1ACCESS(2) Linux Programmer's Manual ACCESS(2)
2
6 access - check real user's permissions for a file
7
9 #include <unistd.h>
10 int access(const char *pathname, int mode);
12
14 access() checks whether the calling process can access the file path‐
15 name. If pathname is a symbolic link, it is dereferenced.
16 The mode specifies the accessibility check(s) to be performed, and is
18 either the value F_OK, or a mask consisting of the bitwise OR of one or
19 more of R_OK, W_OK, and X_OK. F_OK tests for the existence of the
20 file. R_OK, W_OK, and X_OK test whether the file exists and grants
21 read, write, and execute permissions, respectively.
22 The check is done using the calling process's real UID and GID, rather
24 than the effective IDs as is done when actually attempting an operation
25 (e.g., open(2)) on the file. This allows set-user-ID programs to eas‐
26 ily determine the invoking user's authority.
27 If the calling process is privileged (i.e., its real UID is zero), then
29 an X_OK check is successful for a regular file if execute permission is
30 enabled for any of the file owner, group, or other.
31
33 On success (all requested permissions granted, or mode is F_OK and the
34 file exists), zero is returned. On error (at least one bit in mode
35 asked for a permission that is denied, or mode is F_OK and the file
36 does not exist, or some other error occurred), -1 is returned, and
37 errno is set appropriately.
38
40 access() shall fail if:
41 EACCES The requested access would be denied to the file, or search per‐
43 mission is denied for one of the directories in the path prefix
44 of pathname. (See also path_resolution(7).)
45 ELOOP Too many symbolic links were encountered in resolving pathname.
47 ENAMETOOLONG
49 pathname is too long.
50 ENOENT A component of pathname does not exist or is a dangling symbolic
52 link.
53 ENOTDIR
55 A component used as a directory in pathname is not, in fact, a
56 directory.
57 EROFS Write permission was requested for a file on a read-only file
59 system.
60 access() may fail if:
62 EFAULT pathname points outside your accessible address space.
64 EINVAL mode was incorrectly specified.
66 EIO An I/O error occurred.
68 ENOMEM Insufficient kernel memory was available.
70 ETXTBSY
72 Write access was requested to an executable which is being exe‐
73 cuted.
74
76 SVr4, 4.3BSD, POSIX.1-2001.
77
79 Warning: Using access() to check if a user is authorized to, for exam‐
80 ple, open a file before actually doing so using open(2) creates a secu‐
81 rity hole, because the user might exploit the short time interval
82 between checking and opening the file to manipulate it. For this rea‐
83 son, the use of this system call should be avoided. (In the example
84 just described, a safer alternative would be to temporarily switch the
85 process's effective user ID to the real ID and then call open(2).)
86 access() always dereferences symbolic links. If you need to check the
88 permissions on a symbolic link, use faccessat(2) with the flag AT_SYM‐
89 LINK_NOFOLLOW.
90 access() returns an error if any of the access types in mode is denied,
92 even if some of the other access types in mode are permitted.
93 If the calling process has appropriate privileges (i.e., is superuser),
95 POSIX.1-2001 permits an implementation to indicate success for an X_OK
96 check even if none of the execute file permission bits are set. Linux
97 does not do this.
98 A file is accessible only if the permissions on each of the directories
100 in the path prefix of pathname grant search (i.e., execute) access. If
101 any directory is inaccessible, then the access() call will fail,
102 regardless of the permissions on the file itself.
103 Only access bits are checked, not the file type or contents. There‐
105 fore, if a directory is found to be writable, it probably means that
106 files can be created in the directory, and not that the directory can
107 be written as a file. Similarly, a DOS file may be found to be "exe‐
108 cutable," but the execve(2) call will still fail.
109 access() may not work correctly on NFS file systems with UID mapping
111 enabled, because UID mapping is done on the server and hidden from the
112 client, which checks permissions. Similar problems can occur to FUSE
113 mounts.
114
116 In kernel 2.4 (and earlier) there is some strangeness in the handling
117 of X_OK tests for superuser. If all categories of execute permission
118 are disabled for a nondirectory file, then the only access() test that
119 returns -1 is when mode is specified as just X_OK; if R_OK or W_OK is
120 also specified in mode, then access() returns 0 for such files. Early
121 2.6 kernels (up to and including 2.6.3) also behaved in the same way as
122 kernel 2.4.
123 In kernels before 2.6.20, access() ignored the effect of the MS_NOEXEC
125 flag if it was used to mount(2) the underlying file system. Since ker‐
126 nel 2.6.20, access() honors this flag.
127
129 chmod(2), chown(2), faccessat(2), open(2), setgid(2), setuid(2),
130 stat(2), euidaccess(3), credentials(7), path_resolution(7)
131
133 This page is part of release 3.53 of the Linux man-pages project. A
134 description of the project, and information about reporting bugs, can
135 be found at http://www.kernel.org/doc/man-pages/.
136Linux 2013-04-16 ACCESS(2)