readlinkat(2)

1READLINK(2)                Linux Programmer's Manual               READLINK(2)
2
3
4

NAME

6       readlink, readlinkat - read value of a symbolic link
7

SYNOPSIS

9       #include <unistd.h>
10
11       ssize_t readlink(const char *pathname, char *buf, size_t bufsiz);
12
13       #include <fcntl.h>           /* Definition of AT_* constants */
14       #include <unistd.h>
15
16       ssize_t readlinkat(int dirfd, const char *pathname,
17                          char *buf, size_t bufsiz);
18
19   Feature Test Macro Requirements for glibc (see feature_test_macros(7)):
20
21       readlink():
22           _XOPEN_SOURCE >= 500 || _POSIX_C_SOURCE >= 200112L
23               || /* Glibc versions <= 2.19: */ _BSD_SOURCE
24
25       readlinkat():
26           Since glibc 2.10:
27               _POSIX_C_SOURCE >= 200809L
28           Before glibc 2.10:
29               _ATFILE_SOURCE
30

DESCRIPTION

32       readlink()  places  the  contents  of the symbolic link pathname in the
33       buffer buf, which has size bufsiz.  readlink() does not append  a  null
34       byte  to buf.  It will (silently) truncate the contents (to a length of
35       bufsiz characters), in case the buffer is too small to hold all of  the
36       contents.
37
38   readlinkat()
39       The  readlinkat() system call operates in exactly the same way as read‐
40       link(), except for the differences described here.
41
42       If the pathname given in pathname is relative, then it  is  interpreted
43       relative  to  the  directory  referred  to by the file descriptor dirfd
44       (rather than relative to the current working directory of  the  calling
45       process, as is done by readlink() for a relative pathname).
46
47       If  pathname  is relative and dirfd is the special value AT_FDCWD, then
48       pathname is interpreted relative to the current  working  directory  of
49       the calling process (like readlink()).
50
51       If pathname is absolute, then dirfd is ignored.
52
53       Since  Linux 2.6.39, pathname can be an empty string, in which case the
54       call operates on the symbolic link referred to by dirfd  (which  should
55       have been obtained using open(2) with the O_PATH and O_NOFOLLOW flags).
56
57       See openat(2) for an explanation of the need for readlinkat().
58

RETURN VALUE

60       On  success, these calls return the number of bytes placed in buf.  (If
61       the returned value equals bufsiz, then truncation may  have  occurred.)
62       On error, -1 is returned and errno is set to indicate the error.
63

ERRORS

65       EACCES Search  permission is denied for a component of the path prefix.
66              (See also path_resolution(7).)
67
68       EFAULT buf extends outside the process's allocated address space.
69
70       EINVAL bufsiz is not positive.
71
72       EINVAL The named file (i.e., the final filename component of  pathname)
73              is not a symbolic link.
74
75       EIO    An I/O error occurred while reading from the filesystem.
76
77       ELOOP  Too  many  symbolic  links  were  encountered in translating the
78              pathname.
79
80       ENAMETOOLONG
81              A pathname, or a component of a pathname, was too long.
82
83       ENOENT The named file does not exist.
84
85       ENOMEM Insufficient kernel memory was available.
86
87       ENOTDIR
88              A component of the path prefix is not a directory.
89
90       The following additional errors can occur for readlinkat():
91
92       EBADF  dirfd is not a valid file descriptor.
93
94       ENOTDIR
95              pathname is relative and dirfd is a file descriptor referring to
96              a file other than a directory.
97

VERSIONS

99       readlinkat()  was  added to Linux in kernel 2.6.16; library support was
100       added to glibc in version 2.4.
101

CONFORMING TO

103       readlink(): 4.4BSD (readlink() first appeared in 4.2BSD), POSIX.1-2001,
104       POSIX.1-2008.
105
106       readlinkat(): POSIX.1-2008.
107

NOTES

109       In  versions of glibc up to and including glibc 2.4, the return type of
110       readlink() was declared as int.  Nowadays, the return type is  declared
111       as ssize_t, as (newly) required in POSIX.1-2001.
112
113       Using  a  statically sized buffer might not provide enough room for the
114       symbolic link contents.  The  required  size  for  the  buffer  can  be
115       obtained  from the stat.st_size value returned by a call to lstat(2) on
116       the link.  However, the number of bytes written by readlink() and read‐
117       linkat()  should  be checked to make sure that the size of the symbolic
118       link did not increase between the calls.   Dynamically  allocating  the
119       buffer  for  readlink() and readlinkat() also addresses a common porta‐
120       bility problem when using PATH_MAX for the buffer size,  as  this  con‐
121       stant  is not guaranteed to be defined per POSIX if the system does not
122       have such limit.
123
124   Glibc notes
125       On older kernels where readlinkat() is unavailable, the  glibc  wrapper
126       function falls back to the use of readlink().  When pathname is a rela‐
127       tive pathname, glibc constructs a pathname based on the  symbolic  link
128       in /proc/self/fd that corresponds to the dirfd argument.
129

EXAMPLE

131       The following program allocates the buffer needed by readlink() dynami‐
132       cally from the information provided by lstat(2), falling back to a buf‐
133       fer of size PATH_MAX in cases where lstat(2) reports a size of zero.
134
135       #include <sys/types.h>
136       #include <sys/stat.h>
137       #include <limits.h>
138       #include <stdio.h>
139       #include <stdlib.h>
140       #include <unistd.h>
141
142       int
143       main(int argc, char *argv[])
144       {
145           struct stat sb;
146           char *buf;
147           ssize_t nbytes, bufsiz;
148
149           if (argc != 2) {
150               fprintf(stderr, "Usage: %s <pathname>\n", argv[0]);
151               exit(EXIT_FAILURE);
152           }
153
154           if (lstat(argv[1], &sb) == -1) {
155               perror("lstat");
156               exit(EXIT_FAILURE);
157           }
158
159           /* Add one to the link size, so that we can determine whether
160              the buffer returned by readlink() was truncated. */
161
162           bufsiz = sb.st_size + 1;
163
164           /* Some magic symlinks under (for example) /proc and /sys
165              report 'st_size' as zero. In that case, take PATH_MAX as
166              a "good enough" estimate. */
167
168           if (sb.st_size == 0)
169               bufsiz = PATH_MAX;
170
171           buf = malloc(bufsiz);
172           if (buf == NULL) {
173               perror("malloc");
174               exit(EXIT_FAILURE);
175           }
176
177           nbytes = readlink(argv[1], buf, bufsiz);
178           if (nbytes == -1) {
179               perror("readlink");
180               exit(EXIT_FAILURE);
181           }
182
183           printf("'%s' points to '%.*s'\n", argv[1], (int) nbytes, buf);
184
185           /* If the return value was equal to the buffer size, then the
186              the link target was larger than expected (perhaps because the
187              target was changed between the call to lstat() and the call to
188              readlink()). Warn the user that the returned target may have
189              been truncated. */
190
191           if (nbytes == bufsiz)
192               printf("(Returned buffer may have been truncated)\n");
193
194           free(buf);
195           exit(EXIT_SUCCESS);
196       }
197

COLOPHON

203       This page is part of release 5.04 of the Linux  man-pages  project.   A
204       description  of  the project, information about reporting bugs, and the
205       latest    version    of    this    page,    can     be     found     at
206       https://www.kernel.org/doc/man-pages/.
207
208
209
210Linux                             2019-03-06                       READLINK(2)