1statx(2) System Calls Manual statx(2)
2
6 statx - get file status (extended)
7
9 Standard C library (libc, -lc)
10
12 #define _GNU_SOURCE /* See feature_test_macros(7) */
13 #include <fcntl.h> /* Definition of AT_* constants */
14 #include <sys/stat.h>
15 int statx(int dirfd, const char *restrict pathname, int flags,
17 unsigned int mask, struct statx *restrict statxbuf);
18
20 This function returns information about a file, storing it in the buf‐
21 fer pointed to by statxbuf. The returned buffer is a structure of the
22 following type:
23 struct statx {
25 __u32 stx_mask; /* Mask of bits indicating
26 filled fields */
27 __u32 stx_blksize; /* Block size for filesystem I/O */
28 __u64 stx_attributes; /* Extra file attribute indicators */
29 __u32 stx_nlink; /* Number of hard links */
30 __u32 stx_uid; /* User ID of owner */
31 __u32 stx_gid; /* Group ID of owner */
32 __u16 stx_mode; /* File type and mode */
33 __u64 stx_ino; /* Inode number */
34 __u64 stx_size; /* Total size in bytes */
35 __u64 stx_blocks; /* Number of 512B blocks allocated */
36 __u64 stx_attributes_mask;
37 /* Mask to show what's supported
38 in stx_attributes */
39 /* The following fields are file timestamps */
41 struct statx_timestamp stx_atime; /* Last access */
42 struct statx_timestamp stx_btime; /* Creation */
43 struct statx_timestamp stx_ctime; /* Last status change */
44 struct statx_timestamp stx_mtime; /* Last modification */
45 /* If this file represents a device, then the next two
47 fields contain the ID of the device */
48 __u32 stx_rdev_major; /* Major ID */
49 __u32 stx_rdev_minor; /* Minor ID */
50 /* The next two fields contain the ID of the device
52 containing the filesystem where the file resides */
53 __u32 stx_dev_major; /* Major ID */
54 __u32 stx_dev_minor; /* Minor ID */
55 __u64 stx_mnt_id; /* Mount ID */
57 /* Direct I/O alignment restrictions */
59 __u32 stx_dio_mem_align;
60 __u32 stx_dio_offset_align;
61 };
62 The file timestamps are structures of the following type:
64 struct statx_timestamp {
66 __s64 tv_sec; /* Seconds since the Epoch (UNIX time) */
67 __u32 tv_nsec; /* Nanoseconds since tv_sec */
68 };
69 (Note that reserved space and padding is omitted.)
71 Invoking statx():
73 To access a file's status, no permissions are required on the file it‐
74 self, but in the case of statx() with a pathname, execute (search) per‐
75 mission is required on all of the directories in pathname that lead to
76 the file.
77 statx() uses pathname, dirfd, and flags to identify the target file in
79 one of the following ways:
80 An absolute pathname
82 If pathname begins with a slash, then it is an absolute pathname
83 that identifies the target file. In this case, dirfd is ig‐
84 nored.
85 A relative pathname
87 If pathname is a string that begins with a character other than
88 a slash and dirfd is AT_FDCWD, then pathname is a relative path‐
89 name that is interpreted relative to the process's current work‐
90 ing directory.
91 A directory-relative pathname
93 If pathname is a string that begins with a character other than
94 a slash and dirfd is a file descriptor that refers to a direc‐
95 tory, then pathname is a relative pathname that is interpreted
96 relative to the directory referred to by dirfd. (See openat(2)
97 for an explanation of why this is useful.)
98 By file descriptor
100 If pathname is an empty string and the AT_EMPTY_PATH flag is
101 specified in flags (see below), then the target file is the one
102 referred to by the file descriptor dirfd.
103 flags can be used to influence a pathname-based lookup. A value for
105 flags is constructed by ORing together zero or more of the following
106 constants:
107 AT_EMPTY_PATH
109 If pathname is an empty string, operate on the file referred to
110 by dirfd (which may have been obtained using the open(2) O_PATH
111 flag). In this case, dirfd can refer to any type of file, not
112 just a directory.
113 If dirfd is AT_FDCWD, the call operates on the current working
115 directory.
116 AT_NO_AUTOMOUNT
118 Don't automount the terminal ("basename") component of pathname
119 if it is a directory that is an automount point. This allows
120 the caller to gather attributes of an automount point (rather
121 than the location it would mount). This flag has no effect if
122 the mount point has already been mounted over.
123 The AT_NO_AUTOMOUNT flag can be used in tools that scan directo‐
125 ries to prevent mass-automounting of a directory of automount
126 points.
127 All of stat(2), lstat(2), and fstatat(2) act as though AT_NO_AU‐
129 TOMOUNT was set.
130 AT_SYMLINK_NOFOLLOW
132 If pathname is a symbolic link, do not dereference it: instead
133 return information about the link itself, like lstat(2).
134 flags can also be used to control what sort of synchronization the ker‐
136 nel will do when querying a file on a remote filesystem. This is done
137 by ORing in one of the following values:
138 AT_STATX_SYNC_AS_STAT
140 Do whatever stat(2) does. This is the default and is very much
141 filesystem-specific.
142 AT_STATX_FORCE_SYNC
144 Force the attributes to be synchronized with the server. This
145 may require that a network filesystem perform a data writeback
146 to get the timestamps correct.
147 AT_STATX_DONT_SYNC
149 Don't synchronize anything, but rather just take whatever the
150 system has cached if possible. This may mean that the informa‐
151 tion returned is approximate, but, on a network filesystem, it
152 may not involve a round trip to the server - even if no lease is
153 held.
154 The mask argument to statx() is used to tell the kernel which fields
156 the caller is interested in. mask is an ORed combination of the fol‐
157 lowing constants:
158 STATX_TYPE Want stx_mode & S_IFMT
160 STATX_MODE Want stx_mode & ~S_IFMT
161 STATX_NLINK Want stx_nlink
162 STATX_UID Want stx_uid
163 STATX_GID Want stx_gid
164 STATX_ATIME Want stx_atime
165 STATX_MTIME Want stx_mtime
166 STATX_CTIME Want stx_ctime
167 STATX_INO Want stx_ino
168 STATX_SIZE Want stx_size
169 STATX_BLOCKS Want stx_blocks
170 STATX_BASIC_STATS [All of the above]
171 STATX_BTIME Want stx_btime
172 STATX_ALL The same as STATX_BASIC_STATS | STATX_BTIME.
173 It is deprecated and should not be used.
174 STATX_MNT_ID Want stx_mnt_id (since Linux 5.8)
175 STATX_DIOALIGN Want stx_dio_mem_align and stx_dio_offset_align
176 (since Linux 6.1; support varies by filesystem)
177 Note that, in general, the kernel does not reject values in mask other
179 than the above. (For an exception, see EINVAL in errors.) Instead, it
180 simply informs the caller which values are supported by this kernel and
181 filesystem via the statx.stx_mask field. Therefore, do not simply set
182 mask to UINT_MAX (all bits set), as one or more bits may, in the fu‐
183 ture, be used to specify an extension to the buffer.
184 The returned information
186 The status information for the target file is returned in the statx
187 structure pointed to by statxbuf. Included in this is stx_mask which
188 indicates what other information has been returned. stx_mask has the
189 same format as the mask argument and bits are set in it to indicate
190 which fields have been filled in.
191 It should be noted that the kernel may return fields that weren't re‐
193 quested and may fail to return fields that were requested, depending on
194 what the backing filesystem supports. (Fields that are given values
195 despite being unrequested can just be ignored.) In either case,
196 stx_mask will not be equal mask.
197 If a filesystem does not support a field or if it has an unrepre‐
199 sentable value (for instance, a file with an exotic type), then the
200 mask bit corresponding to that field will be cleared in stx_mask even
201 if the user asked for it and a dummy value will be filled in for com‐
202 patibility purposes if one is available (e.g., a dummy UID and GID may
203 be specified to mount under some circumstances).
204 A filesystem may also fill in fields that the caller didn't ask for if
206 it has values for them available and the information is available at no
207 extra cost. If this happens, the corresponding bits will be set in
208 stx_mask.
209 Note: for performance and simplicity reasons, different fields in the
211 statx structure may contain state information from different moments
212 during the execution of the system call. For example, if stx_mode or
213 stx_uid is changed by another process by calling chmod(2) or chown(2),
214 stat() might return the old stx_mode together with the new stx_uid, or
215 the old stx_uid together with the new stx_mode.
216 Apart from stx_mask (which is described above), the fields in the statx
218 structure are:
219 stx_blksize
221 The "preferred" block size for efficient filesystem I/O. (Writ‐
222 ing to a file in smaller chunks may cause an inefficient read-
223 modify-rewrite.)
224 stx_attributes
226 Further status information about the file (see below for more
227 information).
228 stx_nlink
230 The number of hard links on a file.
231 stx_uid
233 This field contains the user ID of the owner of the file.
234 stx_gid
236 This field contains the ID of the group owner of the file.
237 stx_mode
239 The file type and mode. See inode(7) for details.
240 stx_ino
242 The inode number of the file.
243 stx_size
245 The size of the file (if it is a regular file or a symbolic
246 link) in bytes. The size of a symbolic link is the length of
247 the pathname it contains, without a terminating null byte.
248 stx_blocks
250 The number of blocks allocated to the file on the medium, in
251 512-byte units. (This may be smaller than stx_size/512 when the
252 file has holes.)
253 stx_attributes_mask
255 A mask indicating which bits in stx_attributes are supported by
256 the VFS and the filesystem.
257 stx_atime
259 The file's last access timestamp.
260 stx_btime
262 The file's creation timestamp.
263 stx_ctime
265 The file's last status change timestamp.
266 stx_mtime
268 The file's last modification timestamp.
269 stx_dev_major and stx_dev_minor
271 The device on which this file (inode) resides.
272 stx_rdev_major and stx_rdev_minor
274 The device that this file (inode) represents if the file is of
275 block or character device type.
276 stx_mnt_id
278 The mount ID of the mount containing the file. This is the same
279 number reported by name_to_handle_at(2) and corresponds to the
280 number in the first field in one of the records in
281 /proc/self/mountinfo.
282 stx_dio_mem_align
284 The alignment (in bytes) required for user memory buffers for
285 direct I/O (O_DIRECT) on this file, or 0 if direct I/O is not
286 supported on this file.
287 STATX_DIOALIGN (stx_dio_mem_align and stx_dio_offset_align) is
289 supported on block devices since Linux 6.1. The support on reg‐
290 ular files varies by filesystem; it is supported by ext4, f2fs,
291 and xfs since Linux 6.1.
292 stx_dio_offset_align
294 The alignment (in bytes) required for file offsets and I/O seg‐
295 ment lengths for direct I/O (O_DIRECT) on this file, or 0 if di‐
296 rect I/O is not supported on this file. This will only be non‐
297 zero if stx_dio_mem_align is nonzero, and vice versa.
298 For further information on the above fields, see inode(7).
300 File attributes
302 The stx_attributes field contains a set of ORed flags that indicate ad‐
303 ditional attributes of the file. Note that any attribute that is not
304 indicated as supported by stx_attributes_mask has no usable value here.
305 The bits in stx_attributes_mask correspond bit-by-bit to stx_at‐
306 tributes.
307 The flags are as follows:
309 STATX_ATTR_COMPRESSED
311 The file is compressed by the filesystem and may take extra re‐
312 sources to access.
313 STATX_ATTR_IMMUTABLE
315 The file cannot be modified: it cannot be deleted or renamed, no
316 hard links can be created to this file and no data can be writ‐
317 ten to it. See chattr(1).
318 STATX_ATTR_APPEND
320 The file can only be opened in append mode for writing. Random
321 access writing is not permitted. See chattr(1).
322 STATX_ATTR_NODUMP
324 File is not a candidate for backup when a backup program such as
325 dump(8) is run. See chattr(1).
326 STATX_ATTR_ENCRYPTED
328 A key is required for the file to be encrypted by the filesys‐
329 tem.
330 STATX_ATTR_VERITY (since Linux 5.5)
332 The file has fs-verity enabled. It cannot be written to, and
333 all reads from it will be verified against a cryptographic hash
334 that covers the entire file (e.g., via a Merkle tree).
335 STATX_ATTR_DAX (since Linux 5.8)
337 The file is in the DAX (cpu direct access) state. DAX state at‐
338 tempts to minimize software cache effects for both I/O and mem‐
339 ory mappings of this file. It requires a file system which has
340 been configured to support DAX.
341 DAX generally assumes all accesses are via CPU load / store in‐
343 structions which can minimize overhead for small accesses, but
344 may adversely affect CPU utilization for large transfers.
345 File I/O is done directly to/from user-space buffers and memory
347 mapped I/O may be performed with direct memory mappings that by‐
348 pass the kernel page cache.
349 While the DAX property tends to result in data being transferred
351 synchronously, it does not give the same guarantees as the
352 O_SYNC flag (see open(2)), where data and the necessary metadata
353 are transferred together.
354 A DAX file may support being mapped with the MAP_SYNC flag,
356 which enables a program to use CPU cache flush instructions to
357 persist CPU store operations without an explicit fsync(2). See
358 mmap(2) for more information.
359
361 On success, zero is returned. On error, -1 is returned, and errno is
362 set to indicate the error.
363
365 EACCES Search permission is denied for one of the directories in the
366 path prefix of pathname. (See also path_resolution(7).)
367 EBADF pathname is relative but dirfd is neither AT_FDCWD nor a valid
369 file descriptor.
370 EFAULT pathname or statxbuf is NULL or points to a location outside the
372 process's accessible address space.
373 EINVAL Invalid flag specified in flags.
375 EINVAL Reserved flag specified in mask. (Currently, there is one such
377 flag, designated by the constant STATX__RESERVED, with the value
378 0x80000000U.)
379 ELOOP Too many symbolic links encountered while traversing the path‐
381 name.
382 ENAMETOOLONG
384 pathname is too long.
385 ENOENT A component of pathname does not exist, or pathname is an empty
387 string and AT_EMPTY_PATH was not specified in flags.
388 ENOMEM Out of memory (i.e., kernel memory).
390 ENOTDIR
392 A component of the path prefix of pathname is not a directory or
393 pathname is relative and dirfd is a file descriptor referring to
394 a file other than a directory.
395
397 Linux.
398
400 Linux 4.11, glibc 2.28.
401
403 ls(1), stat(1), access(2), chmod(2), chown(2), name_to_handle_at(2),
404 readlink(2), stat(2), utime(2), proc(5), capabilities(7), inode(7),
405 symlink(7)
406Linux man-pages 6.04 2023-03-30 statx(2)