mounted over.  This flag is a GNU
extension.

@item AT_SYMLINK_FOLLOW
If the effective final path component is a symbolic link, the
operation follows the symbolic link and operates on its target.  (For
most functions, this is the default behavior.)

@item AT_SYMLINK_NOFOLLOW
If the effective final path component is a symbolic link, the
operation operates on the symbolic link, without following it.  The
difference in behavior enabled by this flag is similar to the difference
between the @code{lstat} and @code{stat} functions, or the behavior
activated by the @code{O_NOFOLLOW} argument to the @code{open} function.
Even with the @code{AT_SYMLINK_NOFOLLOW} flag present, symbolic links in
a non-final component of the file name are still followed.
@end vtable

@strong{Note:} There is no relationship between these flags and the type
argument to the @code{getauxval} function (with @code{AT_@dots{}}
constants defined in @file{elf.h}).  @xref{Auxiliary Vector}.

@cindex common errors in descriptor-relative functions
@cindex common errors in @code{@dots{}at} functions

The @code{@dots{}at} functions have some common error conditions due to the
nature of descriptor-relative access.  A list of common errors and their
descriptions follows.  Errors specific to a function are described alongside
the function itself.

@table @code
@item EBADF
The file name argument is a relative path but the descriptor argument
is neither @code{AT_FDCWD} nor a valid file descriptor.

@item EINVAL
If the function accepts a @var{flags} argument, the flag combination passed
is not valid for the function.

@item ENOTDIR
The file name argument is a relative file name but the descriptor
argument is associated with a file that is not a directory.
@end table

@node Accessing Directories
@section Accessing Directories
@cindex accessing directories
@cindex reading from a directory
@cindex directories, accessing

The facilities described in this section let you read the contents of a
directory file.  This is useful if you want your program to list all the
files in a directory, perhaps as part of a menu.

@cindex directory stream
The @code{opendir} function opens a @dfn{directory stream} whose
elements are directory entries.  Alternatively @code{fdopendir} can be
used which can have advantages if the program needs to have more
control over the way the directory is opened for reading.  This
allows, for instance, to pass the @code{O_NOATIME} flag to
@code{open}.

You use the @code{readdir} function on the directory stream to
retrieve these entries, represented as @w{@code{struct dirent}}
objects.  The name of the file for each entry is stored in the
@code{d_name} member of this structure.  There are obvious parallels
here to the stream facilities for ordinary files, described in
@ref{I/O on Streams}.

@menu
* Directory Entries::           Format of one directory entry.
* Opening a Directory::         How to open a directory stream.
* Reading/Closing Directory::   How to read directory entries from the stream.
* Simple Directory Lister::     A very simple directory listing program.
* Random Access Directory::     Rereading part of the directory
                                 already read with the same stream.
* Scanning Directory Content::  Get entries for user selected subset of
                                 contents in given directory.
* Simple Directory Lister Mark II::  Revised version of the program.
* Low-level Directory Access::  AS-Safe functions for directory access.
@end menu

@node Directory Entries
@subsection Format of a Directory Entry

@pindex dirent.h
This section describes what you find in a single directory entry, as you
might obtain it from a directory stream.  All the symbols are declared
in the header file @file{dirent.h}.

@deftp {Data Type} {struct dirent}
@standards{POSIX.1, dirent.h}
This is a structure type used to return information about directory
entries.  It contains the following fields:

@table @code
@item char d_name[]
This is the null-terminated file name component.  This is the only
field you can count on in all POSIX systems.

While this field is defined with a specified length, functions such as
@code{readdir} may return a pointer to a @code{struct dirent} where the
@code{d_name} extends beyond the end of the struct.

@item ino_t d_fileno
This is the file serial number.  For BSD compatibility, you can also
refer to this member as @code{d_ino}.  On @gnulinuxhurdsystems{} and most POSIX
systems, for most files this the same as the @code{st_ino} member that
@code{stat} will return for the file.  @xref{File Attributes}.

@item off_t d_off
This value contains the offset of the next directory entry (after this
entry) in the directory stream.  The value may not be compatible with
@code{lseek} or @code{seekdir}, especially if the width of @code{d_off}
is less than 64 bits.  Directory entries are not ordered by offset, and
the @code{d_off} and @code{d_reclen} values are unrelated.  Seeking on
directory streams is not recommended.  The symbol
@code{_DIRENT_HAVE_D_OFF} is defined if the @code{d_ino} member is
available.

@item unsigned char d_namlen
This is the length of the file name, not including the terminating
null character.  Its type is @code{unsigned char} because that is the
integer type of the appropriate size.  This member is a BSD extension.
The symbol @code{_DIRENT_HAVE_D_NAMLEN} is defined if this member is
available.  (It is not available on Linux.)

@item unsigned short int d_reclen
This is the length of the entire directory record.  When iterating
through a buffer filled by @code{getdents64} (@pxref{Low-level Directory
Access}), this value needs to be added to the offset of the current
directory entry to obtain the offset of the next entry.  When using
@code{readdir} and related functions, the value of @code{d_reclen} is
undefined and should not be accessed.  The symbol
@code{_DIRENT_HAVE_D_RECLEN} is defined if this member is available.

@item unsigned char d_type
This is the type of the file, possibly unknown.  The following constants
are defined for its value:

@vtable @code
@item DT_UNKNOWN
The type is unknown.  Only some filesystems have full support to
return the type of the file, others might always return this value.

@item DT_REG
A regular file.

@item DT_DIR
A directory.

@item DT_FIFO
A named pipe, or FIFO.  @xref{FIFO Special Files}.

@item DT_SOCK
A local-domain socket.  @c !!! @xref{Local Domain}.

@item DT_CHR
A character device.

@item DT_BLK
A block device.

@item DT_LNK
A symbolic link.
@end vtable

This member is a BSD extension.  The symbol @code{_DIRENT_HAVE_D_TYPE}
is defined if this member is available.  On systems where it is used, it
corresponds to the file type bits in the @code{st_mode} member of
@code{struct stat}.  If the value cannot be determined the member
value is @code{DT_UNKNOWN}.  These two macros convert between @code{d_type}
values and @code{st_mode} values:

@deftypefun int IFTODT (mode_t @var{mode})
@standards{BSD, dirent.h}
@safety{@prelim{}@mtsafe{}@assafe{}@acsafe{}}
This returns the @code{d_type} value corresponding to @var{mode}.
@end deftypefun

@deftypefun mode_t DTTOIF (int @var{dtype})
@standards{BSD, dirent.h}
@safety{@prelim{}@mtsafe{}@assafe{}@acsafe{}}
This returns the @code{st_mode} value corresponding to @var{dtype}.
@end deftypefun
@end table

This structure may contain additional members in the future.  Their
availability is always announced in the compilation environment by a
macro named @code{_DIRENT_HAVE_D_@var{xxx}} where @var{xxx} is replaced
by the name of the new member.  For instance, the member @code{d_reclen}
available on some systems is announced through the macro
@code{_DIRENT_HAVE_D_RECLEN}.

When a file has multiple names, each name has its own directory entry.
The only way you can tell that the directory entries belong to a
single file is that they have the same value for the @code{d_fileno}
field.

File attributes such as size, modification times etc., are part of the
file itself, not of any particular directory entry.  @xref{File
Attributes}.
@end deftp

@node Opening a Directory
@subsection Opening a Directory Stream

@pindex dirent.h
This section describes how to open a directory stream.  All the symbols
are declared in the header file @file{dirent.h}.

@deftp {Data Type} DIR
@standards{POSIX.1, dirent.h}
The @code{DIR} data type represents a directory stream.
@end deftp

You shouldn't ever allocate objects of the @code{struct dirent} or
@code{DIR} data types, since the directory access functions do that for
you.  Instead, you refer to these objects using the pointers returned by
the following functions.

Directory streams are a high-level interface.  On Linux, alternative
interfaces for accessing directories using file descriptors are
available.  @xref{Low-level Directory Access}.

@deftypefun {DIR *} opendir (const char *@var{dirname})
@standards{POSIX.1, dirent.h}
@safety{@prelim{}@mtsafe{}@asunsafe{@ascuheap{}}@acunsafe{@acsmem{} @acsfd{}}}
@c Besides the safe syscall, we have to allocate the DIR object with
@c __alloc_dir, that calls malloc.
The @code{opendir} function opens and returns a directory stream for
reading the directory whose file name is @var{dirname}.  The stream has
type @code{DIR *}.

If unsuccessful, @code{opendir} returns a null pointer.  In addition to
the usual file name errors (@pxref{File Name Errors}), the
following @code{errno} error conditions are defined for this function:

@table @code
@item EACCES
Read permission is denied for the directory named by @code{dirname}.

@item EMFILE
The process has too many files open.

@item ENFILE
The entire system, or perhaps the file system which contains the
directory, cannot support any additional open files at the moment.
(This problem cannot happen on @gnuhurdsystems{}.)

@item ENOMEM
Not enough memory available.
@end table

The @code{DIR} type is typically implemented using a file descriptor,
and the @code{opendir} function in terms of the @code{open} function.
@xref{Low-Level I/O}.  Directory streams and the underlying
file descriptors are closed on @code{exec} (@pxref{Executing a File}).
@end deftypefun

The directory which is opened for reading by @code{opendir} is
identified by the name.  In some situations this is not sufficient.
Or the way @code{opendir} implicitly creates a file descriptor for the
directory is not the way a program might want it.  In these cases an
alternative interface can be used.

@deftypefun {DIR *} fdopendir (int @var{fd})
@standards{GNU, dirent.h}
@safety{@prelim{}@mtsafe{}@asunsafe{@ascuheap{}}@acunsafe{@acsmem{} @acsfd{}}}
@c The DIR object is allocated with __alloc_dir, that calls malloc.
The @code{fdopendir} function works just like @code{opendir} but
instead of taking a file name and opening a file descriptor for the
directory the caller is required to provide a file descriptor.  This
file descriptor is then used in subsequent uses of the returned
directory stream object.

The caller must make sure the file descriptor is associated with a
directory and it allows reading.

If the @code{fdopendir} call returns successfully the file descriptor
is now under the control of the system.  It can be used in the same
way the descriptor implicitly created by @code{opendir} can be used
but the program must not close the descriptor.

In case the function is unsuccessful it returns a null pointer and the
file descriptor remains to be usable by the program.  The following
@code{errno} error conditions are defined for this function:

@table @code
@item EBADF
The file descriptor is not valid.

@item ENOTDIR
The file descriptor is not associated with a directory.

@item EINVAL
The descriptor does not allow reading the directory content.

@item ENOMEM
Not enough memory available.
@end table
@end deftypefun

In some situations it can be desirable to get hold of the file
descriptor which is created by the @code{opendir} call.  For instance,
to switch the current working directory to the directory just read the
@code{fchdir} function could be used.  Historically the @code{DIR} type
was exposed and programs could access the fields.  This does not happen
in @theglibc{}.  Instead a separate function is provided to allow
access.

@deftypefun int dirfd (DIR *@var{dirstream})
@standards{GNU, dirent.h}
@safety{@prelim{}@mtsafe{}@assafe{}@acsafe{}}
The function @code{dirfd} returns the file descriptor associated with
the directory stream @var{dirstream}.  This descriptor can be used until
the directory is closed with @code{closedir}.  If the directory stream
implementation is not using file descriptors the return value is
@code{-1}.
@end deftypefun

@node Reading/Closing Directory
@subsection Reading and Closing a Directory Stream

@pindex dirent.h
This section describes how to read directory entries from a directory
stream, and how to close the stream when you are done with it.  All the
symbols are declared in the header file @file{dirent.h}.

@deftypefun {struct dirent *} readdir (DIR *@var{dirstream})
@standards{POSIX.1, dirent.h}
@safety{@prelim{}@mtsafe{}@asunsafe{@asulock{}}@acunsafe{@aculock{}}}
@c This function holds dirstream's non-recursive lock, which brings
@c about the usual issues with locks and async signals and cancellation,
@c but the lock taking is not enough to make the returned value safe to
@c use, since it points to a stream's internal buffer that can be
@c overwritten by subsequent calls or even released by closedir.
This function reads the next entry from the directory.  It normally
returns a pointer to a structure containing information about the
file.  This structure is associated with the @var{dirstream} handle
and can be rewritten by a subsequent call.

@strong{Portability Note:} On some systems @code{readdir} may not
return entries for @file{.} and @file{..}, even though these are always
valid file names in any directory.  @xref{File Name Resolution}.

If a directory is modified between a call to @code{readdir} and after
the directory stream was created or @code{rewinddir} was last called on
it, it is unspecified according to POSIX whether newly created or
removed entries appear among the entries returned by repeated
@code{readdir} calls before the end of the directory is reached.
However, due to practical implementation constraints, it is possible
that entries (including unrelated, unmodified entries) appear multiple
times or do not appear at all if the directory is modified while listing
it.  If the application intends to create files in the directory, it may
be necessary to complete the iteration first and create a copy of the
information obtained before creating any new files.  (See below for
instructions regarding copying of @code{d_name}.)  The iteration can be
restarted using @code{rewinddir}.  @xref{Random Access Directory}.

If there are no more entries in the directory or an error is detected,
@code{readdir} returns a null pointer.  The following @code{errno} error
conditions are defined for this function:

@table @code
@item EBADF
The @var{dirstream} argument is not valid.
@end table

To distinguish between an end-of-directory condition or an error, you
must set @code{errno} to zero before calling @code{readdir}.  To avoid
entering an infinite loop, you should stop reading from the directory
after the first error.

@strong{Caution:} The pointer returned by @code{readdir} points to
a buffer within the @code{DIR} object.  The data in that buffer will
be overwritten by the next call to @code{readdir}.  You must take care,
for instance, to copy the @code{d_name} string if you need it later.

Because of this, it is not safe to share a @code{DIR} object among
multiple threads, unless you use your own locking to ensure that
no thread calls @code{readdir} while another thread is still using the
data from the previous call.  In @theglibc{}, it is safe to call
@code{readdir} from multiple threads as long as each thread uses
its own @code{DIR} object.  POSIX.1-2008 does not require this to
be safe, but we are not aware of any operating systems where it
does not work.

@code{readdir_r} allows you to provide your own buffer for the
@code{struct dirent}, but it is less portable than @code{readdir}, and
has problems with very long filenames (see below).  We recommend
you use @code{readdir}, but do not share @code{DIR} objects.
@end deftypefun

@deftypefun int readdir_r (DIR *@var{dirstream}, struct dirent *@var{entry}, struct dirent **@var{result})
@standards{GNU, dirent.h}
@safety{@prelim{}@mtsafe{}@asunsafe{@asulock{}}@acunsafe{@aculock{}}}
This function is a version of @code{readdir} which performs internal
locking.  Like @code{readdir} it returns the next entry from the
directory.  To prevent conflicts between simultaneously running
threads the result is stored inside the @var{entry} object.

@strong{Portability Note:} @code{readdir_r} is deprecated.  It is
recommended to use @code{readdir} instead of @code{readdir_r} for the
following reasons:

@itemize @bullet
@item
On systems which do not define @code{NAME_MAX}, it may not be possible
to use @code{readdir_r} safely because the caller does not specify the
length of the buffer for the directory entry.

@item
On some systems, @code{readdir_r} cannot read directory entries with
very long names.  If such a name is encountered, @theglibc{}
implementation of @code{readdir_r} returns with an error code of
@code{ENAMETOOLONG} after the final directory entry has been read.  On
other systems, @code{readdir_r} may return successfully, but the
@code{d_name} member may not be NUL-terminated or may be truncated.

@item
POSIX-1.2008 does not guarantee that @code{readdir} is thread-safe,
even when access to the same @var{dirstream} is serialized.  But in
current implementations (including @theglibc{}), it is safe to call
@code{readdir} concurrently on different @var{dirstream}s, so there is
no need to use @code{readdir_r} in most multi-threaded programs.  In
the rare case that multiple threads need to read from the same
@var{dirstream}, it is still better to use @code{readdir} and external
synchronization.

@item
It is expected that future versions of POSIX will obsolete
@code{readdir_r} and mandate the level of thread safety for
@code{readdir} which is provided by @theglibc{} and other
implementations today.
@end itemize

Normally @code{readdir_r} returns zero and sets @code{*@var{result}}
to @var{entry}.  If there are no more entries in the directory or an
error is detected, @code{readdir_r} sets @code{*@var{result}} to a
null pointer and returns a nonzero error code, also stored in
@code{errno}, as described for @code{readdir}.

It is also important to look at the definition of the @code{struct
dirent} type.  Simply passing a pointer to an object of this type for
the second parameter of @code{readdir_r} might not be enough.  Some
systems don't define the @code{d_name} element sufficiently long.  In
this case the user has to provide additional space.  There must be room
for at least @code{NAME_MAX + 1} characters in the @code{d_name} array.
Code to call @code{readdir_r} could look like this:

@smallexample
  union
  @{
    struct dirent d;
    char b[offsetof (struct dirent, d_name) + NAME_MAX + 1];
  @} u;

  if (readdir_r (dir, &u.d, &res) == 0)
    @dots{}
@end smallexample
@end deftypefun

To support large filesystems on 32-bit machines there are LFS variants
of the last two functions.

@deftypefun {struct dirent64 *} readdir64 (DIR *@var{dirstream})
@standards{LFS, dirent.h}
@safety{@prelim{}@mtsafe{}@asunsafe{@asulock{}}@acunsafe{@aculock{}}}
The @code{readdir64} function is just like the @code{readdir} function
except that it returns a pointer to a record of type @code{struct
dirent64}.  Some of the members of this data type (notably @code{d_ino})
might have a different size to allow large filesystems.

In all other aspects this function is equivalent to @code{readdir}.
@end deftypefun

@deftypefun int readdir64_r (DIR *@var{dirstream}, struct dirent64 *@var{entry}, struct dirent64 **@var{result})
@standards{LFS, dirent.h}
@safety{@prelim{}@mtsafe{}@asunsafe{@asulock{}}@acunsafe{@aculock{}}}
The deprecated @code{readdir64_r} function is equivalent to the
@code{readdir_r} function except that it takes parameters of base type
@code{struct dirent64} instead of @code{struct dirent} in the second and
third position.  The same precautions mentioned in the documentation of
@code{readdir_r} also apply here.
@end deftypefun

@deftypefun int closedir (DIR *@var{dirstream})
@standards{POSIX.1, dirent.h}
@safety{@prelim{}@mtsafe{}@asunsafe{@ascuheap{} @asulock{/hurd}}@acunsafe{@acsmem{} @acsfd{} @aculock{/hurd}}}
@c No synchronization in the posix implementation, only in the hurd
@c one.  This is regarded as safe because it is undefined behavior if
@c other threads could still be using the dir stream while it's closed.
This function closes the directory stream @var{dirstream}.  It returns
@code{0} on success and @code{-1} on failure.

The following @code{errno} error conditions are defined for this
function:

@table @code
@item EBADF
The @var{dirstream} argument is not valid.
@end table
@end deftypefun

@node Simple Directory Lister
@subsection Simple Program to List a Directory

Here's a simple program that prints the names of the files in
the current working directory:

@smallexample
@include dir.c.texi
@end smallexample

The order in which files appear in a directory tends to be fairly
random.  A more useful program would sort the entries (perhaps by
alphabetizing them) before printing them; see
@ref{Scanning Directory Content}, and @ref{Array Sort Function}.


@node Random Access Directory
@subsection Random Access in a Directory Stream

@pindex dirent.h
This section describes how to reread parts of a directory that you have
already read from an open directory stream.  All the symbols are
declared in the header file @file{dirent.h}.

@deftypefun void rewinddir (DIR *@var{dirstream})
@standards{POSIX.1, dirent.h}
@safety{@prelim{}@mtsafe{}@asunsafe{@asulock{}}@acunsafe{@aculock{}}}
The @code{rewinddir} function is used to reinitialize the directory
stream @var{dirstream}, so that if you call @code{readdir} it
returns information about the first entry in the directory again.  This
function also notices if files have been added or removed to the
directory since it was opened with @code{opendir}.  (Entries for these
files might or might not be returned by @code{readdir} if they were
added or removed since you last called @code{opendir} or
@code{rewinddir}.)

For example, it is recommended to call @code{rewinddir} followed by
@code{readdir} to check if a directory is empty after listing it with
@code{readdir} and deleting all encountered files from it.
@end deftypefun

@deftypefun {long int} telldir (DIR *@var{dirstream})
@standards{BSD, dirent.h}
@safety{@prelim{}@mtsafe{}@asunsafe{@ascuheap{/bsd} @asulock{/bsd}}@acunsafe{@acsmem{/bsd} @aculock{/bsd}}}
@c The implementation is safe on most platforms, but on BSD it uses
@c cookies, buckets and records, and the global array of pointers to
@c dynamically allocated records is guarded by a non-recursive lock.
The @code{telldir} function returns the file position of the directory
stream @var{dirstream}.  You can use this value with @code{seekdir} to
restore the directory stream to that position.

Using the the @code{telldir} function is not recommended.

The value returned by @code{telldir} may not be compatible with the
@code{d_off} field in @code{struct dirent}, and cannot be used with the
@code{lseek} function.  The returned value may not unambiguously
identify the position in the directory stream.
@end deftypefun

@deftypefun void seekdir (DIR *@var{dirstream}, long int @var{pos})
@standards{BSD, dirent.h}
@safety{@prelim{}@mtsafe{}@asunsafe{@ascuheap{/bsd} @asulock{/bsd}}@acunsafe{@acsmem{/bsd} @aculock{/bsd}}}
@c The implementation is safe on most platforms, but on BSD it uses
@c cookies, buckets and records, and the global array of pointers to
@c dynamically allocated records is guarded by a non-recursive lock.
The @code{seekdir} function sets the file position of the directory
stream @var{dirstream} to @var{pos}.  The value @var{pos} must be the
result of a previous call to @code{telldir} on this particular stream;
closing and reopening the directory can invalidate values returned by
@code{telldir}.

Using the the @code{seekdir} function is not recommended.  To seek to
the beginning of the directory stream, use @code{rewinddir}.
@end deftypefun


@node Scanning Directory Content
@subsection Scanning the Content of a Directory

A higher-level interface to the directory handling functions is the
@code{scandir} function.  With its help one can select a subset of the
entries in a directory, possibly sort them and get a list of names as
the result.

@deftypefun int scandir (const char *@var{dir}, struct dirent ***@var{namelist}, int (*@var{selector}) (const struct dirent *), int (*@var{cmp}) (const struct dirent **, const struct dirent **))
@standards{BSD, dirent.h}
@standards{SVID, dirent.h}
@safety{@prelim{}@mtsafe{}@asunsafe{@ascuheap{}}@acunsafe{@acsmem{} @acsfd{}}}
@c The scandir function calls __opendirat, __readdir, and __closedir to
@c go over the named dir; malloc and realloc to allocate the namelist
@c and copies of each selected dirent, besides the selector, if given,
@c and qsort and the cmp functions if the latter is given.  In spite of
@c the cleanup handler that releases memory and the file descriptor in
@c case of synchronous cancellation, an asynchronous cancellation may
@c still leak memory and a file descriptor.  Although readdir is unsafe
@c in general, the use of an internal dir stream for sequential scanning
@c of the directory with copying of dirents before subsequent calls
@c makes the use safe, and the fact that the dir stream is private to
@c each scandir call does away with the lock issues in readdir and
@c closedir.

The @code{scandir} function scans the contents of the directory selected
by @var{dir}.  The result in *@var{namelist} is an array of pointers to
structures of type @code{struct dirent} which describe all selected
directory entries and which is allocated using @code{malloc}.  Instead
of always getting all directory entries returned, the user supplied
function @var{selector} can be used to decide which entries are in the
result.  Only the entries for which @var{selector} returns a non-zero
value are selected.

Finally the entries in *@var{namelist} are sorted using the
user-supplied function @var{cmp}.  The arguments passed to the @var{cmp}
function are of type @code{struct dirent **}, therefore one cannot
directly use the @code{strcmp} or @code{strcoll} functions; instead see
the functions @code{alphasort} and @code{versionsort} below.

The return value of the function is the number of entries placed in
*@var{namelist}.  If it is @code{-1} an error occurred (either the
directory could not be opened for reading or memory allocation failed) and
the global variable @code{errno} contains more information on the error.
@end deftypefun

As described above, the fourth argument to the @code{scandir} function
must be a pointer to a sorting function.  For the convenience of the
programmer @theglibc{} contains implementations of functions which
are very helpful for this purpose.

@deftypefun int alphasort (const struct dirent **@var{a}, const struct dirent **@var{b})
@standards{BSD, dirent.h}
@standards{SVID, dirent.h}
@safety{@prelim{}@mtsafe{@mtslocale{}}@asunsafe{@ascuheap{}}@acunsafe{@acsmem{}}}
@c Calls strcoll.
The @code{alphasort} function behaves like the @code{strcoll} function
(@pxref{String/Array Comparison}).  The difference is that the arguments
are not string pointers but instead they are of type
@code{struct dirent **}.

The return value of @code{alphasort} is less than, equal to, or greater
than zero depending on the order of the two entries @var{a} and @var{b}.
@end deftypefun

@deftypefun int versionsort (const struct dirent **@var{a}, const struct dirent **@var{b})
@standards{GNU, dirent.h}
@safety{@prelim{}@mtsafe{@mtslocale{}}@assafe{}@acsafe{}}
@c Calls strverscmp, which will accesses the locale object multiple
@c times.
The @code{versionsort} function is like @code{alphasort} except that it
uses the @code{strverscmp} function internally.
@end deftypefun

If the filesystem supports large files we cannot use the @code{scandir}
anymore since the @code{dirent} structure might not able to contain all
the information.  The LFS provides the new type @w{@code{struct
dirent64}}.  To use this we need a new function.

@deftypefun int scandir64 (const char *@var{dir}, struct dirent64 ***@var{namelist}, int (*@var{selector}) (const struct dirent64 *), int (*@var{cmp}) (const struct dirent64 **, const struct dirent64 **))
@standards{GNU, dirent.h}
@safety{@prelim{}@mtsafe{}@asunsafe{@ascuheap{}}@acunsafe{@acsmem{} @acsfd{}}}
@c See scandir.
The @code{scandir64} function works like the @code{scandir} function
except that the directory entries it returns are described by elements
of type @w{@code{struct dirent64}}.  The function pointed to by
@var{selector} is again used to select the desired entries, except that
@var{selector} now must point to a function which takes a
@w{@code{struct dirent64 *}} parameter.

Similarly the @var{cmp} function should expect its two arguments to be
of type @code{struct dirent64 **}.
@end deftypefun

As @var{cmp} is now a function of a different type, the functions
@code{alphasort} and @code{versionsort} cannot be supplied for that
argument.  Instead we provide the two replacement functions below.

@deftypefun int alphasort64 (const struct dirent64 **@var{a}, const struct dirent **@var{b})
@standards{GNU, dirent.h}
@safety{@prelim{}@mtsafe{@mtslocale{}}@asunsafe{@ascuheap{}}@acunsafe{@acsmem{}}}
@c See alphasort.
The @code{alphasort64} function behaves like the @code{strcoll} function
(@pxref{String/Array Comparison}).  The difference is that the arguments
are not string pointers but instead they are of type
@code{struct dirent64 **}.

Return value of @code{alphasort64} is less than, equal to, or greater
than zero depending on the order of the two entries @var{a} and @var{b}.
@end deftypefun

@deftypefun int versionsort64 (const struct dirent64 **@var{a}, const struct dirent64 **@var{b})
@standards{GNU, dirent.h}
@safety{@prelim{}@mtsafe{@mtslocale{}}@assafe{}@acsafe{}}
@c See versionsort.
The @code{versionsort64} function is like @code{alphasort64}, excepted that it
uses the @code{strverscmp} function internally.
@end deftypefun

It is important not to mix the use of @code{scandir} and the 64-bit
comparison functions or vice versa.  There are systems on which this
works but on others it will fail miserably.

@node Simple Directory Lister Mark II
@subsection Simple Program to List a Directory, Mark II

Here is a revised version of the directory lister found above
(@pxref{Simple Directory Lister}).  Using the @code{scandir} function we
can avoid the functions which work directly with the directory contents.
After the call the returned entries are available for direct use.

@smallexample
@include dir2.c.texi
@end smallexample

Note the simple selector function in this example.  Since we want to see
all directory entries we always return @code{1}.

@node Low-level Directory Access
@subsection Low-level Directory Access

The stream-based directory functions are not AS-Safe and cannot be
used after @code{vfork}.  @xref{POSIX Safety Concepts}.  The functions
below provide an alternative that can be used in these contexts.

Directory data is obtained from a file descriptor, as created by the
@code{open} function, with or without the @code{O_DIRECTORY} flag.
@xref{Opening and Closing Files}.

@deftypefun ssize_t getdents64 (int @var{fd}, void *@var{buffer}, size_t @var{length})
@standards{Linux, dirent.h}
@safety{@prelim{}@mtsafe{}@assafe{}@acsafe{}}
The @code{getdents64} function reads at most @var{length} bytes of
directory entry data from the file descriptor @var{fd} and stores it
into the byte array starting at @var{buffer}.

On success, the function returns the number of bytes written to the
buffer.  This number is zero if @var{fd} is already at the end of the
directory stream.  On error, the function returns @code{-1} and sets
@code{errno} to the appropriate error code.

The data is stored as a sequence of @code{struct dirent64} records,
which can be traversed using the @code{d_reclen} member.  The buffer
should be large enough to hold the largest possible directory entry.
Note that some file systems support file names longer than
@code{NAME_MAX} bytes (e.g., because they support up to 255 Unicode
characters), so a buffer size of at least 1024 is recommended.

If the directory has been modified since the first call to
@code{getdents64} on the directory (opening the descriptor or seeking to
offset zero), it is possible that the buffer contains entries that have
been encountered before.  Likewise, it is possible that files that are
still present are not reported before the end of the directory is
encountered (and @code{getdents64} returns zero).

This function is specific to Linux.
@end deftypefun

Systems that support @code{getdents64} support seeking on directory
streams.  @xref{File Position Primitive}.  However, the only offset that
works reliably is offset zero, indicating that reading the directory
should start from the beginning.

@node Working with Directory Trees
@section Working with Directory Trees
@cindex directory hierarchy
@cindex hierarchy, directory
@cindex tree, directory

The functions described so far for handling the files in a directory
have allowed you to either retrieve the information bit by bit, or to
process all the files as a group (see @code{scandir}).  Sometimes it is
useful to process whole hierarchies of directories and their contained
files.  The X/Open specification defines two functions to do this.  The
simpler form is derived from an early definition in @w{System V} systems
and therefore this function is available on SVID-derived systems.  The
prototypes and required definitions can be found in the @file{ftw.h}
header.

There are four functions in this family: @code{ftw}, @code{nftw} and
their 64-bit counterparts @code{ftw64} and @code{nftw64}.  These
functions take as one of their arguments a pointer to a callback
function of the appropriate type.

@deftp {Data Type} __ftw_func_t
@standards{GNU, ftw.h}

@smallexample
int (*) (const char *, const struct stat *, int)
@end smallexample

The type of callback functions given to the @code{ftw} function.  The
first parameter points to the file name, the second parameter to an
object of type @code{struct stat} which is filled in for the file named
in the first parameter.

@noindent
The last parameter is a flag giving more information about the current
file.  It can have the following values:

@vtable @code
@item FTW_F
The item is either a normal file or a file which does not fit into one
of the following categories.  This could be special files, sockets etc.
@item FTW_D
The item is a directory.
@item FTW_NS
The @code{stat} call failed and so the information pointed to by the
second parameter is invalid.
@item FTW_DNR
The item is a directory which cannot be read.
@item FTW_SL
The item is a symbolic link.  Since symbolic links are normally followed
seeing this value in a @code{ftw} callback function means the referenced
file does not exist.  The situation for @code{nftw} is different.

This value is only available if the program is compiled with
@code{_XOPEN_EXTENDED} defined before including
the first header.  The original SVID systems do not have symbolic links.
@end vtable

If the sources are compiled with @code{_FILE_OFFSET_BITS == 64} this
type is in fact @code{__ftw64_func_t} since this mode changes
@code{struct stat} to be @code{struct stat64}.
@end deftp

For the LFS interface and for use in the function @code{ftw64}, the
header @file{ftw.h} defines another function type.

@deftp {Data Type} __ftw64_func_t
@standards{GNU, ftw.h}

@smallexample
int (*) (const char *, const struct stat64 *, int)
@end smallexample

This type is used just like @code{__ftw_func_t} for the callback
function, but this time is called from @code{ftw64}.  The second
parameter to the function is a pointer to a variable of type
@code{struct stat64} which is able to represent the larger values.
@end deftp

@deftp {Data Type} __nftw_func_t
@standards{GNU, ftw.h}

@smallexample
int (*) (const char *, const struct stat *, int, struct FTW *)
@end smallexample

The first three arguments are the same as for the @code{__ftw_func_t}
type.  However for the third argument some additional values are defined
to allow finer differentiation:
@vtable @code
@item FTW_DP
The current item is a directory and all subdirectories have already been
visited and reported.  This flag is returned instead of @code{FTW_D} if
the @code{FTW_DEPTH} flag is passed to @code{nftw} (see below).
@item FTW_SLN
The current item is a stale symbolic link.  The file it points to does
not exist.
@end vtable

The last parameter of the callback function is a pointer to a structure
with some extra information as described below.

If the sources are compiled with @code{_FILE_OFFSET_BITS == 64} this
type is in fact @code{__nftw64_func_t} since this mode changes
@code{struct stat} to be @code{struct stat64}.
@end deftp

For the LFS interface there is also a variant of this data type
available which has to be used with the @code{nftw64} function.

@deftp {Data Type} __nftw64_func_t
@standards{GNU, ftw.h}

@smallexample
int (*) (const char *, const struct stat64 *, int, struct FTW *)
@end smallexample

This type is used just like @code{__nftw_func_t} for the callback
function, but this time is called from @code{nftw64}.  The second
parameter to the function is this time a pointer to a variable of type
@code{struct stat64} which is able to represent the larger values.
@end deftp

@deftp {Data Type} {struct FTW}
@standards{XPG4.2, ftw.h}
The information contained in this structure helps in interpreting the
name parameter and gives some information about the current state of the
traversal of the directory hierarchy.

@table @code
@item int base
The value is the offset into the string passed in the first parameter to
the callback function of the beginning of the file name.  The rest of
the string is the path of the file.  This information is especially
important if the @code{FTW_CHDIR} flag was set in calling @code{nftw}
since then the current directory is the one the current item is found
in.
@item int level
Whilst processing, the code tracks how many directories down it has gone
to find the current file.  This nesting level starts at @math{0} for
files in the initial directory (or is zero for the initial file if a
file was passed).
@end table
@end deftp


@deftypefun int ftw (const char *@var{filename}, __ftw_func_t @var{func}, int @var{descriptors})
@standards{SVID, ftw.h}
@safety{@prelim{}@mtsafe{}@asunsafe{@ascuheap{}}@acunsafe{@acsmem{} @acsfd{}}}
@c see nftw for safety details
The @code{ftw} function calls the callback function given in the
parameter @var{func} for every item which is found in the directory
specified by @var{filename} and all directories below.  The function
follows symbolic links if necessary but does not process an item twice.
If @var{filename} is not a directory then it itself is the only object
returned to the callback function.

The file name passed to the callback function is constructed by taking
the @var{filename} parameter and appending the names of all passed
directories and then the local file name.  So the callback function can
use this parameter to access the file.  @code{ftw} also calls
@code{stat} for the file and passes that information on to the callback
function.  If this @code{stat} call is not successful the failure is
indicated by setting the third argument of the callback function to
@code{FTW_NS}.  Otherwise it is set according to the description given
in the account of @code{__ftw_func_t} above.

The callback function is expected to return @math{0} to indicate that no
error occurred and that processing should continue.  If an error
occurred in the callback function or it wants @code{ftw} to return
immediately, the callback function can return a value other than
@math{0}.  This is the only correct way to stop the function.  The
program must not use @code{setjmp} or similar techniques to continue
from another place.  This would leave resources allocated by the
@code{ftw} function unfreed.

The @var{descriptors} parameter to @code{ftw} specifies how many file
descriptors it is allowed to consume.  The function runs faster the more
descriptors it can use.  For each level in the directory hierarchy at
most one descriptor is used, but for very deep ones any limit on open
file descriptors for the process or the system may be exceeded.
Moreover, file descriptor limits in a multi-threaded program apply to
all the threads as a group, and therefore it is a good idea to supply a
reasonable limit to the number of open descriptors.

The return value of the @code{ftw} function is @math{0} if all callback
function calls returned @math{0} and all actions performed by the
@code{ftw} succeeded.  If a function call failed (other than calling
@code{stat} on an item) the function returns @math{-1}.  If a callback
function returns a value other than @math{0} this value is returned as
the return value of @code{ftw}.

When the sources are compiled with @code{_FILE_OFFSET_BITS == 64} on a
32-bit system this function is in fact @code{ftw64}, i.e., the LFS
interface transparently replaces the old interface.
@end deftypefun

@deftypefun int ftw64 (const char *@var{filename}, __ftw64_func_t @var{func}, int @var{descriptors})
@standards{Unix98, ftw.h}
@safety{@prelim{}@mtsafe{}@asunsafe{@ascuheap{}}@acunsafe{@acsmem{} @acsfd{}}}
This function is similar to @code{ftw} but it can work on filesystems
with large files.  File information is reported using a variable of type
@code{struct stat64} which is passed by reference to the callback
function.

When the sources are compiled with @code{_FILE_OFFSET_BITS == 64} on a
32-bit system this function is available under the name @code{ftw} and
transparently replaces the old implementation.
@end deftypefun

@deftypefun int nftw (const char *@var{filename}, __nftw_func_t @var{func}, int @var{descriptors}, int @var{flag})
@standards{XPG4.2, ftw.h}
@safety{@prelim{}@mtsafe{@mtasscwd{}}@asunsafe{@ascuheap{}}@acunsafe{@acsmem{} @acsfd{} @acscwd{}}}
@c ftw_startup calls alloca, malloc, free, xstat/lxstat, tdestroy, and ftw_dir
@c  if FTW_CHDIR, call open, and fchdir, or chdir and getcwd
@c ftw_dir calls open_dir_stream, readdir64, process_entry, closedir
@c  if FTW_CHDIR, also calls fchdir
@c open_dir_stream calls malloc, realloc, readdir64, free, closedir,
@c  then openat64_not_cancel_3 and fdopendir or opendir, then dirfd.
@c process_entry may cal realloc, fxstatat/lxstat/xstat, ftw_dir, and
@c  find_object (tsearch) and add_object (tfind).
@c Since each invocation of *ftw uses its own private search tree, none
@c  of the search tree concurrency issues apply.
The @code{nftw} function works like the @code{ftw} functions.  They call
the callback function @var{func} for all items found in the directory
@var{filename} and below.  At most @var{descriptors} file descriptors
are consumed during the @code{nftw} call.

One difference is that the callback function is of a different type.  It
is of type @w{@code{struct FTW *}} and provides the callback function
with the extra information described above.

A second difference is that @code{nftw} takes a fourth argument, which
is @math{0} or a bitwise-OR combination of any of the following values.

@vtable @code
@item FTW_PHYS
While traversing the directory symbolic links are not followed.  Instead
symbolic links are reported using the @code{FTW_SL} value for the type
parameter to the callback function.  If the file referenced by a
symbolic link does not exist @code{FTW_SLN} is returned instead.
@item FTW_MOUNT
The callback function is only called for items which are on the same
mounted filesystem as the directory given by the @var{filename}
parameter to @code{nftw}.
@item FTW_CHDIR
If this flag is given the current working directory is changed to the
directory of the reported object before the callback function is called.
When @code{ntfw} finally returns the current directory is restored to
its original value.
@item FTW_DEPTH
If this option is specified then all subdirectories and files within
them are processed before processing the top directory itself
(depth-first processing).  This also means the type flag given to the
callback function is @code{FTW_DP} and not @code{FTW_D}.
@item FTW_ACTIONRETVAL
If this option is specified then return values from callbacks
are handled differently.  If the callback returns @code{FTW_CONTINUE},
walking continues normally.  @code{FTW_STOP} means walking stops
and @code{FTW_STOP} is returned to the caller.  If @code{FTW_SKIP_SUBTREE}
is returned by the callback with @code{FTW_D} argument, the subtree
is skipped and walking continues with next sibling of the directory.
If @code{FTW_SKIP_SIBLINGS} is returned by the callback, all siblings
of the current entry are skipped and walking continues in its parent.
No other return values should be returned from the callbacks if
this option is set.  This option is a GNU extension.
@end vtable

The return value is computed in the same way as for @code{ftw}.
@code{nftw} returns @math{0} if no failures occurred and all callback
functions returned @math{0}.  In case of internal errors, such as memory
problems, the return value is @math{-1} and @code{errno} is set
accordingly.  If the return value of a callback invocation was non-zero
then that value is returned.

When the sources are compiled with @code{_FILE_OFFSET_BITS == 64} on a
32-bit system this function is in fact @code{nftw64}, i.e., the LFS
interface transparently replaces the old interface.
@end deftypefun

@deftypefun int nftw64 (const char *@var{filename}, __nftw64_func_t @var{func}, int @var{descriptors}, int @var{flag})
@standards{Unix98, ftw.h}
@safety{@prelim{}@mtsafe{@mtasscwd{}}@asunsafe{@ascuheap{}}@acunsafe{@acsmem{} @acsfd{} @acscwd{}}}
This function is similar to @code{nftw} but it can work on filesystems
with large files.  File information is reported using a variable of type
@code{struct stat64} which is passed by reference to the callback
function.

When the sources are compiled with @code{_FILE_OFFSET_BITS == 64} on a
32-bit system this function is available under the name @code{nftw} and
transparently replaces the old implementation.
@end deftypefun


@node Hard Links
@section Hard Links
@cindex hard link
@cindex link, hard
@cindex multiple names for one file
@cindex file names, multiple

In POSIX systems, one file can have many names at the same time.  All of
the names are equally real, and no one of them is preferred to the
others.

To add a name to a file, use the @code{link} function.  (The new name is
also called a @dfn{hard link} to the file.)  Creating a new link to a
file does not copy the contents of the file; it simply makes a new name
by which the file can be known, in addition to the file's existing name
or names.

One file can have names in several directories, so the organization
of the file system is not a strict hierarchy or tree.

In most implementations, it is not possible to have hard links to the
same file in multiple file systems.  @code{link} reports an error if you
try to make a hard link to the file from another file system when this
cannot be done.

The prototype for the @code{link} function is declared in the header
file @file{unistd.h}.
@pindex unistd.h

@deftypefun int link (const char *@var{oldname}, const char *@var{newname})
@standards{POSIX.1, unistd.h}
@safety{@prelim{}@mtsafe{}@assafe{}@acsafe{}}
The @code{link} function makes a new link to the existing file named by
@var{oldname}, under the new name @var{newname}.

This function returns a value of @code{0} if it is successful and
@code{-1} on failure.  In addition to the usual file name errors
(@pxref{File Name Errors}) for both @var{oldname} and @var{newname}, the
following @code{errno} error conditions are defined for this function:

@table @code
@item EACCES
You are not allowed to write to the directory in which the new link is
to be written.
@ignore
Some implementations also require that the existing file be accessible
by the caller, and use this error to report failure for that reason.
@end ignore

@item EEXIST
There is already a file named @var{newname}.  If you want to replace
this link with a new link, you must remove the old link explicitly first.

@item EMLINK
There are already too many links to the file named by @var{oldname}.
(The maximum number of links to a file is @w{@code{LINK_MAX}}; see
@ref{Limits for Files}.)

@item ENOENT
The file named by @var{oldname} doesn't exist.  You can't make a link to
a file that doesn't exist.

@item ENOSPC
The directory or file system that would contain the new link is full
and cannot be extended.

@item EPERM
On @gnulinuxhurdsystems{} and some others, you cannot make links to
directories.
Many systems allow only privileged users to do so.  This error
is used to report the problem.

@item EROFS
The directory containing the new link can't be modified because it's on
a read-only file system.

@item EXDEV
The directory specified in @var{newname} is on a different file system
than the existing file.

@item EIO
A hardware error occurred while trying to read or write the to filesystem.
@end table
@end deftypefun

@deftypefun int linkat (int oldfd, const char *@var{oldname}, int newfd, const char *@var{newname}, int flags)
@standards{POSIX.1, unistd.h}
@safety{@prelim{}@mtsafe{}@assafe{}@acsafe{}}

The @code{linkat} function is analogous to the @code{link} function,
except that it identifies its source and target using a combination of a
file descriptor (referring to a directory) and a file name.
@xref{Descriptor-Relative Access}.  For @code{linkat}, if a file name is
not absolute, it is resolved relative to the corresponding file
descriptor.  As usual, the special value @code{AT_FDCWD} denotes the
current directory.

The @var{flags} argument is a combination of the following flags:

@table @code
@item AT_SYMLINK_FOLLOW
If the source path identified by @var{oldfd} and @var{oldname} is a
symbolic link, @code{linkat} follows the symbolic link and creates a
link to its target.  If the flag is not set, a link for the symbolic
link itself is created; this is not supported by all file systems and
@code{linkat} can fail in this case.

@item AT_EMPTY_PATH
If this flag is specified, @var{oldname} can be an empty string.  In
this case, a new link to the file denoted by the descriptor @var{oldfd}
is created, which may have been opened with @code{O_PATH} or
@code{O_TMPFILE}.  This flag is a GNU extension.
@end table
@end deftypefun

@node Symbolic Links
@section Symbolic Links
@cindex soft link
@cindex link, soft
@cindex symbolic link
@cindex link, symbolic

@gnusystems{} support @dfn{soft links} or @dfn{symbolic links}.  This
is a kind of ``file'' that is essentially a pointer to another file
name.  Unlike hard links, symbolic links can be made to directories or
across file systems with no restrictions.  You can also make a symbolic
link to a name which is not the name of any file.  (Opening this link
will fail until a file by that name is created.)  Likewise, if the
symbolic link points to an existing file which is later deleted, the
symbolic link continues to point to the same file name even though the
name no longer names any file.

The reason symbolic links work the way they do is that special things
happen when you try to open the link.  The @code{open} function realizes
you have specified the name of a link, reads the file name contained in
the link, and opens that file name instead.  The @code{stat} function
likewise operates on the file that the symbolic link points to, instead
of on the link itself.

By contrast, other operations such as deleting or renaming the file
operate on the link itself.  The functions @code{readlink} and
@code{lstat} also refrain from following symbolic links, because their
purpose is to obtain information about the link.  @code{link}, the
function that makes a hard link, does too.  It makes a hard link to the
symbolic link, which one rarely wants.

Some systems have, for some functions operating on files, a limit on
how many symbolic links are followed when resolving a path name.  The
limit if it exists is published in the @file{sys/param.h} header file.

@deftypevr Macro int MAXSYMLINKS
@standards{BSD, sys/param.h}

The macro @code{MAXSYMLINKS} specifies how many symlinks some function
will follow before returning @code{ELOOP}.  Not all functions behave the
same and this value is not the same as that returned for
@code{_SC_SYMLOOP} by @code{sysconf}.  In fact, the @code{sysconf}
result can indicate that there is no fixed limit although
@code{MAXSYMLINKS} exists and has a finite value.
@end deftypevr

Prototypes for most of the functions listed in this section are in
@file{unistd.h}.
@pindex unistd.h

@deftypefun int symlink (const char *@var{oldname}, const char *@var{newname})
@standards{BSD, unistd.h}
@safety{@prelim{}@mtsafe{}@assafe{}@acsafe{}}
The @code{symlink} function makes a symbolic link to @var{oldname} named
@var{newname}.

The normal return value from @code{symlink} is @code{0}.  A return value
of @code{-1} indicates an error.  In addition to the usual file name
syntax errors (@pxref{File Name Errors}), the following @code{errno}
error conditions are defined for this function:

@table @code
@item EEXIST
There is already an existing file named @var{newname}.

@item EROFS
The file @var{newname} would exist on a read-only file system.

@item ENOSPC
The directory or file system cannot be extended to make the new link.

@item EIO
A hardware error occurred while reading or writing data on the disk.

@comment not sure about these
@ignore
@item ELOOP
There are too many levels of indirection.  This can be the result of
circular symbolic links to directories.

@item EDQUOT
The new link can't be created because the user's disk quota has been
exceeded.
@end ignore
@end table
@end deftypefun

@deftypefun ssize_t readlink (const char *@var{filename}, char *@var{buffer}, size_t @var{size})
@standards{BSD, unistd.h}
@safety{@prelim{}@mtsafe{}@assafe{}@acsafe{}}
The @code{readlink} function gets the value of the symbolic link
@var{filename}.  The file name that the link points to is copied into
@var{buffer}.  This file name string is @emph{not} null-terminated;
@code{readlink} normally returns the number of characters copied.  The
@var{size} argument specifies the maximum number of characters to copy,
usually the allocation size of @var{buffer}.

If the return value equals @var{size}, you cannot tell whether or not
there was room to return the entire name.  So make a bigger buffer and
call @code{readlink} again.  Here is an example:

@smallexample
char *
readlink_malloc (const char *filename)
@{
  size_t size = 50;
  char *buffer = NULL;

  while (1)
    @{
      buffer = xreallocarray (buffer, size, 2);
      size *= 2;
      ssize_t nchars = readlink (filename, buffer, size);
      if (nchars < 0)
        @{
          free (buffer);
          return NULL;
        @}
      if (nchars < size)
        return buffer;
    @}
@}
@end smallexample

@c @group  Invalid outside example.
A value of @code{-1} is returned in case of error.  In addition to the
usual file name errors (@pxref{File Name Errors}), the following
@code{errno} error conditions are defined for this function:

@table @code
@item EINVAL
The named file is not a symbolic link.

@item EIO
A hardware error occurred while reading or writing data on the disk.
@end table
@c @end group
@end deftypefun

In some situations it is desirable to resolve all the
symbolic links to get the real
name of a file where no prefix names a symbolic link which is followed
and no filename in the path is @code{.} or @code{..}.  This is for
instance desirable if files have to be compared in which case different
names can refer to the same inode.

@deftypefun {char *} canonicalize_file_name (const char *@var{name})
@standards{GNU, stdlib.h}
@safety{@prelim{}@mtsafe{}@asunsafe{@ascuheap{}}@acunsafe{@acsmem{} @acsfd{}}}
@c Calls realpath.

The @code{canonicalize_file_name} function returns the absolute name of
the file named by @var{name} which contains no @code{.}, @code{..}
components nor any repeated path separators (@code{/}) or symlinks.  The
result is passed back as the return value of the function in a block of
memory allocated with @code{malloc}.  If the result is not used anymore
the memory should be freed with a call to @code{free}.

If any of the path components are missing the function returns a NULL
pointer.  This is also what is returned if the length of the path
reaches or exceeds @code{PATH_MAX} characters.  In any case
@code{errno} is set accordingly.

@table @code
@item ENAMETOOLONG
The resulting path is too long.  This error only occurs on systems which
have a limit on the file name length.

@item EACCES
At least one of the path components is not readable.

@item ENOENT
The input file name is empty.

@item ENOENT
At least one of the path components does not exist.

@item ELOOP
More than @code{MAXSYMLINKS} many symlinks have been followed.
@end table

This function is a GNU extension and is declared in @file{stdlib.h}.
@end deftypefun

The Unix standard includes a similar function which differs from
@code{canonicalize_file_name} in that the user has to provide the buffer
where the result is placed in.

@deftypefun {char *} realpath (const char *restrict @var{name}, char *restrict @var{resolved})
@standards{XPG, stdlib.h}
@safety{@prelim{}@mtsafe{}@asunsafe{@ascuheap{}}@acunsafe{@acsmem{} @acsfd{}}}
@c Calls malloc, realloc, getcwd, lxstat64, readlink, alloca.

A call to @code{realpath} where the @var{resolved} parameter is
@code{NULL} behaves exactly like @code{canonicalize_file_name}.  The
function allocates a buffer for the file name and returns a pointer to
it.  If @var{resolved} is not @code{NULL} it points to a buffer into
which the result is copied.  It is the callers responsibility to
allocate a buffer which is large enough.  On systems which define
@code{PATH_MAX} this means the buffer must be large enough for a
pathname of this size.  For systems without limitations on the pathname
length the requirement cannot be met and programs should not call
@code{realpath} with anything but @code{NULL} for the second parameter.

One other difference is that the buffer @var{resolved} (if nonzero) will
contain the part of the path component which does not exist or is not
readable if the function returns @code{NULL} and @code{errno} is set to
@code{EACCES} or @code{ENOENT}.

This function is declared in @file{stdlib.h}.
@end deftypefun

The advantage of using this function is that it is more widely
available.  The drawback is that it reports failures for long paths on
systems which have no limits on the file name length.

@node Deleting Files
@section Deleting Files
@cindex deleting a file
@cindex removing a file
@cindex unlinking a file

You can delete a file with @code{unlink} or @code{remove}.

Deletion actually deletes a file name.  If this is the file's only name,
then the file is deleted as well.  If the file has other remaining names
(@pxref{Hard Links}), it remains accessible under those names.

@deftypefun int unlink (const char *@var{filename})
@standards{POSIX.1, unistd.h}
@safety{@prelim{}@mtsafe{}@assafe{}@acsafe{}}
The @code{unlink} function deletes the file name @var{filename}.  If
this is a file's sole name, the file itself is also deleted.  (Actually,
if any process has the file open when this happens, deletion is
postponed until all processes have closed the file.)

@pindex unistd.h
The function @code{unlink} is declared in the header file @file{unistd.h}.

This function returns @code{0} on successful completion, and @code{-1}
on error.  In addition to the usual file name errors
(@pxref{File Name Errors}), the following @code{errno} error conditions are
defined for this function:

@table @code
@item EACCES
Write permission is denied for the directory from which the file is to be
removed, or the directory has the sticky bit set and you do not own the file.

@item EBUSY
This error indicates that the file is being used by the system in such a
way that it can't be unlinked.  For example, you might see this error if
the file name specifies the root directory or a mount point for a file
system.

@item ENOENT
The file name to be deleted doesn't exist.

@item EPERM
On some systems @code{unlink} cannot be used to delete the name of a
directory, or at least can only be used this way by a privileged user.
To avoid such problems, use @code{rmdir} to delete directories.  (On
@gnulinuxhurdsystems{} @code{unlink} can never delete the name of a directory.)

@item EROFS
The directory containing the file name to be deleted is on a read-only
file system and can't be modified.
@end table
@end deftypefun

@deftypefun int unlinkat (int @var{filedes}, const char *@var{filename}, int @var{flags})
@standards{POSIX.1-2008, unistd.h}
@comment Unaudited and therefore marked AC-Unsafe and AS-Unsafe by default
@safety{@mtsafe{}@asunsafe{@asucorrupt{}}@acunsafe{@acucorrupt{}}}
This function is a descriptor-relative version of the @code{unlink}
function above.  @xref{Descriptor-Relative Access}.  The @var{flags}
argument may either be @code{0} or contain the flag @code{AT_REMOVEDIR}:

@table @code
@item AT_REMOVEDIR
This flag causes @code{unlinkat} to perform an @code{rmdir} operation on
@code{filename} instead of performing the equivalent of @code{unlink}.
@end table

Compared to @code{unlink}, some additional error conditions can occur due to
descriptor-relative access.  @xref{Descriptor-Relative Access}.  In
addition to this, the following other errors can also occur:

@table @code
@item EISDIR
The effective final path derived from @var{filename} and @var{filedes} is a
directory but @code{AT_REMOVEDIR} was not passed in @code{flags}.
@end table
@end deftypefun

@deftypefun int rmdir (const char *@var{filename})
@standards{POSIX.1, unistd.h}
@safety{@prelim{}@mtsafe{}@assafe{}@acsafe{}}
@cindex directories, deleting
@cindex deleting a directory
The @code{rmdir} function deletes a directory.  The directory must be
empty before it can be removed; in other words, it can only contain
entries for @file{.} and @file{..}.

In most other respects, @code{rmdir} behaves like @code{unlink}.  There
are two additional @code{errno} error conditions defined for
@code{rmdir}:

@table @code
@item ENOTEMPTY
@itemx EEXIST
The directory to be deleted is not empty.
@end table

These two error codes are synonymous; some systems use one, and some use
the other.  @gnulinuxhurdsystems{} always use @code{ENOTEMPTY}.

The prototype for this function is declared in the header file
@file{unistd.h}.
@pindex unistd.h
@end deftypefun

@deftypefun int remove (const char *@var{filename})
@standards{ISO, stdio.h}
@safety{@prelim{}@mtsafe{}@assafe{}@acsafe{}}
@c Calls unlink and rmdir.
This is the @w{ISO C} function to remove a file.  It works like
@code{unlink} for files and like @code{rmdir} for directories.
@code{remove} is declared in @file{stdio.h}.
@pindex stdio.h
@end deftypefun

@node Renaming Files
@section Renaming Files

The @code{rename} function is used to change a file's name.

@cindex renaming a file
@deftypefun int rename (const char *@var{oldname}, const char *@var{newname})
@standards{ISO, stdio.h}
@safety{@prelim{}@mtsafe{}@assafe{}@acsafe{}}
@c In the absence of a rename syscall, there's an emulation with link
@c and unlink, but it's racy, even more so if newname exists and is
@c unlinked first.
The @code{rename} function renames the file @var{oldname} to
@var{newname}.  The file formerly accessible under the name
@var{oldname} is afterwards accessible as @var{newname} instead.  (If
the file had any other names aside from @var{oldname}, it continues to
have those names.)

The directory containing the name @var{newname} must be on the same file
system as the directory containing the name @var{oldname}.

One special case for @code{rename} is when @var{oldname} and
@var{newname} are two names for the same file.  The consistent way to
handle this case is to delete @var{oldname}.  However, in this case
POSIX requires that @code{rename} do nothing and report success---which
is inconsistent.  We don't know what your operating system will do.

If @var{oldname} is not a directory, then any existing file named
@var{newname} is removed during the renaming operation.  However, if
@var{newname} is the name of a directory, @code{rename} fails in this
case.

If @var{oldname} is a directory, then either @var{newname} must not
exist or it must name a directory that is empty.  In the latter case,
the existing directory named @var{newname} is deleted first.  The name
@var{newname} must not specify a subdirectory of the directory
@code{oldname} which is being renamed.

One useful feature of @code{rename} is that the meaning of @var{newname}
changes ``atomically'' from any previously existing file by that name to
its new meaning (i.e., the file that was called @var{oldname}).  There is
no instant at which @var{newname} is non-existent ``in between'' the old
meaning and the new meaning.  If there is a system crash during the
operation, it is possible for both names to still exist; but
@var{newname} will always be intact if it exists at all.

If @code{rename} fails, it returns @code{-1}.  In addition to the usual
file name errors (@pxref{File Name Errors}), the following
@code{errno} error conditions are defined for this function:

@table @code
@item EACCES
One of the directories containing @var{newname} or @var{oldname}
refuses write permission; or @var{newname} and @var{oldname} are
directories and write permission is refused for one of them.

@item EBUSY
A directory named by @var{oldname} or @var{newname} is being used by
the system in a way that prevents the renaming from working.  This includes
directories that are mount points for filesystems, and directories
that are the current working directories of processes.

@item ENOTEMPTY
@itemx EEXIST
The directory @var{newname} isn't empty.  @gnulinuxhurdsystems{} always return
@code{ENOTEMPTY} for this, but some other systems return @code{EEXIST}.

@item EINVAL
@var{oldname} is a directory that contains @var{newname}.

@item EISDIR
@var{newname} is a directory but the @var{oldname} isn't.

@item EMLINK
The parent directory of @var{newname} would have too many links
(entries).

@item ENOENT
The file @var{oldname} doesn't exist.

@item ENOSPC
The directory that would contain @var{newname} has no room for another
entry, and there is no space left in the file system to expand it.

@item EROFS
The operation would involve writing to a directory on a read-only file
system.

@item EXDEV
The two file names @var{newname} and @var{oldname} are on different
file systems.
@end table
@end deftypefun

@deftypefun int renameat (int @var{oldfiledes}, const char *@var{oldname}, int @var{newfiledes}, const char *@var{newname})
@standards{POSIX.1-2008, stdio.h}
@comment Unaudited and therefore marked AC-Unsafe and AS-Unsafe by default
@safety{@mtsafe{}@asunsafe{@asucorrupt{}}@acunsafe{@acucorrupt{}}}
This function is a descriptor-relative version of the @code{rename}
function above.  @xref{Descriptor-Relative Access}.  If @var{oldname} or
@var{newname} is a relative path, it is interpreted relative to the
directory associated with @var{oldfiledes} or @var{newfiledes},
respectively.  Absolute paths are interpreted in the usual way.

Compared to @code{rename}, some additional error conditions can occur.
@xref{Descriptor-Relative Access}.
@end deftypefun

@node Creating Directories
@section Creating Directories
@cindex creating a directory
@cindex directories, creating

@pindex mkdir
Directories are created with the @code{mkdir} function.  (There is also
a shell command @code{mkdir} which does the same thing.)
@c !!! umask

@deftypefun int mkdir (const char *@var{filename}, mode_t @var{mode})
@standards{POSIX.1, sys/stat.h}
@safety{@prelim{}@mtsafe{}@assafe{}@acsafe{}}
The @code{mkdir} function creates a new, empty directory with name
@var{filename}.

The argument @var{mode} specifies the file permissions for the new
directory file.  @xref{Permission Bits}, for more information about
this.

A return value of @code{0} indicates successful completion, and
@code{-1} indicates failure.  In addition to the usual file name syntax
errors (@pxref{File Name Errors}), the following @code{errno} error
conditions are defined for this function:

@table @code
@item EACCES
Write permission is denied for the parent directory in which the new
directory is to be added.

@item EEXIST
A file named @var{filename} already exists.

@item EMLINK
The parent directory has too many links (entries).

Well-designed file systems never report this error, because they permit
more links than your disk could possibly hold.  However, you must still
take account of the possibility of this error, as it could result from
network access to a file system on another machine.

@item ENOSPC
The file system doesn't have enough room to create the new directory.

@item EROFS
The parent directory of the directory being created is on a read-only
file system and cannot be modified.
@end table

To use this function, your program should include the header file
@file{sys/stat.h}.
@pindex sys/stat.h
@end deftypefun

@deftypefun int mkdirat (int @var{filedes}, const char *@var{filename}, mode_t @var{mode})
@standards{POSIX.1-2008, sys/stat.h}
@comment Unaudited and therefore marked AC-Unsafe and AS-Unsafe by default
@safety{@mtsafe{}@asunsafe{@asucorrupt{}}@acunsafe{@acucorrupt{}}}
This function is a descriptor-relative version of the @code{mkdir}
function above.  @xref{Descriptor-Relative Access}.

Compared to @code{mkdir}, some additional error conditions can occur.
@xref{Descriptor-Relative Access}.
@end deftypefun

@node File Attributes
@section File Attributes

@pindex ls
When you issue an @samp{ls -l} shell command on a file, it gives you
information about the size of the file, who owns it, when it was last
modified, etc.  These are called the @dfn{file attributes}, and are
associated with the file itself and not a particular one of its names.

This section contains information about how you can inquire about and
modify the attributes of a file.

@menu
* Attribute Meanings::          The names of the file attributes,
                                 and what their values mean.
* Reading Attributes::          How to read the attributes of a file.
* Testing File Type::           Distinguishing ordinary files,
                                 directories, links@dots{}
* File Owner::                  How ownership for new files is determined,
			         and how to change it.
* Permission Bits::             How information about a file's access
                                 mode is stored.
* Access Permission::           How the system decides who can access a file.
* Setting Permissions::         How permissions for new files are assigned,
			         and how to change them.
* Testing File Access::         How to find out if your process can
                                 access a file.
* File Times::                  About the time attributes of a file.
* File Size::			Manually changing the size of a file.
* Storage Allocation::          Allocate backing storage for files.
@end menu

@node Attribute Meanings
@subsection The meaning of the File Attributes
@cindex status of a file
@cindex attributes of a file
@cindex file attributes

When you read the attributes of a file, they come back in a structure
called @code{struct stat}.  This section describes the names of the
attributes, their data types, and what they mean.  For the functions
to read the attributes of a file, see @ref{Reading Attributes}.

The header file @file{sys/stat.h} declares all the symbols defined
in this section.
@pindex sys/stat.h

@deftp {Data Type} {struct stat}
@standards{POSIX.1, sys/stat.h}
The @code{stat} structure type is used to return information about the
attributes of a file.  It contains at least the following members:

@table @code
@item mode_t st_mode
Specifies the mode of the file.  This includes file type information
(@pxref{Testing File Type}) and the file permission bits
(@pxref{Permission Bits}).

@item ino_t st_ino
The file serial number, which distinguishes this file from all other
files on the same device.

@item dev_t st_dev
Identifies the device containing the file.  The @code{st_ino} and
@code{st_dev}, taken together, uniquely identify the file.  The
@code{st_dev} value is not necessarily consistent across reboots or
system crashes, however.

@item nlink_t st_nlink
The number of hard links to the file.  This count keeps track of how
many directories have entries for this file.  If the count is ever
decremented to zero, then the file itself is discarded as soon as no
process still holds it open.  Symbolic links are not counted in the
total.

@item uid_t st_uid
The user ID of the file's owner.  @xref{File Owner}.

@item gid_t st_gid
The group ID of the file.  @xref{File Owner}.

@item off_t st_size
This specifies the size of a regular file in bytes.  For files that are
really devices this field isn't usually meaningful.  For symbolic links
this specifies the length of the file name the link refers to.

@item time_t st_atime
This is the last access time for the file.  @xref{File Times}.

@item unsigned long int st_atime_usec
This is the fractional part of the last access time for the file.
@xref{File Times}.

@item time_t st_mtime
This is the time of the last modification to the contents of the file.
@xref{File Times}.

@item unsigned long int st_mtime_usec
This is the fractional part of the time of the last modification to the
contents of the file.  @xref{File Times}.

@item time_t st_ctime
This is the time of the last modification to the attributes of the file.
@xref{File Times}.

@item unsigned long int st_ctime_usec
This is the fractional part of the time of the last modification to the
attributes of the file.  @xref{File Times}.

@c !!! st_rdev
@item blkcnt_t st_blocks
This is the amount of disk space that the file occupies, measured in
units of 512-byte blocks.

The number of disk blocks is not strictly proportional to the size of
the file, for two reasons: the file system may use some blocks for
internal record keeping; and the file may be sparse---it may have
``holes'' which contain zeros but do not actually take up space on the
disk.

You can tell (approximately) whether a file is sparse by comparing this
value with @code{st_size}, like this:

@smallexample
(st.st_blocks * 512 < st.st_size)
@end smallexample

This test is not perfect because a file that is just slightly sparse
might not be detected as sparse at all.  For practical applications,
this is not a problem.

@item unsigned int st_blksize
The optimal block size for reading or writing this file, in bytes.  You
might use this size for allocating the buffer space for reading or
writing the file.  (This is unrelated to @code{st_blocks}.)
@end table
@end deftp

The extensions for the Large File Support (LFS) require, even on 32-bit
machines, types which can handle file sizes up to @twoexp{63}.
Therefore a new definition of @code{struct stat} is necessary.

@deftp {Data Type} {struct stat64}
@standards{LFS, sys/stat.h}
The members of this type are the same and have the same names as those
in @code{struct stat}.  The only difference is that the members
@code{st_ino}, @code{st_size}, and @code{st_blocks} have a different
type to support larger values.

@table @code
@item mode_t st_mode
Specifies the mode of the file.  This includes file type information
(@pxref{Testing File Type}) and the file permission bits
(@pxref{Permission Bits}).

@item ino64_t st_ino
The file serial number, which distinguishes this file from all other
files on the same device.

@item dev_t st_dev
Identifies the device containing the file.  The @code{st_ino} and
@code{st_dev}, taken together, uniquely identify the file.  The
@code{st_dev} value is not necessarily consistent across reboots or
system crashes, however.

@item nlink_t st_nlink
The number of hard links to the file.  This count keeps track of how
many directories have entries for this file.  If the count is ever
decremented to zero, then the file itself is discarded as soon as no
process still holds it open.  Symbolic links are not counted in the
total.

@item uid_t st_uid
The user ID of the file's owner.  @xref{File Owner}.

@item gid_t st_gid
The group ID of the file.  @xref{File Owner}.

@item off64_t st_size
This specifies the size of a regular file in bytes.  For files that are
really devices this field isn't usually meaningful.  For symbolic links
this specifies the length of the file name the link refers to.

@item time_t st_atime
This is the last access time for the file.  @xref{File Times}.

@item unsigned long int st_atime_usec
This is the fractional part of the last access time for the file.
@xref{File Times}.

@item time_t st_mtime
This is the time of the last modification to the contents of the file.
@xref{File Times}.

@item unsigned long int st_mtime_usec
This is the fractional part of the time of the last modification to the
contents of the file.  @xref{File Times}.

@item time_t st_ctime
This is the time of the last modification to the attributes of the file.
@xref{File Times}.

@item unsigned long int st_ctime_usec
This is the fractional part of the time of the last modification to the
attributes of the file.  @xref{File Times}.

@c !!! st_rdev
@item blkcnt64_t st_blocks
This is the amount of disk space that the file occupies, measured in
units of 512-byte blocks.

@item unsigned int st_blksize
The optimal block size for reading of writing this file, in bytes.  You
might use this size for allocating the buffer space for reading of
writing the file.  (This is unrelated to @code{st_blocks}.)
@end table
@end deftp

Some of the file attributes have special data type names which exist
specifically for those attributes.  (They are all aliases for well-known
integer types that you know and love.)  These typedef names are defined
in the header file @file{sys/types.h} as well as in @file{sys/stat.h}.
Here is a list of them.

@deftp {Data Type} mode_t
@standards{POSIX.1, sys/types.h}
This is an integer data type used to represent file modes.  In
@theglibc{}, this is an unsigned type no narrower than @code{unsigned
int}.
@end deftp

@cindex inode number
@deftp {Data Type} ino_t
@standards{POSIX.1, sys/types.h}
This is an unsigned integer type used to represent file serial numbers.
(In Unix jargon, these are sometimes called @dfn{inode numbers}.)
In @theglibc{}, this type is no narrower than @code{unsigned int}.

If the source is compiled with @code{_FILE_OFFSET_BITS == 64} this type
is transparently replaced by @code{ino64_t}.
@end deftp

@deftp {Data Type} ino64_t
@standards{Unix98, sys/types.h}
This is an unsigned integer type used to represent file serial numbers
for the use in LFS.  In @theglibc{}, this type is no narrower than
@code{unsigned int}.

When compiling with @code{_FILE_OFFSET_BITS == 64} this type is
available under the name @code{ino_t}.
@end deftp

@deftp {Data Type} dev_t
@standards{POSIX.1, sys/types.h}
This is an arithmetic data type used to represent file device numbers.
In @theglibc{}, this is an integer type no narrower than @code{int}.
@end deftp

@deftp {Data Type} nlink_t
@standards{POSIX.1, sys/types.h}
This is an integer type used to represent file link counts.
@end deftp

@deftp {Data Type} blkcnt_t
@standards{Unix98, sys/types.h}
This is a signed integer type used to represent block counts.
In @theglibc{}, this type is no narrower than @code{int}.

If the source is compiled with @code{_FILE_OFFSET_BITS == 64} this type
is transparently replaced by @code{blkcnt64_t}.
@end deftp

@deftp {Data Type} blkcnt64_t
@standards{Unix98, sys/types.h}
This is a signed integer type used to represent block counts for the
use in LFS.  In @theglibc{}, this type is no narrower than @code{int}.

When compiling with @code{_FILE_OFFSET_BITS == 64} this type is
available under the name @code{blkcnt_t}.
@end deftp

@node Reading Attributes
@subsection Reading the Attributes of a File

To examine the attributes of files, use the functions @code{stat},
@code{fstat} and @code{lstat}.  They return the attribute information in
a @code{struct stat} object.  All three functions are declared in the
header file @file{sys/stat.h}.

@deftypefun int stat (const char *@var{filename}, struct stat *@var{buf})
@standards{POSIX.1, sys/stat.h}
@safety{@prelim{}@mtsafe{}@assafe{}@acsafe{}}
The @code{stat} function returns information about the attributes of the
file named by @w{@var{filename}} in the structure pointed to by @var{buf}.

If @var{filename} is the name of a symbolic link, the attributes you get
describe the file that the link points to.  If the link points to a
nonexistent file name, then @code{stat} fails reporting a nonexistent
file.

The return value is @code{0} if the operation is successful, or
@code{-1} on failure.  In addition to the usual file name errors
(@pxref{File Name Errors}, the following @code{errno} error conditions
are defined for this function:

@table @code
@item ENOENT
The file named by @var{filename} doesn't exist.
@end table

When the sources are compiled with @code{_FILE_OFFSET_BITS == 64} this
function is in fact @code{stat64} since the LFS interface transparently
replaces the normal implementation.
@end deftypefun

@deftypefun int stat64 (const char *@var{filename}, struct stat64 *@var{buf})
@standards{Unix98, sys/stat.h}
@safety{@prelim{}@mtsafe{}@assafe{}@acsafe{}}
This function is similar to @code{stat} but it is also able to work on
files larger than @twoexp{31} bytes on 32-bit systems.  To be able to do
this the result is stored in a variable of type @code{struct stat64} to
which @var{buf} must point.

When the sources are compiled with @code{_FILE_OFFSET_BITS == 64} this
function is available under the name @code{stat} and so transparently
replaces the interface for small files on 32-bit machines.
@end deftypefun

@deftypefun int fstat (int @var{filedes}, struct stat *@var{buf})
@standards{POSIX.1, sys/stat.h}
@safety{@prelim{}@mtsafe{}@assafe{}@acsafe{}}
The @code{fstat} function is like @code{stat}, except that it takes an
open file descriptor as an argument instead of a file name.
@xref{Low-Level I/O}.

Like @code{stat}, @code{fstat} returns @code{0} on success and @code{-1}
on failure.  The following @code{errno} error conditions are defined for
@code{fstat}:

@table @code
@item EBADF
The @var{filedes} argument is not a valid file descriptor.
@end table

When the sources are compiled with @code{_FILE_OFFSET_BITS == 64} this
function is in fact @code{fstat64} since the LFS interface transparently
replaces the normal implementation.
@end deftypefun

@deftypefun int fstat64 (int @var{filedes}, struct stat64 *@var{buf})
@standards{Unix98, sys/stat.h}
@safety{@prelim{}@mtsafe{}@assafe{}@acsafe{}}
This function is similar to @code{fstat} but is able to work on large
files on 32-bit platforms.  For large files the file descriptor
@var{filedes} should be obtained by @code{open64} or @code{creat64}.
The @var{buf} pointer points to a variable of type @code{struct stat64}
which is able to represent the larger values.

When the sources are compiled with @code{_FILE_OFFSET_BITS == 64} this
function is available under the name @code{fstat} and so transparently
replaces the interface for small files on 32-bit machines.
@end deftypefun

@deftypefun int fstatat (int @var{filedes}, const char *@var{filename}, struct stat *@var{buf}, int @var{flags})
@standards{POSIX.1, sys/stat.h}
@safety{@prelim{}@mtsafe{}@assafe{}@acsafe{}}
This function is a descriptor-relative version of the @code{fstat}
function above.  @xref{Descriptor-Relative Access}.  The @var{flags}
argument can contain a combination of the flags @code{AT_EMPTY_PATH},
@code{AT_NO_AUTOMOUNT}, @code{AT_SYMLINK_NOFOLLOW}.

Compared to @code{fstat}, the following additional error conditions can
occur:

@table @code
@item EBADF
The @var{filedes} argument is not a valid file descriptor.

@item EINVAL
The @var{flags} argument is not valid for this function.

@item ENOTDIR
The descriptor @var{filedes} is not associated with a directory, and
@var{filename} is a relative file name.

@item ENOENT
The file named by @var{filename} does not exist, it's a dangling symbolic link
and @var{flags} does not contain @code{AT_SYMLINK_NOFOLLOW}, or @var{filename}
is an empty string and @var{flags} does not contain @code{AT_EMPTY_PATH}.
@end table

When the sources are compiled with @code{_FILE_OFFSET_BITS == 64} this
function is in fact @code{fstatat64} since the LFS interface transparently
replaces the normal implementation.
@end deftypefun

@deftypefun int fstatat64 (int @var{filedes}, const char *@var{filename}, struct stat64 *@var{buf}, int @var{flags})
@standards{GNU, sys/stat.h}
@safety{@prelim{}@mtsafe{}@assafe{}@acsafe{}}
This function is the large-file variant of @code{fstatat}, similar to
how @code{fstat64} is the variant of @code{fstat}.

When the sources are compiled with @code{_FILE_OFFSET_BITS == 64} this
function is available under the name @code{fstatat} and so transparently
replaces the interface for small files on 32-bit machines.
@end deftypefun

@deftypefun int lstat (const char *@var{filename}, struct stat *@var{buf})
@standards{BSD, sys/stat.h}
@safety{@prelim{}@mtsafe{}@assafe{}@acsafe{}}
@c Direct system call through lxstat, sometimes with an xstat conv call
@c afterwards.
The @code{lstat} function is like @code{stat}, except that it does not
follow symbolic links.  If @var{filename} is the name of a symbolic
link, @code{lstat} returns information about the link itself; otherwise
@code{lstat} works like @code{stat}.  @xref{Symbolic Links}.

When the sources are compiled with @code{_FILE_OFFSET_BITS == 64} this
function is in fact @code{lstat64} since the LFS interface transparently
replaces the normal implementation.
@end deftypefun

@deftypefun int lstat64 (const char *@var{filename}, struct stat64 *@var{buf})
@standards{Unix98, sys/stat.h}
@safety{@prelim{}@mtsafe{}@assafe{}@acsafe{}}
@c Direct system call through lxstat64, sometimes with an xstat conv
@c call afterwards.
This function is similar to @code{lstat} but it is also able to work on
files larger than @twoexp{31} bytes on 32-bit systems.  To be able to do
this the result is stored in a variable of type @code{struct stat64} to
which @var{buf} must point.

When the sources are compiled with @code{_FILE_OFFSET_BITS == 64} this
function is available under the name @code{lstat} and so transparently
replaces the interface for small files on 32-bit machines.
@end deftypefun

@node Testing File Type
@subsection Testing the Type of a File

The @dfn{file mode}, stored in the @code{st_mode} field of the file
attributes, contains two kinds of information: the file type code, and
the access permission bits.  This section discusses only the type code,
which you can use to tell whether the file is a directory, socket,
symbolic link, and so on.  For details about access permissions see
@ref{Permission Bits}.

There are two ways you can access the file type information in a file
mode.  Firstly, for each file type there is a @dfn{predicate macro}
which examines a given file mode and returns whether it is of that type
or not.  Secondly, you can mask out the rest of the file mode to leave
just the file type code, and compare this against constants for each of
the supported file types.

All of the symbols listed in this section are defined in the header file
@file{sys/stat.h}.
@pindex sys/stat.h

The following predicate macros test the type of a file, given the value
@var{m} which is the @code{st_mode} field returned by @code{stat} on
that file:

@deftypefn Macro int S_ISDIR (mode_t @var{m})
@standards{POSIX, sys/stat.h}
@safety{@prelim{}@mtsafe{}@assafe{}@acsafe{}}
This macro returns non-zero if the file is a directory.
@end deftypefn

@deftypefn Macro int S_ISCHR (mode_t @var{m})
@standards{POSIX, sys/stat.h}
@safety{@prelim{}@mtsafe{}@assafe{}@acsafe{}}
This macro returns non-zero if the file is a character special file (a
device like a terminal).
@end deftypefn

@deftypefn Macro int S_ISBLK (mode_t @var{m})
@standards{POSIX, sys/stat.h}
@safety{@prelim{}@mtsafe{}@assafe{}@acsafe{}}
This macro returns non-zero if the file is a block special file (a device
like a disk).
@end deftypefn

@deftypefn Macro int S_ISREG (mode_t @var{m})
@standards{POSIX, sys/stat.h}
@safety{@prelim{}@mtsafe{}@assafe{}@acsafe{}}
This macro returns non-zero if the file is a regular file.
@end deftypefn

@deftypefn Macro int S_ISFIFO (mode_t @var{m})
@standards{POSIX, sys/stat.h}
@safety{@prelim{}@mtsafe{}@assafe{}@acsafe{}}
This macro returns non-zero if the file is a FIFO special file, or a
pipe.  @xref{Pipes and FIFOs}.
@end deftypefn

@deftypefn Macro int S_ISLNK (mode_t @var{m})
@standards{GNU, sys/stat.h}
@safety{@prelim{}@mtsafe{}@assafe{}@acsafe{}}
This macro returns non-zero if the file is a symbolic link.
@xref{Symbolic Links}.
@end deftypefn

@deftypefn Macro int S_ISSOCK (mode_t @var{m})
@standards{GNU, sys/stat.h}
@safety{@prelim{}@mtsafe{}@assafe{}@acsafe{}}
This macro returns non-zero if the file is a socket.  @xref{Sockets}.
@end deftypefn

An alternate non-POSIX method of testing the file type is supported for
compatibility with BSD.  The mode can be bitwise AND-ed with
@code{S_IFMT} to extract the file type code, and compared to the
appropriate constant.  For example,

@smallexample
S_ISCHR (@var{mode})
@end smallexample

@noindent
is equivalent to:

@smallexample
((@var{mode} & S_IFMT) == S_IFCHR)
@end smallexample

@deftypevr Macro int S_IFMT
@standards{BSD, sys/stat.h}
This is a bit mask used to extract the file type code from a mode value.
@end deftypevr

These are the symbolic names for the different file type codes:

@vtable @code
@item S_IFDIR
@standards{BSD, sys/stat.h}
This is the file type constant of a directory file.

@item S_IFCHR
@standards{BSD, sys/stat.h}
This is the file type constant of a character-oriented device file.

@item S_IFBLK
@standards{BSD, sys/stat.h}
This is the file type constant of a block-oriented device file.

@item S_IFREG
@standards{BSD, sys/stat.h}
This is the file type constant of a regular file.

@item S_IFLNK
@standards{BSD, sys/stat.h}
This is the file type constant of a symbolic link.

@item S_IFSOCK
@standards{BSD, sys/stat.h}
This is the file type constant of a socket.

@item S_IFIFO
@standards{BSD, sys/stat.h}
This is the file type constant of a FIFO or pipe.
@end vtable

The POSIX.1b standard introduced a few more objects which possibly can
be implemented as objects in the filesystem.  These are message queues,
semaphores, and shared memory objects.  To allow differentiating these
objects from other files the POSIX standard introduced three new test
macros.  But unlike the other macros they do not take the value of the
@code{st_mode} field as the parameter.  Instead they expect a pointer to
the whole @code{struct stat} structure.

@deftypefn Macro int S_TYPEISMQ (struct stat *@var{s})
@standards{POSIX, sys/stat.h}
@safety{@prelim{}@mtsafe{}@assafe{}@acsafe{}}
If the system implements POSIX message queues as distinct objects and the
file is a message queue object, this macro returns a non-zero value.
In all other cases the result is zero.
@end deftypefn

@deftypefn Macro int S_TYPEISSEM (struct stat *@var{s})
@standards{POSIX, sys/stat.h}
@safety{@prelim{}@mtsafe{}@assafe{}@acsafe{}}
If the system implements POSIX semaphores as distinct objects and the
file is a semaphore object, this macro returns a non-zero value.
In all other cases the result is zero.
@end deftypefn

@deftypefn Macro int S_TYPEISSHM (struct stat *@var{s})
@standards{POSIX, sys/stat.h}
@safety{@prelim{}@mtsafe{}@assafe{}@acsafe{}}
If the system implements POSIX shared memory objects as distinct objects
and the file is a shared memory object, this macro returns a non-zero
value.  In all other cases the result is zero.
@end deftypefn

@node File Owner
@subsection File Owner
@cindex file owner
@cindex owner of a file
@cindex group owner of a file

Every file has an @dfn{owner} which is one of the registered user names
defined on the system.  Each file also has a @dfn{group} which is one of
the defined groups.  The file owner can often be useful for showing you
who edited the file (especially when you edit with GNU Emacs), but its
main purpose is for access control.

The file owner and group play a role in determining access because the
file has one set of access permission bits for the owner, another set
that applies to users who belong to the file's group, and a third set of
bits that applies to everyone else.  @xref{Access Permission}, for the
details of how access is decided based on this data.

When a file is created, its owner is set to the effective user ID of the
process that creates it (@pxref{Process Persona}).  The file's group ID
may be set to either the effective group ID of the process, or the group
ID of the directory that contains the file, depending on the system
where the file is stored.  When you access a remote file system, it
behaves according to its own rules, not according to the system your
program is running on.  Thus, your program must be prepared to encounter
either kind of behavior no matter what kind of system you run it on.

@pindex chown
@pindex chgrp
You can change the owner and/or group owner of an existing file using
the @code{chown} function.  This is the primitive for the @code{chown}
and @code{chgrp} shell commands.

@pindex unistd.h
The prototype for this function is declared in @file{unistd.h}.

@deftypefun int chown (const char *@var{filename}, uid_t @var{owner}, gid_t @var{group})
@standards{POSIX.1, unistd.h}
@safety{@prelim{}@mtsafe{}@assafe{}@acsafe{}}
The @code{chown} function changes the owner of the file @var{filename} to
@var{owner}, and its group owner to @var{group}.

Changing the owner of the file on certain systems clears the set-user-ID
and set-group-ID permission bits.  (This is because those bits may not
be appropriate for the new owner.)  Other file permission bits are not
changed.

The return value is @code{0} on success and @code{-1} on failure.
In addition to the usual file name errors (@pxref{File Name Errors}),
the following @code{errno} error conditions are defined for this function:

@table @code
@item EPERM
This process lacks permission to make the requested change.

Only privileged users or the file's owner can change the file's group.
On most file systems, only privileged users can change the file owner;
some file systems allow you to change the owner if you are currently the
owner.  When you access a remote file system, the behavior you encounter
is determined by the system that actually holds the file, not by the
system your program is running on.

@xref{Options for Files}, for information about the
@code{_POSIX_CHOWN_RESTRICTED} macro.

@item EROFS
The file is on a read-only file system.
@end table
@end deftypefun

@deftypefun int fchown (int @var{filedes}, uid_t @var{owner}, gid_t @var{group})
@standards{BSD, unistd.h}
@safety{@prelim{}@mtsafe{}@assafe{}@acsafe{}}
This is like @code{chown}, except that it changes the owner of the open
file with descriptor @var{filedes}.

The return value from @code{fchown} is @code{0} on success and @code{-1}
on failure.  The following @code{errno} error codes are defined for this
function:

@table @code
@item EBADF
The @var{filedes} argument is not a valid file descriptor.

@item EINVAL
The @var{filedes} argument corresponds to a pipe or socket, not an ordinary
file.

@item EPERM
This process lacks permission to make the requested change.  For details
see @code{chmod} above.

@item EROFS
The file resides on a read-only file system.
@end table
@end deftypefun

@node Permission Bits
@subsection The Mode Bits for Access Permission

The @dfn{file mode}, stored in the @code{st_mode} field of the file
attributes, contains two kinds of information: the file type code, and
the access permission bits.  This section discusses only the access
permission bits, which control who can read or write the file.
@xref{Testing File Type}, for information about the file type code.

All of the symbols listed in this section are defined in the header file
@file{sys/stat.h}.
@pindex sys/stat.h

@cindex file permission bits
These symbolic constants are defined for the file mode bits that control
access permission for the file:

@vtable @code
@item S_IRUSR
@itemx S_IREAD
@standards{POSIX.1, sys/stat.h}
@standardsx{S_IREAD, BSD, sys/stat.h}
Read permission bit for the owner of the file.  On many systems this bit
is 0400.  @code{S_IREAD} is an obsolete synonym provided for BSD
compatibility.

@item S_IWUSR
@itemx S_IWRITE
@standards{POSIX.1, sys/stat.h}
@standardsx{S_IWRITE, BSD, sys/stat.h}
Write permission bit for the owner of the file.  Usually 0200.
@w{@code{S_IWRITE}} is an obsolete synonym provided for BSD compatibility.

@item S_IXUSR
@itemx S_IEXEC
@standards{POSIX.1, sys/stat.h}
@standardsx{S_IEXEC, BSD, sys/stat.h}
Execute (for ordinary files) or search (for directories) permission bit
for the owner of the file.  Usually 0100.  @code{S_IEXEC} is an obsolete
synonym provided for BSD compatibility.

@item S_IRWXU
@standards{POSIX.1, sys/stat.h}
This is equivalent to @samp{(S_IRUSR | S_IWUSR | S_IXUSR)}.

@item S_IRGRP
@standards{POSIX.1, sys/stat.h}
Read permission bit for the group owner of the file.  Usually 040.

@item S_IWGRP
@standards{POSIX.1, sys/stat.h}
Write permission bit for the group owner of the file.  Usually 020.

@item S_IXGRP
@standards{POSIX.1, sys/stat.h}
Execute or search permission bit for the group owner of the file.
Usually 010.

@item S_IRWXG
@standards{POSIX.1, sys/stat.h}
This is equivalent to @samp{(S_IRGRP | S_IWGRP | S_IXGRP)}.

@item S_IROTH
@standards{POSIX.1, sys/stat.h}
Read permission bit for other users.  Usually 04.

@item S_IWOTH
@standards{POSIX.1, sys/stat.h}
Write permission bit for other users.  Usually 02.

@item S_IXOTH
@standards{POSIX.1, sys/stat.h}
Execute or search permission bit for other users.  Usually 01.

@item S_IRWXO
@standards{POSIX.1, sys/stat.h}
This is equivalent to @samp{(S_IROTH | S_IWOTH | S_IXOTH)}.

@item S_ISUID
@standards{POSIX, sys/stat.h}
This is the set-user-ID on execute bit, usually 04000.
@xref{How Change Persona}.

@item S_ISGID
@standards{POSIX, sys/stat.h}
This is the set-group-ID on execute bit, usually 02000.
@xref{How Change Persona}.

@cindex sticky bit
@item S_ISVTX
@standards{BSD, sys/stat.h}
This is the @dfn{sticky} bit, usually 01000.

For a directory it gives permission to delete a file in that directory
only if you own that file.  Ordinarily, a user can either delete all the
files in a directory or cannot delete any of them (based on whether the
user has write permission for the directory).  The same restriction
applies---you must have both write permission for the directory and own
the file you want to delete.  The one exception is that the owner of the
directory can delete any file in the directory, no matter who owns it
(provided the owner has given himself write permission for the
directory).  This is commonly used for the @file{/tmp} directory, where
anyone may create files but not delete files created by other users.

Originally the sticky bit on an executable file modified the swapping
policies of the system.  Normally, when a program terminated, its pages
in core were immediately freed and reused.  If the sticky bit was set on
the executable file, the system kept the pages in core for a while as if
the program were still running.  This was advantageous for a program
likely to be run many times in succession.  This usage is obsolete in
modern systems.  When a program terminates, its pages always remain in
core as long as there is no shortage of memory in the system.  When the
program is next run, its pages will still be in core if no shortage
arose since the last run.

On some modern systems where the sticky bit has no useful meaning for an
executable file, you cannot set the bit at all for a non-directory.
If you try, @code{chmod} fails with @code{EFTYPE};
@pxref{Setting Permissions}.

Some systems (particularly SunOS) have yet another use for the sticky
bit.  If the sticky bit is set on a file that is @emph{not} executable,
it means the opposite: never cache the pages of this file at all.  The
main use of this is for the files on an NFS server machine which are
used as the swap area of diskless client machines.  The idea is that the
pages of the file will be cached in the client's memory, so it is a
waste of the server's memory to cache them a second time.  With this
usage the sticky bit also implies that the filesystem may fail to record
the file's modification time onto disk reliably (the idea being that
no-one cares for a swap file).

This bit is only available on BSD systems (and those derived from
them).  Therefore one has to use the @code{_GNU_SOURCE} feature select
macro, or not define any feature test macros, to get the definition
(@pxref{Feature Test Macros}).
@end vtable

The actual bit values of the symbols are listed in the table above
so you can decode file mode values when debugging your programs.
These bit values are correct for most systems, but they are not
guaranteed.

@strong{Warning:} Writing explicit numbers for file permissions is bad
practice.  Not only is it not portable, it also requires everyone who
reads your program to remember what the bits mean.  To make your program
clean use the symbolic names.

@node Access Permission
@subsection How Your Access to a File is Decided
@cindex permission to access a file
@cindex access permission for a file
@cindex file access permission

Recall that the operating system normally decides access permission for
a file based on the effective user and group IDs of the process and its
supplementary group IDs, together with the file's owner, group and
permission bits.  These concepts are discussed in detail in @ref{Process
Persona}.

If the effective user ID of the process matches the owner user ID of the
file, then permissions for read, write, and execute/search are
controlled by the corresponding ``user'' (or ``owner'') bits.  Likewise,
if any of the effective group ID or supplementary group IDs of the
process matches the group owner ID of the file, then permissions are
controlled by the ``group'' bits.  Otherwise, permissions are controlled
by the ``other'' bits.

Privileged users, like @samp{root}, can access any file regardless of
its permission bits.  As a special case, for a file to be executable
even by a privileged user, at least one of its execute bits must be set.

@node Setting Permissions
@subsection Assigning File Permissions

@cindex file creation mask
@cindex umask
The primitive functions for creating files (for example, @code{open} or
@code{mkdir}) take a @var{mode} argument, which specifies the file
permissions to give the newly created file.  This mode is modified by
the process's @dfn{file creation mask}, or @dfn{umask}, before it is
used.

The bits that are set in the file creation mask identify permissions
that are always to be disabled for newly created files.  For example, if
you set all the ``other'' access bits in the mask, then newly created
files are not accessible at all to processes in the ``other'' category,
even if the @var{mode} argument passed to the create function would
permit such access.  In other words, the file creation mask is the
complement of the ordinary access permissions you want to grant.

Programs that create files typically specify a @var{mode} argument that
includes all the permissions that make sense for the particular file.
For an ordinary file, this is typically read and write permission for
all classes of users.  These permissions are then restricted as
specified by the individual user's own file creation mask.

@findex chmod
To change the permission of an existing file given its name, call
@code{chmod}.  This function uses the specified permission bits and
ignores the file creation mask.

@pindex umask
In normal use, the file creation mask is initialized by the user's login
shell (using the @code{umask} shell command), and inherited by all
subprocesses.  Application programs normally don't need to worry about
the file creation mask.  It will automatically do what it is supposed to
do.

When your program needs to create a file and bypass the umask for its
access permissions, the easiest way to do this is to use @code{fchmod}
after opening the file, rather than changing the umask.  In fact,
changing the umask is usually done only by shells.  They use the
@code{umask} function.

The functions in this section are declared in @file{sys/stat.h}.
@pindex sys/stat.h

@deftypefun mode_t umask (mode_t @var{mask})
@standards{POSIX.1, sys/stat.h}
@safety{@prelim{}@mtsafe{}@assafe{}@acsafe{}}
The @code{umask} function sets the file creation mask of the current
process to @var{mask}, and returns the previous value of the file
creation mask.

Here is an example showing how to read the mask with @code{umask}
without changing it permanently:

@smallexample
mode_t
read_umask (void)
@{
  mode_t mask = umask (0);
  umask (mask);
  return mask;
@}
@end smallexample

@noindent
However, on @gnuhurdsystems{} it is better to use @code{getumask} if
you just want to read the mask value, because it is reentrant.
@end deftypefun

@deftypefun mode_t getumask (void)
@standards{GNU, sys/stat.h}
@safety{@prelim{}@mtsafe{}@assafe{}@acsafe{}}
Return the current value of the file creation mask for the current
process.  This function is a GNU extension and is only available on
@gnuhurdsystems{}.
@end deftypefun

@deftypefun int chmod (const char *@var{filename}, mode_t @var{mode})
@standards{POSIX.1, sys/stat.h}
@safety{@prelim{}@mtsafe{}@assafe{}@acsafe{}}
The @code{chmod} function sets the access permission bits for the file
named by @var{filename} to @var{mode}.

If @var{filename} is a symbolic link, @code{chmod} changes the
permissions of the file pointed to by the link, not those of the link
itself.

This function returns @code{0} if successful and @code{-1} if not.  In
addition to the usual file name errors (@pxref{File Name
Errors}), the following @code{errno} error conditions are defined for
this function:

@table @code
@item ENOENT
The named file doesn't exist.

@item EPERM
This process does not have permission to change the access permissions
of this file.  Only the file's owner (as judged by the effective user ID
of the process) or a privileged user can change them.

@item EROFS
The file resides on a read-only file system.

@item EFTYPE
@var{mode} has the @code{S_ISVTX} bit (the ``sticky bit'') set,
and the named file is not a directory.  Some systems do not allow setting the
sticky bit on non-directory files, and some do (and only some of those
assign a useful meaning to the bit for non-directory files).

You only get @code{EFTYPE} on systems where the sticky bit has no useful
meaning for non-directory files, so it is always safe to just clear the
bit in @var{mode} and call @code{chmod} again.  @xref{Permission Bits},
for full details on the sticky bit.
@end table
@end deftypefun

@deftypefun int fchmod (int @var{filedes}, mode_t @var{mode})
@standards{BSD, sys/stat.h}
@safety{@prelim{}@mtsafe{}@assafe{}@acsafe{}}
This is like @code{chmod}, except that it changes the permissions of the
currently open file given by @var{filedes}.

The return value from @code{fchmod} is @code{0} on success and @code{-1}
on failure.  The following @code{errno} error codes are defined for this
function:

@table @code
@item EBADF
The @var{filedes} argument is not a valid file descriptor.

@item EINVAL
The @var{filedes} argument corresponds to a pipe or socket, or something
else that doesn't really have access permissions.

@item EPERM
This process does not have permission to change the access permissions
of this file.  Only the file's owner (as judged by the effective user ID
of the process) or a privileged user can change them.

@item EROFS
The file resides on a read-only file system.
@end table
@end deftypefun

@node Testing File Access
@subsection Testing Permission to Access a File
@cindex testing access permission
@cindex access, testing for
@cindex setuid programs and file access

In some situations it is desirable to allow programs to access files or
devices even if this is not possible with the permissions granted to the
user.  One possible solution is to set the setuid-bit of the program
file.  If such a program is started the @emph{effective} user ID of the
process is changed to that of the owner of the program file.  So to
allow write access to files like @file{/etc/passwd}, which normally can
be written only by the super-user, the modifying program will have to be
owned by @code{root} and the setuid-bit must be set.

But besides the files the program is intended to change the user should
not be allowed to access any file to which s/he would not have access
anyway.  The program therefore must explicitly check whether @emph{the
user} would have the necessary access to a file, before it reads or
writes the file.

To do this, use the function @code{access}, which checks for access
permission based on the process's @emph{real} user ID rather than the
effective user ID.  (The setuid feature does not alter the real user ID,
so it reflects the user who actually ran the program.)

There is another way you could check this access, which is easy to
describe, but very hard to use.  This is to examine the file mode bits
and mimic the system's own access computation.  This method is
undesirable because many systems have additional access control
features; your program cannot portably mimic them, and you would not
want to try to keep track of the diverse features that different systems
have.  Using @code{access} is simple and automatically does whatever is
appropriate for the system you are using.

@code{access} is @emph{only} appropriate to use in setuid programs.
A non-setuid program will always use the effective ID rather than the
real ID.

@pindex unistd.h
The symbols in this section are declared in @file{unistd.h}.

@deftypefun int access (const char *@var{filename}, int @var{how})
@standards{POSIX.1, unistd.h}
@safety{@prelim{}@mtsafe{}@assafe{}@acsafe{}}
The @code{access} function checks to see whether the file named by
@var{filename} can be accessed in the way specified by the @var{how}
argument.  The @var{how} argument either can be the bitwise OR of the
flags @code{R_OK}, @code{W_OK}, @code{X_OK}, or the existence test
@code{F_OK}.

This function uses the @emph{real} user and group IDs of the calling
process, rather than the @emph{effective} IDs, to check for access
permission.  As a result, if you use the function from a @code{setuid}
or @code{setgid} program (@pxref{How Change Persona}), it gives
information relative to the user who actually ran the program.

The return value is @code{0} if the access is permitted, and @code{-1}
otherwise.  (In other words, treated as a predicate function,
@code{access} returns true if the requested access is @emph{denied}.)

In addition to the usual file name errors (@pxref{File Name
Errors}), the following @code{errno} error conditions are defined for
this function:

@table @code
@item EACCES
The access specified by @var{how} is denied.

@item ENOENT
The file doesn't exist.

@item EROFS
Write permission was requested for a file on a read-only file system.
@end table
@end deftypefun

These macros are defined in the header file @file{unistd.h} for use
as the @var{how} argument to the @code{access} function.  The values
are integer constants.
@pindex unistd.h

@deftypevr Macro int R_OK
@standards{POSIX.1, unistd.h}
Flag meaning test for read permission.
@end deftypevr

@deftypevr Macro int W_OK
@standards{POSIX.1, unistd.h}
Flag meaning test for write permission.
@end deftypevr

@deftypevr Macro int X_OK
@standards{POSIX.1, unistd.h}
Flag meaning test for execute/search permission.
@end deftypevr

@deftypevr Macro int F_OK
@standards{POSIX.1, unistd.h}
Flag meaning test for existence of the file.
@end deftypevr

@deftypefun int faccessat (int @var{filedes}, const char *@var{filename}, int @var{how}, int @var{flags})
@standards{POSIX.1-2008, unistd.h}
@comment Unaudited and therefore marked AC-Unsafe and AS-Unsafe by default
@safety{@mtsafe{}@asunsafe{@asucorrupt{}}@acunsafe{@acucorrupt{}}}
This function is a descriptor-relative version of the @code{access}
function above.  @xref{Descriptor-Relative Access}.  The @var{flags}
argument can contain a combination of the flags @code{AT_EACCESS} described
below, @code{AT_EMPTY_PATH}, and @code{AT_SYMLINK_NOFOLLOW}.

@vtable @code
@item AT_EACCESS
This flag when passed to the @code{faccessat} function causes it to perform
access checks using effective user and group IDs instead of real IDs, which
is the default and matches the @code{access} function.
@end vtable

Compared to @code{access}, some additional error conditions can occur.
@xref{Descriptor-Relative Access}.

This function may not work correctly on older kernels missing the
@code{faccessat2} system call.
@end deftypefun

@node File Times
@subsection File Times

@cindex file access time
@cindex file modification time
@cindex file attribute modification time
Each file has three time stamps associated with it:  its access time,
its modification time, and its attribute modification time.  These
correspond to the @code{st_atime}, @code{st_mtime}, and @code{st_ctime}
members of the @code{stat} structure; see @ref{File Attributes}.

All of these times are represented in calendar time format, as
@code{time_t} objects.  This data type is defined in @file{time.h}.
For more information about representation and manipulation of time
values, see @ref{Calendar Time}.
@pindex time.h

Reading from a file updates its access time attribute, and writing
updates its modification time.  When a file is created, all three
time stamps for that file are set to the current time.  In addition, the
attribute change time and modification time fields of the directory that
contains the new entry are updated.

Adding a new name for a file with the @code{link} function updates the
attribute change time field of the file being linked, and both the
attribute change time and modification time fields of the directory
containing the new name.  These same fields are affected if a file name
is deleted with @code{unlink}, @code{remove} or @code{rmdir}.  Renaming
a file with @code{rename} affects only the attribute change time and
modification time fields of the two parent directories involved, and not
the times for the file being renamed.

Changing the attributes of a file (for example, with @code{chmod})
updates its attribute change time field.

You can also change some of the time stamps of a file explicitly using
the @code{utime} function---all except the attribute change time.  You
need to include the header file @file{utime.h} to use this facility.
@pindex utime.h

@deftp {Data Type} {struct utimbuf}
@standards{POSIX.1, utime.h}
The @code{utimbuf} structure is used with the @code{utime} function to
specify new access and modification times for a file.  It contains the
following members:

@table @code
@item time_t actime
This is the access time for the file.

@item time_t modtime
This is the modification time for the file.
@end table
@end deftp

@deftypefun int utime (const char *@var{filename}, const struct utimbuf *@var{times})
@standards{POSIX.1, utime.h}
@safety{@prelim{}@mtsafe{}@assafe{}@acsafe{}}
@c In the absence of a utime syscall, it non-atomically converts times
@c to a struct timeval and calls utimes.
This function is used to modify the file times associated with the file
named @var{filename}.

If @var{times} is a null pointer, then the access and modification times
of the file are set to the current time.  Otherwise, they are set to the
values from the @code{actime} and @code{modtime} members (respectively)
of the @code{utimbuf} structure pointed to by @var{times}.

The attribute modification time for the file is set to the current time
in either case (since changing the time stamps is itself a modification
of the file attributes).

The @code{utime} function returns @code{0} if successful and @code{-1}
on failure.  In addition to the usual file name errors
(@pxref{File Name Errors}), the following @code{errno} error conditions
are defined for this function:

@table @code
@item EACCES
There is a permission problem in the case where a null pointer was
passed as the @var{times} argument.  In order to update the time stamp on
the file, you must either be the owner of the file, have write
permission for the file, or be a privileged user.

@item ENOENT
The file doesn't exist.

@item EPERM
If the @var{times} argument is not a null pointer, you must either be
the owner of the file or be a privileged user.

@item EROFS
The file lives on a read-only file system.
@end table
@end deftypefun

Each of the three time stamps has a corresponding microsecond part,
which extends its resolution.  These fields are called
@code{st_atime_usec}, @code{st_mtime_usec}, and @code{st_ctime_usec};
each has a value between 0 and 999,999, which indicates the time in
microseconds.  They correspond to the @code{tv_usec} field of a
@code{timeval} structure; see @ref{Time Types}.

The @code{utimes} function is like @code{utime}, but also lets you specify
the fractional part of the file times.  The prototype for this function is
in the header file @file{sys/time.h}.
@pindex sys/time.h

@deftypefun int utimes (const char *@var{filename}, const struct timeval @var{tvp}@t{[2]})
@standards{BSD, sys/time.h}
@safety{@prelim{}@mtsafe{}@assafe{}@acsafe{}}
@c In the absence of a utimes syscall, it non-atomically converts tvp
@c to struct timespec array and issues a utimensat syscall, or to
@c struct utimbuf and calls utime.
This function sets the file access and modification times of the file
@var{filename}.  The new file access time is specified by
@code{@var{tvp}[0]}, and the new modification time by
@code{@var{tvp}[1]}.  Similar to @code{utime}, if @var{tvp} is a null
pointer then the access and modification times of the file are set to
the current time.  This function comes from BSD.

The return values and error conditions are the same as for the @code{utime}
function.
@end deftypefun

@deftypefun int lutimes (const char *@var{filename}, const struct timeval @var{tvp}@t{[2]})
@standards{BSD, sys/time.h}
@safety{@prelim{}@mtsafe{}@assafe{}@acsafe{}}
@c Since there's no lutimes syscall, it non-atomically converts tvp
@c to struct timespec array and issues a utimensat syscall.
This function is like @code{utimes}, except that it does not follow
symbolic links.  If @var{filename} is the name of a symbolic link,
@code{lutimes} sets the file access and modification times of the
symbolic link special file itself (as seen by @code{lstat};
@pxref{Symbolic Links}) while @code{utimes} sets the file access and
modification times of the file the symbolic link refers to.  This
function comes from FreeBSD, and is not available on all platforms (if
not available, it will fail with @code{ENOSYS}).

The return values and error conditions are the same as for the @code{utime}
function.
@end deftypefun

@deftypefun int futimes (int @var{fd}, const struct timeval @var{tvp}@t{[2]})
@standards{BSD, sys/time.h}
@safety{@prelim{}@mtsafe{}@assafe{}@acsafe{}}
@c Since there's no futimes syscall, it non-atomically converts tvp
@c to struct timespec array and issues a utimensat syscall, falling back
@c to utimes on a /proc/self/fd symlink.
This function is like @code{utimes}, except that it takes an open file
descriptor as an argument instead of a file name.  @xref{Low-Level
I/O}.  This function comes from FreeBSD, and is not available on all
platforms (if not available, it will fail with @code{ENOSYS}).

Like @code{utimes}, @code{futimes} returns @code{0} on success and @code{-1}
on failure.  The following @code{errno} error conditions are defined for
@code{futimes}:

@table @code
@item EACCES
There is a permission problem in the case where a null pointer was
passed as the @var{times} argument.  In order to update the time stamp on
the file, you must either be the owner of the file, have write
permission for the file, or be a privileged user.

@item EBADF
The @var{filedes} argument is not a valid file descriptor.

@item EINVAL
At least one of the fields in the @code{tvp} array passed has an invalid
value.

@item EPERM
If the @var{times} argument is not a null pointer, you must either be
the owner of the file or be a privileged user.

@item EROFS
The file lives on a read-only file system.
@end table
@end deftypefun

@deftypefun int futimens (int @var{filedes}, const struct timespec @var{tsp}@t{[2]})
@standards{POSIX.1-2008, sys/stat.h}
@comment Unaudited and therefore marked AC-Unsafe and AS-Unsafe by default
@safety{@mtsafe{}@asunsafe{@asucorrupt{}}@acunsafe{@acucorrupt{}}}
This function is like @code{futimes}, except that it sets the file access
and modification timestamps with nanosecond precision.  The argument
@code{tsp} is used similarly to @code{futimes}' @code{tvp}, but has a
@code{const struct timespec} type that can express calendar time with
nanosecond precision.  @xref{Time Types}.
@end deftypefun

@deftypefun int utimensat (int @var{filedes}, const char *@var{filename}, const struct timespec @var{tsp}@t{[2]}, int @var{flags})
@standards{POSIX.1-2008, sys/stat.h}
@comment Unaudited and therefore marked AC-Unsafe and AS-Unsafe by default
@safety{@mtsafe{}@asunsafe{@asucorrupt{}}@acunsafe{@acucorrupt{}}}
This function is a descriptor-relative version of the @code{futimens}
function above.  @xref{Descriptor-Relative Access}.  The @var{flags}
argument can contain a combination of the flags @code{AT_EMPTY_PATH}, and
@code{AT_SYMLINK_NOFOLLOW}.  The call:

@smallexample
futimens (@var{filedes}, @var{tsp})
@end smallexample

is equivalent to:

@smallexample
utimensat (@var{filedes}, @code{NULL}, @var{tsp}, 0)
@end smallexample

Compared to @code{futimens}, some additional error conditions can occur due
to descriptor-relative access.  @xref{Descriptor-Relative Access}.  In
addition to this, the following other errors can also occur:

@table @code
@item EINVAL
The @var{filename} argument is NULL, @var{filedes} is not @code{AT_FDCWD},
and @var{flags} is not @code{0}.

@item ELOOP
There are too many levels of indirection.  This can be the result of
circular symbolic links to directories.

@item ENAMETOOLONG
The resulting path is too long.  This error only occurs on systems which
have a limit on the file name length.

@item ENOENT
The @var{filename} argument is an empty string and @var{flags} does not
contain @code{AT_EMPTY_PATH}, or @var{filename} does not refer to an
existing file.

@item ESRCH
Search permission was denied for one of the prefix components of the the
@var{filename} argument.
@end table
@end deftypefun

@node File Size
@subsection File Size

Normally file sizes are maintained automatically.  A file begins with a
size of @math{0} and is automatically extended when data is written past
its end.  It is also possible to empty a file completely by an
@code{open} or @code{fopen} call.

However, sometimes it is necessary to @emph{reduce} the size of a file.
This can be done with the @code{truncate} and @code{ftruncate} functions.
They were introduced in BSD Unix.  @code{ftruncate} was later added to
POSIX.1.

Some systems allow you to extend a file (creating holes) with these
functions.  This is useful when using memory-mapped I/O
(@pxref{Memory-mapped I/O}), where files are not automatically extended.
However, it is not portable but must be implemented if @code{mmap}
allows mapping of files (i.e., @code{_POSIX_MAPPED_FILES} is defined).

Using these functions on anything other than a regular file gives
@emph{undefined} results.  On many systems, such a call will appear to
succeed, without actually accomplishing anything.

@deftypefun int truncate (const char *@var{filename}, off_t @var{length})
@standards{X/Open, unistd.h}
@safety{@prelim{}@mtsafe{}@assafe{}@acsafe{}}
@c In the absence of a truncate syscall, we use open and ftruncate.

The @code{truncate} function changes the size of @var{filename} to
@var{length}.  If @var{length} is shorter than the previous length, data
at the end will be lost.  The file must be writable by the user to
perform this operation.

If @var{length} is longer, holes will be added to the end.  However, some
systems do not support this feature and will leave the file unchanged.

When the source file is compiled with @code{_FILE_OFFSET_BITS == 64} the
@code{truncate} function is in fact @code{truncate64} and the type
@code{off_t} has 64 bits which makes it possible to handle files up to
@twoexp{63} bytes in length.

The return value is @math{0} for success, or @math{-1} for an error.  In
addition to the usual file name errors, the following errors may occur:

@table @code

@item EACCES
The file is a directory or not writable.

@item EINVAL
@var{length} is negative.

@item EFBIG
The operation would extend the file beyond the limits of the operating system.

@item EIO
A hardware I/O error occurred.

@item EPERM
The file is "append-only" or "immutable".

@item EINTR
The operation was interrupted by a signal.

@end table

@end deftypefun

@deftypefun int truncate64 (const char *@var{name}, off64_t @var{length})
@standards{Unix98, unistd.h}
@safety{@prelim{}@mtsafe{}@assafe{}@acsafe{}}
@c In the absence of a syscall, try truncate if length fits.
This function is similar to the @code{truncate} function.  The
difference is that the @var{length} argument is 64 bits wide even on 32
bits machines, which allows the handling of files with sizes up to
@twoexp{63} bytes.

When the source file is compiled with @code{_FILE_OFFSET_BITS == 64} on a
32 bits machine this function is actually available under the name
@code{truncate} and so transparently replaces the 32 bits interface.
@end deftypefun

@deftypefun int ftruncate (int @var{fd}, off_t @var{length})
@standards{POSIX, unistd.h}
@safety{@prelim{}@mtsafe{}@assafe{}@acsafe{}}

This is like @code{truncate}, but it works on a file descriptor @var{fd}
for an opened file instead of a file name to identify the object.  The
file must be opened for writing to successfully carry out the operation.

The POSIX standard leaves it implementation defined what happens if the
specified new @var{length} of the file is bigger than the original size.
The @code{ftruncate} function might simply leave the file alone and do
nothing or it can increase the size to the desired size.  In this later
case the extended area should be zero-filled.  So using @code{ftruncate}
is no reliable way to increase the file size but if it is possible it is
probably the fastest way.  The function also operates on POSIX shared
memory segments if these are implemented by the system.

@code{ftruncate} is especially useful in combination with @code{mmap}.
Since the mapped region must have a fixed size one cannot enlarge the
file by writing something beyond the last mapped page.  Instead one has
to enlarge the file itself and then remap the file with the new size.
The example below shows how this works.

When the source file is compiled with @code{_FILE_OFFSET_BITS == 64} the
@code{ftruncate} function is in fact @code{ftruncate64} and the type
@code{off_t} has 64 bits which makes it possible to handle files up to
@twoexp{63} bytes in length.

The return value is @math{0} for success, or @math{-1} for an error.  The
following errors may occur:

@table @code

@item EBADF
@var{fd} does not correspond to an open file.

@item EACCES
@var{fd} is a directory or not open for writing.

@item EINVAL
@var{length} is negative.

@item EFBIG
The operation would extend the file beyond the limits of the operating system.
@c or the open() call -- with the not-yet-discussed feature of opening
@c files with extra-large offsets.

@item EIO
A hardware I/O error occurred.

@item EPERM
The file is "append-only" or "immutable".

@item EINTR
The operation was interrupted by a signal.

@c ENOENT is also possible on Linux --- however it only occurs if the file
@c descriptor has a `file' structure but no `inode' structure.  I'm not
@c sure how such an fd could be created.  Perhaps it's a bug.

@end table

@end deftypefun

@deftypefun int ftruncate64 (int @var{id}, off64_t @var{length})
@standards{Unix98, unistd.h}
@safety{@prelim{}@mtsafe{}@assafe{}@acsafe{}}
@c In the absence of a syscall, try ftruncate if length fits.
This function is similar to the @code{ftruncate} function.  The
difference is that the @var{length} argument is 64 bits wide even on 32
bits machines which allows the handling of files with sizes up to
@twoexp{63} bytes.

When the source file is compiled with @code{_FILE_OFFSET_BITS == 64} on a
32 bits machine this function is actually available under the name
@code{ftruncate} and so transparently replaces the 32 bits interface.
@end deftypefun

As announced here is a little example of how to use @code{ftruncate} in
combination with @code{mmap}:

@smallexample
int fd;
void *start;
size_t len;

int
add (off_t at, void *block, size_t size)
@{
  if (at + size > len)
    @{
      /* Resize the file and remap.  */
      size_t ps = sysconf (_SC_PAGESIZE);
      size_t ns = (at + size + ps - 1) & ~(ps - 1);
      void *np;
      if (ftruncate (fd, ns) < 0)
        return -1;
      np = mmap (NULL, ns, PROT_READ|PROT_WRITE, MAP_SHARED, fd, 0);
      if (np == MAP_FAILED)
        return -1;
      start = np;
      len = ns;
    @}
  memcpy ((char *) start + at, block, size);
  return 0;
@}
@end smallexample

The function @code{add} writes a block of memory at an arbitrary
position in the file.  If the current size of the file is too small it
is extended.  Note that it is extended by a whole number of pages.  This
is a requirement of @code{mmap}.  The program has to keep track of the
real size, and when it has finished a final @code{ftruncate} call should
set the real size of the file.

@node Storage Allocation
@subsection Storage Allocation
@cindex allocating file storage
@cindex file allocation
@cindex storage allocating

@cindex file fragmentation
@cindex fragmentation of files
@cindex sparse files
@cindex files, sparse
Most file systems support allocating large files in a non-contiguous
fashion: the file is split into @emph{fragments} which are allocated
sequentially, but the fragments themselves can be scattered across the
disk.  File systems generally try to avoid such fragmentation because it
decreases performance, but if a file gradually increases in size, there
might be no other option than to fragment it.  In addition, many file
systems support @emph{sparse files} with @emph{holes}: regions of null
bytes for which no backing storage has been allocated by the file
system.  When the holes are finally overwritten with data, fragmentation
can occur as well.

Explicit allocation of storage for yet-unwritten parts of the file can
help the system to avoid fragmentation.  Additionally, if storage
pre-allocation fails, it is possible to report the out-of-disk error
early, often without filling up the entire disk.  However, due to
deduplication, copy-on-write semantics, and file compression, such
pre-allocation may not reliably prevent the out-of-disk-space error from
occurring later.  Checking for write errors is still required, and
writes to memory-mapped regions created with @code{mmap} can still
result in @code{SIGBUS}.

@deftypefun int posix_fallocate (int @var{fd}, off_t @var{offset}, off_t @var{length})
@safety{@prelim{}@mtsafe{}@assafe{}@acsafe{}}
@c If the file system does not support allocation,
@c @code{posix_fallocate} has a race with file extension (if
@c @var{length} is zero) or with concurrent writes of non-NUL bytes (if
@c @var{length} is positive).

Allocate backing store for the region of @var{length} bytes starting at
byte @var{offset} in the file for the descriptor @var{fd}.  The file
length is increased to @samp{@var{length} + @var{offset}} if necessary.

@var{fd} must be a regular file opened for writing, or @code{EBADF} is
returned.  If there is insufficient disk space to fulfill the allocation
request, @code{ENOSPC} is returned.

@strong{Note:} If @code{fallocate} is not available (because the file
system does not support it), @code{posix_fallocate} is emulated, which
has the following drawbacks:

@itemize @bullet
@item
It is very inefficient because all file system blocks in the requested
range need to be examined (even if they have been allocated before) and
potentially rewritten.  In contrast, with proper @code{fallocate}
support (see below), the file system can examine the internal file
allocation data structures and eliminate holes directly, maybe even
using unwritten extents (which are pre-allocated but uninitialized on
disk).

@item
There is a race condition if another thread or process modifies the
underlying file in the to-be-allocated area.  Non-null bytes could be
overwritten with null bytes.

@item
If @var{fd} has been opened with the @code{O_WRONLY} flag, the function
will fail with an @code{errno} value of @code{EBADF}.

@item
If @var{fd} has been opened with the @code{O_APPEND} flag, the function
will fail with an @code{errno} value of @code{EBADF}.

@item
If @var{length} is zero, @code{ftruncate} is used to increase the file
size as requested, without allocating file system blocks.  There is a
race condition which means that @code{ftruncate} can accidentally
truncate the file if it has been extended concurrently.
@end itemize

On Linux, if an application does not benefit from emulation or if the
emulation is harmful due to its inherent race conditions, the
application can use the Linux-specific @code{fallocate} function, with a
zero flag argument.  For the @code{fallocate} function, @theglibc{} does
not perform allocation emulation if the file system does not support
allocation.  Instead, an @code{EOPNOTSUPP} is returned to the caller.

@end deftypefun

@deftypefun int posix_fallocate64 (int @var{fd}, off64_t @var{offset}, off64_t @var{length})
@safety{@prelim{}@mtsafe{}@assafe{}@acsafe{}}

This function is a variant of @code{posix_fallocate64} which accepts
64-bit file offsets on all platforms.

@end deftypefun

@node Making Special Files
@section Making Special Files
@cindex creating special files
@cindex special files

The @code{mknod} function is the primitive for making special files,
such as files that correspond to devices.  @Theglibc{} includes
this function for compatibility with BSD.

The prototype for @code{mknod} is declared in @file{sys/stat.h}.
@pindex sys/stat.h

@deftypefun int mknod (const char *@var{filename}, mode_t @var{mode}, dev_t @var{dev})
@standards{BSD, sys/stat.h}
@safety{@prelim{}@mtsafe{}@assafe{}@acsafe{}}
@c Instead of issuing the syscall directly, we go through xmknod.
@c Although the internal xmknod takes a dev_t*, that could lead to
@c @mtsrace races, it's passed a pointer to mknod's dev.
The @code{mknod} function makes a special file with name @var{filename}.
The @var{mode} specifies the mode of the file, and may include the various
special file bits, such as @code{S_IFCHR} (for a character special file)
or @code{S_IFBLK} (for a block special file).  @xref{Testing File Type}.

The @var{dev} argument specifies which device the special file refers to.
Its exact interpretation depends on the kind of special file being created.

The return value is @code{0} on success and @code{-1} on error.  In addition
to the usual file name errors (@pxref{File Name Errors}), the
following @code{errno} error conditions are defined for this function:

@table @code
@item EPERM
The calling process is not privileged.  Only the superuser can create
special files.

@item ENOSPC
The directory or file system that would contain the new file is full
and cannot be extended.

@item EROFS
The directory containing the new file can't be modified because it's on
a read-only file system.

@item EEXIST
There is already a file named @var{filename}.  If you want to replace
this file, you must remove the old file explicitly first.
@end table
@end deftypefun

@node Temporary Files
@section Temporary Files

If you need to use a temporary file in your program, you can use the
@code{tmpfile} function to open it.  Or you can use the @code{tmpnam}
(better: @code{tmpnam_r}) function to provide a name for a temporary
file and then you can open it in the usual way with @code{fopen}.

The @code{tempnam} function is like @code{tmpnam} but lets you choose
what directory temporary files will go in, and something about what
their file names will look like.  Important for multi-threaded programs
is that @code{tempnam} is reentrant, while @code{tmpnam} is not since it
returns a pointer to a static buffer.

These facilities are declared in the header file @file{stdio.h}.
@pindex stdio.h

@deftypefun {FILE *} tmpfile (void)
@standards{ISO, stdio.h}
@safety{@prelim{}@mtsafe{}@asunsafe{@ascuheap{} @asulock{}}@acunsafe{@acsmem{} @acsfd{} @aculock{}}}
@c The unsafety issues are those of fdopen, plus @acsfd because of the
@c open.
@c __path_search (internal buf, !dir, const pfx, !try_tmpdir) ok
@c  libc_secure_genenv only if try_tmpdir
@c  xstat64, strlen, strcmp, sprintf
@c __gen_tempname (internal tmpl, __GT_FILE) ok
@c  strlen, memcmp, getpid, open/mkdir/lxstat64 ok
@c  HP_TIMING_NOW if available ok
@c  gettimeofday (!tz) first time, or every time if no HP_TIMING_NOW ok
@c  static value is used and modified without synchronization ok
@c   but the use is as a source of non-cryptographic randomness
@c   with retries in case of collision, so it should be safe
@c unlink, fdopen
This function creates a temporary binary file for update mode, as if by
calling @code{fopen} with mode @code{"wb+"}.  The file is deleted
automatically when it is closed or when the program terminates.  (On
some other @w{ISO C} systems the file may fail to be deleted if the program
terminates abnormally).

This function is reentrant.

When the sources are compiled with @code{_FILE_OFFSET_BITS == 64} on a
32-bit system this function is in fact @code{tmpfile64}, i.e., the LFS
interface transparently replaces the old interface.
@end deftypefun

@deftypefun {FILE *} tmpfile64 (void)
@standards{Unix98, stdio.h}
@safety{@prelim{}@mtsafe{}@asunsafe{@ascuheap{} @asulock{}}@acunsafe{@acsmem{} @acsfd{} @aculock{}}}
This function is similar to @code{tmpfile}, but the stream it returns a
pointer to was opened using @code{tmpfile64}.  Therefore this stream can
be used for files larger than @twoexp{31} bytes on 32-bit machines.

Please note that the return type is still @code{FILE *}.  There is no
special @code{FILE} type for the LFS interface.

If the sources are compiled with @code{_FILE_OFFSET_BITS == 64} on a 32
bits machine this function is available under the name @code{tmpfile}
and so transparently replaces the old interface.
@end deftypefun

@deftypefun {char *} tmpnam (char *@var{result})
@standards{ISO, stdio.h}
@safety{@prelim{}@mtunsafe{@mtasurace{:tmpnam/!result}}@asunsafe{}@acsafe{}}
@c The passed-in buffer should not be modified concurrently with the
@c call.
@c __path_search (static or passed-in buf, !dir, !pfx, !try_tmpdir) ok
@c __gen_tempname (internal tmpl, __GT_NOCREATE) ok
This function constructs and returns a valid file name that does not
refer to any existing file.  If the @var{result} argument is a null
pointer, the return value is a pointer to an internal static string,
which might be modified by subsequent calls and therefore makes this
function non-reentrant.  Otherwise, the @var{result} argument should be
a pointer to an array of at least @code{L_tmpnam} characters, and the
result is written into that array.

It is possible for @code{tmpnam} to fail if you call it too many times
without removing previously-created files.  This is because the limited
length of the temporary file names gives room for only a finite number
of different names.  If @code{tmpnam} fails it returns a null pointer.

@strong{Warning:} Between the time the pathname is constructed and the
file is created another process might have created a file with the same
name using @code{tmpnam}, leading to a possible security hole.  The
implementation generates names which can hardly be predicted, but when
opening the file you should use the @code{O_EXCL} flag.  Using
@code{tmpfile} or @code{mkstemp} is a safe way to avoid this problem.
@end deftypefun

@deftypefun {char *} tmpnam_r (char *@var{result})
@standards{GNU, stdio.h}
@safety{@prelim{}@mtsafe{}@assafe{}@acsafe{}}
This function is nearly identical to the @code{tmpnam} function, except
that if @var{result} is a null pointer it returns a null pointer.

This guarantees reentrancy because the non-reentrant situation of
@code{tmpnam} cannot happen here.

@strong{Warning}: This function has the same security problems as
@code{tmpnam}.
@end deftypefun

@deftypevr Macro int L_tmpnam
@standards{ISO, stdio.h}
The value of this macro is an integer constant expression that
represents the minimum size of a string large enough to hold a file name
generated by the @code{tmpnam} function.
@end deftypevr

@deftypevr Macro int TMP_MAX
@standards{ISO, stdio.h}
The macro @code{TMP_MAX} is a lower bound for how many temporary names
you can create with @code{tmpnam}.  You can rely on being able to call
@code{tmpnam} at least this many times before it might fail saying you
have made too many temporary file names.

With @theglibc{}, you can create a very large number of temporary
file names.  If you actually created the files, you would probably run
out of disk space before you ran out of names.  Some other systems have
a fixed, small limit on the number of temporary files.  The limit is
never less than @code{25}.
@end deftypevr

@deftypefun {char *} tempnam (const char *@var{dir}, const char *@var{prefix})
@standards{SVID, stdio.h}
@safety{@prelim{}@mtsafe{@mtsenv{}}@asunsafe{@ascuheap{}}@acunsafe{@acsmem{}}}
@c There's no way (short of being setuid) to avoid getenv("TMPDIR"),
@c even with a non-NULL dir.
@c
@c __path_search (internal buf, dir, pfx, try_tmpdir) unsafe getenv
@c __gen_tempname (internal tmpl, __GT_NOCREATE) ok
@c strdup
This function generates a unique temporary file name.  If @var{prefix}
is not a null pointer, up to five characters of this string are used as
a prefix for the file name.  The return value is a string newly
allocated with @code{malloc}, so you should release its storage with
@code{free} when it is no longer needed.

Because the string is dynamically allocated this function is reentrant.

The directory prefix for the temporary file name is determined by
testing each of the following in sequence.  The directory must exist and
be writable.

@itemize @bullet
@item
The environment variable @code{TMPDIR}, if it is defined.  For security
reasons this only happens if the program is not SUID or SGID enabled.

@item
The @var{dir} argument, if it is not a null pointer.

@item
The value of the @code{P_tmpdir} macro.

@item
The directory @file{/tmp}.
@end itemize

This function is defined for SVID compatibility.

@strong{Warning:} Between the time the pathname is constructed and the
file is created another process might have created a file with the same
name using @code{tempnam}, leading to a possible security hole.  The
implementation generates names which can hardly be predicted, but when
opening the file you should use the @code{O_EXCL} flag.  Using
@code{tmpfile} or @code{mkstemp} is a safe way to avoid this problem.
@end deftypefun
@cindex TMPDIR environment variable

@c !!! are we putting SVID/GNU/POSIX.1/BSD in here or not??
@deftypevr {SVID Macro} {char *} P_tmpdir
@standards{SVID, stdio.h}
This macro is the name of the default directory for temporary files.
@end deftypevr

Older Unix systems did not have the functions just described.  Instead
they used @code{mktemp} and @code{mkstemp}.  Both of these functions
work by modifying a file name template string you pass.  The last six
characters of this string must be @samp{XXXXXX}.  These six @samp{X}s
are replaced with six characters which make the whole string a unique
file name.  Usually the template string is something like
@samp{/tmp/@var{prefix}XXXXXX}, and each program uses a unique @var{prefix}.

@strong{NB:} Because @code{mktemp} and @code{mkstemp} modify the
template string, you @emph{must not} pass string constants to them.
String constants are normally in read-only storage, so your program
would crash when @code{mktemp} or @code{mkstemp} tried to modify the
string.  These functions are declared in the header file @file{stdlib.h}.
@pindex stdlib.h

@deftypefun {char *} mktemp (char *@var{template})
@standards{Unix, stdlib.h}
@safety{@prelim{}@mtsafe{}@assafe{}@acsafe{}}
@c __gen_tempname (caller tmpl, __GT_NOCREATE) ok
The @code{mktemp} function generates a unique file name by modifying
@var{template} as described above.  If successful, it returns
@var{template} as modified.  If @code{mktemp} cannot find a unique file
name, it makes @var{template} an empty string and returns that.  If
@var{template} does not end with @samp{XXXXXX}, @code{mktemp} returns a
null pointer.

@strong{Warning:} Between the time the pathname is constructed and the
file is created another process might have created a file with the same
name using @code{mktemp}, leading to a possible security hole.  The
implementation generates names which can hardly be predicted, but when
opening the file you should use the @code{O_EXCL} flag.  Using
@code{mkstemp} is a safe way to avoid this problem.
@end deftypefun

@deftypefun int mkstemp (char *@var{template})
@standards{BSD, stdlib.h}
@safety{@prelim{}@mtsafe{}@assafe{}@acsafe{@acsfd{}}}
@c __gen_tempname (caller tmpl, __GT_FILE) ok
The @code{mkstemp} function generates a unique file name just as
@code{mktemp} does, but it also opens the file for you with @code{open}
(@pxref{Opening and Closing Files}).  If successful, it modifies
@var{template} in place and returns a file descriptor for that file open
for reading and writing.  If @code{mkstemp} cannot create a
uniquely-named file, it returns @code{-1}.  If @var{template} does not
end with @samp{XXXXXX}, @code{mkstemp} returns @code{-1} and does not
modify @var{template}.

The file is opened using mode @code{0600}.  If the file is meant to be
used by other users this mode must be changed explicitly.
@end deftypefun

Unlike @code{mktemp}, @code{mkstemp} is actually guaranteed to create a
unique file that cannot possibly clash with any other program trying to
create a temporary file.  This is because it works by calling
@code{open} with the @code{O_EXCL} flag, which says you want to create a
new file and get an error if the file already exists.

@deftypefun {char *} mkdtemp (char *@var{template})
@standards{BSD, stdlib.h}
@safety{@prelim{}@mtsafe{}@assafe{}@acsafe{}}
@c __gen_tempname (caller tmpl, __GT_DIR) ok
The @code{mkdtemp} function creates a directory with a unique name.  If
it succeeds, it overwrites @var{template} with the name of the
directory, and returns @var{template}.  As with @code{mktemp} and
@code{mkstemp}, @var{template} should be a string ending with
@samp{XXXXXX}.

If @code{mkdtemp} cannot create an uniquely named directory, it returns
@code{NULL} and sets @code{errno} appropriately.  If @var{template} does
not end with @samp{XXXXXX}, @code{mkdtemp} returns @code{NULL} and does
not modify @var{template}.  @code{errno} will be set to @code{EINVAL} in
this case.

The directory is created using mode @code{0700}.
@end deftypefun

The directory created by @code{mkdtemp} cannot clash with temporary
files or directories created by other users.  This is because directory
creation always works like @code{open} with @code{O_EXCL}.
@xref{Creating Directories}.

The @code{mkdtemp} function comes from OpenBSD.

@c FIXME these are undocumented:
@c fchmodat
@c fchownat
@c futimesat
@c fstatat (there's a commented-out safety assessment for this one)
@c statx
@c mkfifoat
@c name_to_handle_at
@c openat
@c open_by_handle_at
@c readlinkat
@c renameat2
@c scandirat
@c symlinkat
@c mknodat