1CHOWN(2) Linux Programmer's Manual CHOWN(2)
2
6 chown, fchown, lchown, fchownat - change ownership of a file
7
9 #include <unistd.h>
10 int chown(const char *pathname, uid_t owner, gid_t group);
12 int fchown(int fd, uid_t owner, gid_t group);
13 int lchown(const char *pathname, uid_t owner, gid_t group);
14 #include <fcntl.h> /* Definition of AT_* constants */
16 #include <unistd.h>
17 int fchownat(int dirfd, const char *pathname,
19 uid_t owner, gid_t group, int flags);
20 Feature Test Macro Requirements for glibc (see feature_test_macros(7)):
22 fchown(), lchown():
24 /* Since glibc 2.12: */ _POSIX_C_SOURCE >= 200809L
25 || _XOPEN_SOURCE >= 500
26 || /* Glibc <= 2.19: */ _BSD_SOURCE
27 fchownat():
29 Since glibc 2.10:
30 _POSIX_C_SOURCE >= 200809L
31 Before glibc 2.10:
32 _ATFILE_SOURCE
33
35 These system calls change the owner and group of a file. The chown(),
36 fchown(), and lchown() system calls differ only in how the file is
37 specified:
38 * chown() changes the ownership of the file specified by pathname,
40 which is dereferenced if it is a symbolic link.
41 * fchown() changes the ownership of the file referred to by the open
43 file descriptor fd.
44 * lchown() is like chown(), but does not dereference symbolic links.
46 Only a privileged process (Linux: one with the CAP_CHOWN capability)
48 may change the owner of a file. The owner of a file may change the
49 group of the file to any group of which that owner is a member. A
50 privileged process (Linux: with CAP_CHOWN) may change the group arbi‐
51 trarily.
52 If the owner or group is specified as -1, then that ID is not changed.
54 When the owner or group of an executable file is changed by an unprivi‐
56 leged user, the S_ISUID and S_ISGID mode bits are cleared. POSIX does
57 not specify whether this also should happen when root does the chown();
58 the Linux behavior depends on the kernel version, and since Linux
59 2.2.13, root is treated like other users. In case of a non-group-exe‐
60 cutable file (i.e., one for which the S_IXGRP bit is not set) the S_IS‐
61 GID bit indicates mandatory locking, and is not cleared by a chown().
62 When the owner or group of an executable file is changed (by any user),
64 all capability sets for the file are cleared.
65 fchownat()
67 The fchownat() system call operates in exactly the same way as chown(),
68 except for the differences described here.
69 If the pathname given in pathname is relative, then it is interpreted
71 relative to the directory referred to by the file descriptor dirfd
72 (rather than relative to the current working directory of the calling
73 process, as is done by chown() for a relative pathname).
74 If pathname is relative and dirfd is the special value AT_FDCWD, then
76 pathname is interpreted relative to the current working directory of
77 the calling process (like chown()).
78 If pathname is absolute, then dirfd is ignored.
80 The flags argument is a bit mask created by ORing together 0 or more of
82 the following values;
83 AT_EMPTY_PATH (since Linux 2.6.39)
85 If pathname is an empty string, operate on the file referred to
86 by dirfd (which may have been obtained using the open(2) O_PATH
87 flag). In this case, dirfd can refer to any type of file, not
88 just a directory. If dirfd is AT_FDCWD, the call operates on
89 the current working directory. This flag is Linux-specific; de‐
90 fine _GNU_SOURCE to obtain its definition.
91 AT_SYMLINK_NOFOLLOW
93 If pathname is a symbolic link, do not dereference it: instead
94 operate on the link itself, like lchown(). (By default, fchow‐
95 nat() dereferences symbolic links, like chown().)
96 See openat(2) for an explanation of the need for fchownat().
98
100 On success, zero is returned. On error, -1 is returned, and errno is
101 set to indicate the error.
102
104 Depending on the filesystem, errors other than those listed below can
105 be returned.
106 The more general errors for chown() are listed below.
108 EACCES Search permission is denied on a component of the path prefix.
110 (See also path_resolution(7).)
111 EBADF (fchown()) fd is not a valid open file descriptor.
113 EBADF (fchownat()) pathname is relative but dirfd is neither AT_FDCWD
115 nor a valid file descriptor.
116 EFAULT pathname points outside your accessible address space.
118 EINVAL (fchownat()) Invalid flag specified in flags.
120 EIO (fchown()) A low-level I/O error occurred while modifying the
122 inode.
123 ELOOP Too many symbolic links were encountered in resolving pathname.
125 ENAMETOOLONG
127 pathname is too long.
128 ENOENT The file does not exist.
130 ENOMEM Insufficient kernel memory was available.
132 ENOTDIR
134 A component of the path prefix is not a directory.
135 ENOTDIR
137 (fchownat()) pathname is relative and dirfd is a file descriptor
138 referring to a file other than a directory.
139 EPERM The calling process did not have the required permissions (see
141 above) to change owner and/or group.
142 EPERM The file is marked immutable or append-only. (See
144 ioctl_iflags(2).)
145 EROFS The named file resides on a read-only filesystem.
147
149 fchownat() was added to Linux in kernel 2.6.16; library support was
150 added to glibc in version 2.4.
151
153 chown(), fchown(), lchown(): 4.4BSD, SVr4, POSIX.1-2001, POSIX.1-2008.
154 The 4.4BSD version can be used only by the superuser (that is, ordinary
156 users cannot give away files).
157 fchownat(): POSIX.1-2008.
159
161 Ownership of new files
162 When a new file is created (by, for example, open(2) or mkdir(2)), its
163 owner is made the same as the filesystem user ID of the creating
164 process. The group of the file depends on a range of factors, includ‐
165 ing the type of filesystem, the options used to mount the filesystem,
166 and whether or not the set-group-ID mode bit is enabled on the parent
167 directory. If the filesystem supports the -o grpid (or, synonymously
168 -o bsdgroups) and -o nogrpid (or, synonymously -o sysvgroups) mount(8)
169 options, then the rules are as follows:
170 * If the filesystem is mounted with -o grpid, then the group of a new
172 file is made the same as that of the parent directory.
173 * If the filesystem is mounted with -o nogrpid and the set-group-ID bit
175 is disabled on the parent directory, then the group of a new file is
176 made the same as the process's filesystem GID.
177 * If the filesystem is mounted with -o nogrpid and the set-group-ID bit
179 is enabled on the parent directory, then the group of a new file is
180 made the same as that of the parent directory.
181 As at Linux 4.12, the -o grpid and -o nogrpid mount options are sup‐
183 ported by ext2, ext3, ext4, and XFS. Filesystems that don't support
184 these mount options follow the -o nogrpid rules.
185 Glibc notes
187 On older kernels where fchownat() is unavailable, the glibc wrapper
188 function falls back to the use of chown() and lchown(). When pathname
189 is a relative pathname, glibc constructs a pathname based on the sym‐
190 bolic link in /proc/self/fd that corresponds to the dirfd argument.
191 NFS
193 The chown() semantics are deliberately violated on NFS filesystems
194 which have UID mapping enabled. Additionally, the semantics of all
195 system calls which access the file contents are violated, because
196 chown() may cause immediate access revocation on already open files.
197 Client side caching may lead to a delay between the time where owner‐
198 ship have been changed to allow access for a user and the time where
199 the file can actually be accessed by the user on other clients.
200 Historical details
202 The original Linux chown(), fchown(), and lchown() system calls sup‐
203 ported only 16-bit user and group IDs. Subsequently, Linux 2.4 added
204 chown32(), fchown32(), and lchown32(), supporting 32-bit IDs. The
205 glibc chown(), fchown(), and lchown() wrapper functions transparently
206 deal with the variations across kernel versions.
207 In versions of Linux prior to 2.1.81 (and distinct from 2.1.46),
209 chown() did not follow symbolic links. Since Linux 2.1.81, chown()
210 does follow symbolic links, and there is a new system call lchown()
211 that does not follow symbolic links. Since Linux 2.1.86, this new call
212 (that has the same semantics as the old chown()) has got the same
213 syscall number, and chown() got the newly introduced number.
214
216 The following program changes the ownership of the file named in its
217 second command-line argument to the value specified in its first com‐
218 mand-line argument. The new owner can be specified either as a numeric
219 user ID, or as a username (which is converted to a user ID by using
220 getpwnam(3) to perform a lookup in the system password file).
221 Program source
223 #include <pwd.h>
224 #include <stdio.h>
225 #include <stdlib.h>
226 #include <unistd.h>
227 int
229 main(int argc, char *argv[])
230 {
231 uid_t uid;
232 struct passwd *pwd;
233 char *endptr;
234 if (argc != 3 || argv[1][0] == '\0') {
236 fprintf(stderr, "%s <owner> <file>\n", argv[0]);
237 exit(EXIT_FAILURE);
238 }
239 uid = strtol(argv[1], &endptr, 10); /* Allow a numeric string */
241 if (*endptr != '\0') { /* Was not pure numeric string */
243 pwd = getpwnam(argv[1]); /* Try getting UID for username */
244 if (pwd == NULL) {
245 perror("getpwnam");
246 exit(EXIT_FAILURE);
247 }
248 uid = pwd->pw_uid;
250 }
251 if (chown(argv[2], uid, -1) == -1) {
253 perror("chown");
254 exit(EXIT_FAILURE);
255 }
256 exit(EXIT_SUCCESS);
258 }
259
261 chgrp(1), chown(1), chmod(2), flock(2), path_resolution(7), symlink(7)
262
264 This page is part of release 5.13 of the Linux man-pages project. A
265 description of the project, information about reporting bugs, and the
266 latest version of this page, can be found at
267 https://www.kernel.org/doc/man-pages/.
268Linux 2021-08-27 CHOWN(2)