1MOUNT(2) Linux Programmer's Manual MOUNT(2)
2
6 mount - mount filesystem
7
9 #include <sys/mount.h>
10 int mount(const char *source, const char *target,
12 const char *filesystemtype, unsigned long mountflags,
13 const void *data);
14
16 mount() attaches the filesystem specified by source (which is often a
17 pathname referring to a device, but can also be the pathname of a di‐
18 rectory or file, or a dummy string) to the location (a directory or
19 file) specified by the pathname in target.
20 Appropriate privilege (Linux: the CAP_SYS_ADMIN capability) is required
22 to mount filesystems.
23 Values for the filesystemtype argument supported by the kernel are
25 listed in /proc/filesystems (e.g., "btrfs", "ext4", "jfs", "xfs",
26 "vfat", "fuse", "tmpfs", "cgroup", "proc", "mqueue", "nfs", "cifs",
27 "iso9660"). Further types may become available when the appropriate
28 modules are loaded.
29 The data argument is interpreted by the different filesystems. Typi‐
31 cally it is a string of comma-separated options understood by this
32 filesystem. See mount(8) for details of the options available for each
33 filesystem type. This argument may be specified as NULL, if there are
34 no options.
35 A call to mount() performs one of a number of general types of opera‐
37 tion, depending on the bits specified in mountflags. The choice of
38 which operation to perform is determined by testing the bits set in
39 mountflags, with the tests being conducted in the order listed here:
40 * Remount an existing mount: mountflags includes MS_REMOUNT.
42 * Create a bind mount: mountflags includes MS_BIND.
44 * Change the propagation type of an existing mount: mountflags in‐
46 cludes one of MS_SHARED, MS_PRIVATE, MS_SLAVE, or MS_UNBINDABLE.
47 * Move an existing mount to a new location: mountflags includes
49 MS_MOVE.
50 * Create a new mount: mountflags includes none of the above flags.
52 Each of these operations is detailed later in this page. Further flags
54 may be specified in mountflags to modify the behavior of mount(), as
55 described below.
56 Additional mount flags
58 The list below describes the additional flags that can be specified in
59 mountflags. Note that some operation types ignore some or all of these
60 flags, as described later in this page.
61 MS_DIRSYNC (since Linux 2.5.19)
63 Make directory changes on this filesystem synchronous. (This
64 property can be obtained for individual directories or subtrees
65 using chattr(1).)
66 MS_LAZYTIME (since Linux 4.0)
68 Reduce on-disk updates of inode timestamps (atime, mtime, ctime)
69 by maintaining these changes only in memory. The on-disk time‐
70 stamps are updated only when:
71 (a) the inode needs to be updated for some change unrelated to
73 file timestamps;
74 (b) the application employs fsync(2), syncfs(2), or sync(2);
76 (c) an undeleted inode is evicted from memory; or
78 (d) more than 24 hours have passed since the inode was written
80 to disk.
81 This mount option significantly reduces writes needed to update
83 the inode's timestamps, especially mtime and atime. However, in
84 the event of a system crash, the atime and mtime fields on disk
85 might be out of date by up to 24 hours.
86 Examples of workloads where this option could be of significant
88 benefit include frequent random writes to preallocated files, as
89 well as cases where the MS_STRICTATIME mount option is also en‐
90 abled. (The advantage of combining MS_STRICTATIME and MS_LAZY‐
91 TIME is that stat(2) will return the correctly updated atime,
92 but the atime updates will be flushed to disk only in the cases
93 listed above.)
94 MS_MANDLOCK
96 Permit mandatory locking on files in this filesystem. (Manda‐
97 tory locking must still be enabled on a per-file basis, as de‐
98 scribed in fcntl(2).) Since Linux 4.5, this mount option re‐
99 quires the CAP_SYS_ADMIN capability and a kernel configured with
100 the CONFIG_MANDATORY_FILE_LOCKING option.
101 MS_NOATIME
103 Do not update access times for (all types of) files on this
104 filesystem.
105 MS_NODEV
107 Do not allow access to devices (special files) on this filesys‐
108 tem.
109 MS_NODIRATIME
111 Do not update access times for directories on this filesystem.
112 This flag provides a subset of the functionality provided by
113 MS_NOATIME; that is, MS_NOATIME implies MS_NODIRATIME.
114 MS_NOEXEC
116 Do not allow programs to be executed from this filesystem.
117 MS_NOSUID
119 Do not honor set-user-ID and set-group-ID bits or file capabili‐
120 ties when executing programs from this filesystem. In addition,
121 SELinux domain transitions require the permission nosuid_transi‐
122 tion, which in turn needs also the policy capability nnp_no‐
123 suid_transition.
124 MS_RDONLY
126 Mount filesystem read-only.
127 MS_REC (since Linux 2.4.11)
129 Used in conjunction with MS_BIND to create a recursive bind
130 mount, and in conjunction with the propagation type flags to re‐
131 cursively change the propagation type of all of the mounts in a
132 subtree. See below for further details.
133 MS_RELATIME (since Linux 2.6.20)
135 When a file on this filesystem is accessed, update the file's
136 last access time (atime) only if the current value of atime is
137 less than or equal to the file's last modification time (mtime)
138 or last status change time (ctime). This option is useful for
139 programs, such as mutt(1), that need to know when a file has
140 been read since it was last modified. Since Linux 2.6.30, the
141 kernel defaults to the behavior provided by this flag (unless
142 MS_NOATIME was specified), and the MS_STRICTATIME flag is re‐
143 quired to obtain traditional semantics. In addition, since
144 Linux 2.6.30, the file's last access time is always updated if
145 it is more than 1 day old.
146 MS_SILENT (since Linux 2.6.17)
148 Suppress the display of certain (printk()) warning messages in
149 the kernel log. This flag supersedes the misnamed and obsolete
150 MS_VERBOSE flag (available since Linux 2.4.12), which has the
151 same meaning.
152 MS_STRICTATIME (since Linux 2.6.30)
154 Always update the last access time (atime) when files on this
155 filesystem are accessed. (This was the default behavior before
156 Linux 2.6.30.) Specifying this flag overrides the effect of
157 setting the MS_NOATIME and MS_RELATIME flags.
158 MS_SYNCHRONOUS
160 Make writes on this filesystem synchronous (as though the O_SYNC
161 flag to open(2) was specified for all file opens to this
162 filesystem).
163 MS_NOSYMFOLLOW (since Linux 5.10)
165 Do not follow symbolic links when resolving paths. Symbolic
166 links can still be created, and readlink(1), readlink(2), real‐
167 path(1), and realpath(3) all still work properly.
168 From Linux 2.4 onward, some of the above flags are settable on a per-
170 mount basis, while others apply to the superblock of the mounted
171 filesystem, meaning that all mounts of the same filesystem share those
172 flags. (Previously, all of the flags were per-superblock.)
173 The per-mount-point flags are as follows:
175 * Since Linux 2.4: MS_NODEV, MS_NOEXEC, and MS_NOSUID flags are set‐
177 table on a per-mount-point basis.
178 * Additionally, since Linux 2.6.16: MS_NOATIME and MS_NODIRATIME.
180 * Additionally, since Linux 2.6.20: MS_RELATIME.
182 The following flags are per-superblock: MS_DIRSYNC, MS_LAZYTIME,
184 MS_MANDLOCK, MS_SILENT, and MS_SYNCHRONOUS. The initial settings of
185 these flags are determined on the first mount of the filesystem, and
186 will be shared by all subsequent mounts of the same filesystem. Subse‐
187 quently, the settings of the flags can be changed via a remount opera‐
188 tion (see below). Such changes will be visible via all mount points
189 associated with the filesystem.
190 Since Linux 2.6.16, MS_RDONLY can be set or cleared on a per-mount-
192 point basis as well as on the underlying filesystem superblock. The
193 mounted filesystem will be writable only if neither the filesystem nor
194 the mountpoint are flagged as read-only.
195 Remounting an existing mount
197 An existing mount may be remounted by specifying MS_REMOUNT in mount‐
198 flags. This allows you to change the mountflags and data of an exist‐
199 ing mount without having to unmount and remount the filesystem. target
200 should be the same value specified in the initial mount() call.
201 The source and filesystemtype arguments are ignored.
203 The mountflags and data arguments should match the values used in the
205 original mount() call, except for those parameters that are being de‐
206 liberately changed.
207 The following mountflags can be changed: MS_LAZYTIME, MS_MANDLOCK,
209 MS_NOATIME, MS_NODEV, MS_NODIRATIME, MS_NOEXEC, MS_NOSUID, MS_RELATIME,
210 MS_RDONLY, MS_STRICTATIME (whose effect is to clear the MS_NOATIME and
211 MS_RELATIME flags), and MS_SYNCHRONOUS. Attempts to change the setting
212 of the MS_DIRSYNC and MS_SILENT flags during a remount are silently ig‐
213 nored. Note that changes to per-superblock flags are visible via all
214 mount points of the associated filesystem (because the per-superblock
215 flags are shared by all mount points).
216 Since Linux 3.17, if none of MS_NOATIME, MS_NODIRATIME, MS_RELATIME, or
218 MS_STRICTATIME is specified in mountflags, then the remount operation
219 preserves the existing values of these flags (rather than defaulting to
220 MS_RELATIME).
221 Since Linux 2.6.26, the MS_REMOUNT flag can be used with MS_BIND to
223 modify only the per-mount-point flags. This is particularly useful for
224 setting or clearing the "read-only" flag on a mount point without
225 changing the underlying filesystem. Specifying mountflags as:
226 MS_REMOUNT | MS_BIND | MS_RDONLY
228 will make access through this mountpoint read-only, without affecting
230 other mount points.
231 Creating a bind mount
233 If mountflags includes MS_BIND (available since Linux 2.4), then per‐
234 form a bind mount. A bind mount makes a file or a directory subtree
235 visible at another point within the single directory hierarchy. Bind
236 mounts may cross filesystem boundaries and span chroot(2) jails.
237 The filesystemtype and data arguments are ignored.
239 The remaining bits (other than MS_REC, described below) in the mount‐
241 flags argument are also ignored. (The bind mount has the same mount
242 options as the underlying mount point.) However, see the discussion of
243 remounting above, for a method of making an existing bind mount read-
244 only.
245 By default, when a directory is bind mounted, only that directory is
247 mounted; if there are any submounts under the directory tree, they are
248 not bind mounted. If the MS_REC flag is also specified, then a recur‐
249 sive bind mount operation is performed: all submounts under the source
250 subtree (other than unbindable mounts) are also bind mounted at the
251 corresponding location in the target subtree.
252 Changing the propagation type of an existing mount
254 If mountflags includes one of MS_SHARED, MS_PRIVATE, MS_SLAVE, or
255 MS_UNBINDABLE (all available since Linux 2.6.15), then the propagation
256 type of an existing mount is changed. If more than one of these flags
257 is specified, an error results.
258 The only other flags that can be specified while changing the propaga‐
260 tion type are MS_REC (described below) and MS_SILENT (which is ig‐
261 nored).
262 The source, filesystemtype, and data arguments are ignored.
264 The meanings of the propagation type flags are as follows:
266 MS_SHARED
268 Make this mount point shared. Mount and unmount events immedi‐
269 ately under this mount point will propagate to the other mount
270 points that are members of this mount's peer group. Propagation
271 here means that the same mount or unmount will automatically oc‐
272 cur under all of the other mount points in the peer group. Con‐
273 versely, mount and unmount events that take place under peer
274 mount points will propagate to this mount point.
275 MS_PRIVATE
277 Make this mount point private. Mount and unmount events do not
278 propagate into or out of this mount point.
279 MS_SLAVE
281 If this is a shared mount point that is a member of a peer group
282 that contains other members, convert it to a slave mount. If
283 this is a shared mount point that is a member of a peer group
284 that contains no other members, convert it to a private mount.
285 Otherwise, the propagation type of the mount point is left un‐
286 changed.
287 When a mount point is a slave, mount and unmount events propa‐
289 gate into this mount point from the (master) shared peer group
290 of which it was formerly a member. Mount and unmount events un‐
291 der this mount point do not propagate to any peer.
292 A mount point can be the slave of another peer group while at
294 the same time sharing mount and unmount events with a peer group
295 of which it is a member.
296 MS_UNBINDABLE
298 Make this mount unbindable. This is like a private mount, and
299 in addition this mount can't be bind mounted. When a recursive
300 bind mount (mount() with the MS_BIND and MS_REC flags) is per‐
301 formed on a directory subtree, any unbindable mounts within the
302 subtree are automatically pruned (i.e., not replicated) when
303 replicating that subtree to produce the target subtree.
304 By default, changing the propagation type affects only the target mount
306 point. If the MS_REC flag is also specified in mountflags, then the
307 propagation type of all mount points under target is also changed.
308 For further details regarding mount propagation types (including the
310 default propagation type assigned to new mounts), see mount_name‐
311 spaces(7).
312 Moving a mount
314 If mountflags contains the flag MS_MOVE (available since Linux 2.4.18),
315 then move a subtree: source specifies an existing mount point and tar‐
316 get specifies the new location to which that mount point is to be relo‐
317 cated. The move is atomic: at no point is the subtree unmounted.
318 The remaining bits in the mountflags argument are ignored, as are the
320 filesystemtype and data arguments.
321 Creating a new mount point
323 If none of MS_REMOUNT, MS_BIND, MS_MOVE, MS_SHARED, MS_PRIVATE,
324 MS_SLAVE, or MS_UNBINDABLE is specified in mountflags, then mount()
325 performs its default action: creating a new mount point. source speci‐
326 fies the source for the new mount point, and target specifies the di‐
327 rectory at which to create the mount point.
328 The filesystemtype and data arguments are employed, and further bits
330 may be specified in mountflags to modify the behavior of the call.
331
333 On success, zero is returned. On error, -1 is returned, and errno is
334 set to indicate the error.
335
337 The error values given below result from filesystem type independent
338 errors. Each filesystem type may have its own special errors and its
339 own special behavior. See the Linux kernel source code for details.
340 EACCES A component of a path was not searchable. (See also path_reso‐
342 lution(7).)
343 EACCES Mounting a read-only filesystem was attempted without giving the
345 MS_RDONLY flag.
346 The filesystem may be read-only for various reasons, including:
348 it resides on a read-only optical disk; it is resides on a de‐
349 vice with a physical switch that has been set to mark the device
350 read-only; the filesystem implementation was compiled with read-
351 only support; or errors were detected when initially mounting
352 the filesystem, so that it was marked read-only and can't be re‐
353 mounted as read-write (until the errors are fixed).
354 Some filesystems instead return the error EROFS on an attempt to
356 mount a read-only filesystem.
357 EACCES The block device source is located on a filesystem mounted with
359 the MS_NODEV option.
360 EBUSY An attempt was made to stack a new mount directly on top of an
362 existing mount point that was created in this mount namespace
363 with the same source and target.
364 EBUSY source cannot be remounted read-only, because it still holds
366 files open for writing.
367 EFAULT One of the pointer arguments points outside the user address
369 space.
370 EINVAL source had an invalid superblock.
372 EINVAL A remount operation (MS_REMOUNT) was attempted, but source was
374 not already mounted on target.
375 EINVAL A move operation (MS_MOVE) was attempted, but the mount tree un‐
377 der source includes unbindable mounts and target is a mount
378 point that has propagation type MS_SHARED.
379 EINVAL A move operation (MS_MOVE) was attempted, but the parent mount
381 of source mount has propagation type MS_SHARED.
382 EINVAL A move operation (MS_MOVE) was attempted, but source was not a
384 mount point, or was '/'.
385 EINVAL A bind operation (MS_BIND) was requested where source referred a
387 mount namespace magic link (i.e., a /proc/[pid]/ns/mnt magic
388 link or a bind mount to such a link) and the propagation type of
389 the parent mount of target was MS_SHARED, but propagation of the
390 requested bind mount could lead to a circular dependency that
391 might prevent the mount namespace from ever being freed.
392 EINVAL mountflags includes more than one of MS_SHARED, MS_PRIVATE,
394 MS_SLAVE, or MS_UNBINDABLE.
395 EINVAL mountflags includes MS_SHARED, MS_PRIVATE, MS_SLAVE, or MS_UN‐
397 BINDABLE and also includes a flag other than MS_REC or
398 MS_SILENT.
399 EINVAL An attempt was made to bind mount an unbindable mount.
401 EINVAL In an unprivileged mount namespace (i.e., a mount namespace
403 owned by a user namespace that was created by an unprivileged
404 user), a bind mount operation (MS_BIND) was attempted without
405 specifying (MS_REC), which would have revealed the filesystem
406 tree underneath one of the submounts of the directory being
407 bound.
408 ELOOP Too many links encountered during pathname resolution.
410 ELOOP A move operation was attempted, and target is a descendant of
412 source.
413 EMFILE (In case no block device is required:) Table of dummy devices is
415 full.
416 ENAMETOOLONG
418 A pathname was longer than MAXPATHLEN.
419 ENODEV filesystemtype not configured in the kernel.
421 ENOENT A pathname was empty or had a nonexistent component.
423 ENOMEM The kernel could not allocate a free page to copy filenames or
425 data into.
426 ENOTBLK
428 source is not a block device (and a device was required).
429 ENOTDIR
431 target, or a prefix of source, is not a directory.
432 ENXIO The major number of the block device source is out of range.
434 EPERM The caller does not have the required privileges.
436 EROFS Mounting a read-only filesystem was attempted without giving the
438 MS_RDONLY flag. See EACCES, above.
439
441 The definitions of MS_DIRSYNC, MS_MOVE, MS_PRIVATE, MS_REC, MS_RELA‐
442 TIME, MS_SHARED, MS_SLAVE, MS_STRICTATIME, and MS_UNBINDABLE were added
443 to glibc headers in version 2.12.
444
446 This function is Linux-specific and should not be used in programs in‐
447 tended to be portable.
448
450 Since Linux 2.4 a single filesystem can be mounted at multiple mount
451 points, and multiple mounts can be stacked on the same mount point.
452 The mountflags argument may have the magic number 0xC0ED (MS_MGC_VAL)
454 in the top 16 bits. (All of the other flags discussed in DESCRIPTION
455 occupy the low order 16 bits of mountflags.) Specifying MS_MGC_VAL was
456 required in kernel versions prior to 2.4, but since Linux 2.4 is no
457 longer required and is ignored if specified.
458 The original MS_SYNC flag was renamed MS_SYNCHRONOUS in 1.1.69 when a
460 different MS_SYNC was added to <mman.h>.
461 Before Linux 2.4 an attempt to execute a set-user-ID or set-group-ID
463 program on a filesystem mounted with MS_NOSUID would fail with EPERM.
464 Since Linux 2.4 the set-user-ID and set-group-ID bits are just silently
465 ignored in this case.
466 Mount namespaces
468 Starting with kernel 2.4.19, Linux provides mount namespaces. A mount
469 namespace is the set of filesystem mounts that are visible to a
470 process. Mount namespaces can be (and usually are) shared between mul‐
471 tiple processes, and changes to the namespace (i.e., mounts and un‐
472 mounts) by one process are visible to all other processes sharing the
473 same namespace. (The pre-2.4.19 Linux situation can be considered as
474 one in which a single namespace was shared by every process on the sys‐
475 tem.)
476 A child process created by fork(2) shares its parent's mount namespace;
478 the mount namespace is preserved across an execve(2).
479 A process can obtain a private mount namespace if: it was created using
481 the clone(2) CLONE_NEWNS flag, in which case its new namespace is ini‐
482 tialized to be a copy of the namespace of the process that called
483 clone(2); or it calls unshare(2) with the CLONE_NEWNS flag, which
484 causes the caller's mount namespace to obtain a private copy of the
485 namespace that it was previously sharing with other processes, so that
486 future mounts and unmounts by the caller are invisible to other pro‐
487 cesses (except child processes that the caller subsequently creates)
488 and vice versa.
489 For further details on mount namespaces, see mount_namespaces(7).
491 Parental relationship between mount points
493 Each mount point has a parent mount point. The overall parental rela‐
494 tionship of all mount points defines the single directory hierarchy
495 seen by the processes within a mount namespace.
496 The parent of a new mount point is defined when the mount point is cre‐
498 ated. In the usual case, the parent of a new mount is the mount point
499 of the filesystem containing the directory or file at which the new
500 mount is attached. In the case where a new mount is stacked on top of
501 an existing mount, the parent of the new mount is the previous mount
502 that was stacked at that location.
503 The parental relationship between mount points can be discovered via
505 the /proc/[pid]/mountinfo file (see below).
506 /proc/[pid]/mounts and /proc/[pid]/mountinfo
508 The Linux-specific /proc/[pid]/mounts file exposes the list of mount
509 points in the mount namespace of the process with the specified ID.
510 The /proc/[pid]/mountinfo file exposes even more information about
511 mount points, including the propagation type and mount ID information
512 that makes it possible to discover the parental relationship between
513 mount points. See proc(5) and mount_namespaces(7) for details of this
514 file.
515
517 mountpoint(1), chroot(2), ioctl_iflags(2), pivot_root(2), umount(2),
518 mount_namespaces(7), path_resolution(7), findmnt(8), lsblk(8),
519 mount(8), umount(8)
520
522 This page is part of release 5.12 of the Linux man-pages project. A
523 description of the project, information about reporting bugs, and the
524 latest version of this page, can be found at
525 https://www.kernel.org/doc/man-pages/.
526Linux 2021-03-22 MOUNT(2)