1UMOUNT(2) Linux Programmer's Manual UMOUNT(2)
2
6 umount, umount2 - unmount filesystem
7
9 #include <sys/mount.h>
10 int umount(const char *target);
12 int umount2(const char *target, int flags);
13
15 umount() and umount2() remove the attachment of the (topmost) filesys‐
16 tem mounted on target.
17 Appropriate privilege (Linux: the CAP_SYS_ADMIN capability) is required
19 to unmount filesystems.
20 Linux 2.1.116 added the umount2() system call, which, like umount(),
22 unmounts a target, but allows additional flags controlling the behavior
23 of the operation:
24 MNT_FORCE (since Linux 2.1.116)
26 Ask the filesystem to abort pending requests before attempting
27 the unmount. This may allow the unmount to complete without
28 waiting for an inaccessible server, but could cause data loss.
29 If, after aborting requests, some processes still have active
30 references to the filesystem, the unmount will still fail. As
31 at Linux 4.12, MNT_FORCE is supported only on the following
32 filesystems: 9p (since Linux 2.6.16), ceph (since Linux 2.6.34),
33 cifs (since Linux 2.6.12), fuse (since Linux 2.6.16), lustre
34 (since Linux 3.11), and NFS (since Linux 2.1.116).
35 MNT_DETACH (since Linux 2.4.11)
37 Perform a lazy unmount: make the mount point unavailable for new
38 accesses, immediately disconnect the filesystem and all filesys‐
39 tems mounted below it from each other and from the mount table,
40 and actually perform the unmount when the mount point ceases to
41 be busy.
42 MNT_EXPIRE (since Linux 2.6.8)
44 Mark the mount point as expired. If a mount point is not cur‐
45 rently in use, then an initial call to umount2() with this flag
46 fails with the error EAGAIN, but marks the mount point as ex‐
47 pired. The mount point remains expired as long as it isn't ac‐
48 cessed by any process. A second umount2() call specifying
49 MNT_EXPIRE unmounts an expired mount point. This flag cannot be
50 specified with either MNT_FORCE or MNT_DETACH.
51 UMOUNT_NOFOLLOW (since Linux 2.6.34)
53 Don't dereference target if it is a symbolic link. This flag
54 allows security problems to be avoided in set-user-ID-root pro‐
55 grams that allow unprivileged users to unmount filesystems.
56
58 On success, zero is returned. On error, -1 is returned, and errno is
59 set to indicate the error.
60
62 The error values given below result from filesystem type independent
63 errors. Each filesystem type may have its own special errors and its
64 own special behavior. See the Linux kernel source code for details.
65 EAGAIN A call to umount2() specifying MNT_EXPIRE successfully marked an
67 unbusy filesystem as expired.
68 EBUSY target could not be unmounted because it is busy.
70 EFAULT target points outside the user address space.
72 EINVAL target is not a mount point.
74 EINVAL umount2() was called with MNT_EXPIRE and either MNT_DETACH or
76 MNT_FORCE.
77 EINVAL (since Linux 2.6.34)
79 umount2() was called with an invalid flag value in flags.
80 ENAMETOOLONG
82 A pathname was longer than MAXPATHLEN.
83 ENOENT A pathname was empty or had a nonexistent component.
85 ENOMEM The kernel could not allocate a free page to copy filenames or
87 data into.
88 EPERM The caller does not have the required privileges.
90
92 MNT_DETACH and MNT_EXPIRE are available in glibc since version 2.11.
93
95 These functions are Linux-specific and should not be used in programs
96 intended to be portable.
97
99 umount() and shared mount points
100 Shared mount points cause any mount activity on a mount point, includ‐
101 ing umount() operations, to be forwarded to every shared mount point in
102 the peer group and every slave mount of that peer group. This means
103 that umount() of any peer in a set of shared mounts will cause all of
104 its peers to be unmounted and all of their slaves to be unmounted as
105 well.
106 This propagation of unmount activity can be particularly surprising on
108 systems where every mount point is shared by default. On such systems,
109 recursively bind mounting the root directory of the filesystem onto a
110 subdirectory and then later unmounting that subdirectory with MNT_DE‐
111 TACH will cause every mount in the mount namespace to be lazily un‐
112 mounted.
113 To ensure umount() does not propagate in this fashion, the mount point
115 may be remounted using a mount(2) call with a mount_flags argument that
116 includes both MS_REC and MS_PRIVATE prior to umount() being called.
117 Historical details
119 The original umount() function was called as umount(device) and would
120 return ENOTBLK when called with something other than a block device.
121 In Linux 0.98p4, a call umount(dir) was added, in order to support
122 anonymous devices. In Linux 2.3.99-pre7, the call umount(device) was
123 removed, leaving only umount(dir) (since now devices can be mounted in
124 more than one place, so specifying the device does not suffice).
125
127 mount(2), mount_namespaces(7), path_resolution(7), mount(8), umount(8)
128
130 This page is part of release 5.12 of the Linux man-pages project. A
131 description of the project, information about reporting bugs, and the
132 latest version of this page, can be found at
133 https://www.kernel.org/doc/man-pages/.
134Linux 2021-03-22 UMOUNT(2)