1STATFS(2) Linux Programmer's Manual STATFS(2)
2
6 statfs, fstatfs - get filesystem statistics
7
9 #include <sys/vfs.h> /* or <sys/statfs.h> */
10 int statfs(const char *path, struct statfs *buf);
12 int fstatfs(int fd, struct statfs *buf);
13
15 The statfs() system call returns information about a mounted filesys‐
16 tem. path is the pathname of any file within the mounted filesystem.
17 buf is a pointer to a statfs structure defined approximately as fol‐
18 lows:
19 struct statfs {
21 __fsword_t f_type; /* Type of filesystem (see below) */
22 __fsword_t f_bsize; /* Optimal transfer block size */
23 fsblkcnt_t f_blocks; /* Total data blocks in filesystem */
24 fsblkcnt_t f_bfree; /* Free blocks in filesystem */
25 fsblkcnt_t f_bavail; /* Free blocks available to
26 unprivileged user */
27 fsfilcnt_t f_files; /* Total inodes in filesystem */
28 fsfilcnt_t f_ffree; /* Free inodes in filesystem */
29 fsid_t f_fsid; /* Filesystem ID */
30 __fsword_t f_namelen; /* Maximum length of filenames */
31 __fsword_t f_frsize; /* Fragment size (since Linux 2.6) */
32 __fsword_t f_flags; /* Mount flags of filesystem
33 (since Linux 2.6.36) */
34 __fsword_t f_spare[xxx];
35 /* Padding bytes reserved for future use */
36 };
37 The following filesystem types may appear in f_type:
39 ADFS_SUPER_MAGIC 0xadf5
41 AFFS_SUPER_MAGIC 0xadff
42 AFS_SUPER_MAGIC 0x5346414f
43 ANON_INODE_FS_MAGIC 0x09041934 /* Anonymous inode FS (for
44 pseudofiles that have no name;
45 e.g., epoll, signalfd, bpf) */
46 AUTOFS_SUPER_MAGIC 0x0187
47 BDEVFS_MAGIC 0x62646576
48 BEFS_SUPER_MAGIC 0x42465331
49 BFS_MAGIC 0x1badface
50 BINFMTFS_MAGIC 0x42494e4d
51 BPF_FS_MAGIC 0xcafe4a11
52 BTRFS_SUPER_MAGIC 0x9123683e
53 BTRFS_TEST_MAGIC 0x73727279
54 CGROUP_SUPER_MAGIC 0x27e0eb /* Cgroup pseudo FS */
55 CGROUP2_SUPER_MAGIC 0x63677270 /* Cgroup v2 pseudo FS */
56 CIFS_MAGIC_NUMBER 0xff534d42
57 CODA_SUPER_MAGIC 0x73757245
58 COH_SUPER_MAGIC 0x012ff7b7
59 CRAMFS_MAGIC 0x28cd3d45
60 DEBUGFS_MAGIC 0x64626720
61 DEVFS_SUPER_MAGIC 0x1373 /* Linux 2.6.17 and earlier */
62 DEVPTS_SUPER_MAGIC 0x1cd1
63 ECRYPTFS_SUPER_MAGIC 0xf15f
64 EFIVARFS_MAGIC 0xde5e81e4
65 EFS_SUPER_MAGIC 0x00414a53
66 EXT_SUPER_MAGIC 0x137d /* Linux 2.0 and earlier */
67 EXT2_OLD_SUPER_MAGIC 0xef51
68 EXT2_SUPER_MAGIC 0xef53
69 EXT3_SUPER_MAGIC 0xef53
70 EXT4_SUPER_MAGIC 0xef53
71 F2FS_SUPER_MAGIC 0xf2f52010
72 FUSE_SUPER_MAGIC 0x65735546
73 FUTEXFS_SUPER_MAGIC 0xbad1dea /* Unused */
74 HFS_SUPER_MAGIC 0x4244
75 HOSTFS_SUPER_MAGIC 0x00c0ffee
76 HPFS_SUPER_MAGIC 0xf995e849
77 HUGETLBFS_MAGIC 0x958458f6
78 ISOFS_SUPER_MAGIC 0x9660
79 JFFS2_SUPER_MAGIC 0x72b6
80 JFS_SUPER_MAGIC 0x3153464a
81 MINIX_SUPER_MAGIC 0x137f /* original minix FS */
82 MINIX_SUPER_MAGIC2 0x138f /* 30 char minix FS */
83 MINIX2_SUPER_MAGIC 0x2468 /* minix V2 FS */
84 MINIX2_SUPER_MAGIC2 0x2478 /* minix V2 FS, 30 char names */
85 MINIX3_SUPER_MAGIC 0x4d5a /* minix V3 FS, 60 char names */
86 MQUEUE_MAGIC 0x19800202 /* POSIX message queue FS */
87 MSDOS_SUPER_MAGIC 0x4d44
88 MTD_INODE_FS_MAGIC 0x11307854
89 NCP_SUPER_MAGIC 0x564c
90 NFS_SUPER_MAGIC 0x6969
91 NILFS_SUPER_MAGIC 0x3434
92 NSFS_MAGIC 0x6e736673
93 NTFS_SB_MAGIC 0x5346544e
94 OCFS2_SUPER_MAGIC 0x7461636f
95 OPENPROM_SUPER_MAGIC 0x9fa1
96 OVERLAYFS_SUPER_MAGIC 0x794c7630
97 PIPEFS_MAGIC 0x50495045
98 PROC_SUPER_MAGIC 0x9fa0 /* /proc FS */
99 PSTOREFS_MAGIC 0x6165676c
100 QNX4_SUPER_MAGIC 0x002f
101 QNX6_SUPER_MAGIC 0x68191122
102 RAMFS_MAGIC 0x858458f6
103 REISERFS_SUPER_MAGIC 0x52654973
104 ROMFS_MAGIC 0x7275
105 SECURITYFS_MAGIC 0x73636673
106 SELINUX_MAGIC 0xf97cff8c
107 SMACK_MAGIC 0x43415d53
108 SMB_SUPER_MAGIC 0x517b
109 SMB2_MAGIC_NUMBER 0xfe534d42
110 SOCKFS_MAGIC 0x534f434b
111 SQUASHFS_MAGIC 0x73717368
112 SYSFS_MAGIC 0x62656572
113 SYSV2_SUPER_MAGIC 0x012ff7b6
114 SYSV4_SUPER_MAGIC 0x012ff7b5
115 TMPFS_MAGIC 0x01021994
116 TRACEFS_MAGIC 0x74726163
117 UDF_SUPER_MAGIC 0x15013346
118 UFS_MAGIC 0x00011954
119 USBDEVICE_SUPER_MAGIC 0x9fa2
120 V9FS_MAGIC 0x01021997
121 VXFS_SUPER_MAGIC 0xa501fcf5
122 XENFS_SUPER_MAGIC 0xabba1974
123 XENIX_SUPER_MAGIC 0x012ff7b4
124 XFS_SUPER_MAGIC 0x58465342
125 _XIAFS_SUPER_MAGIC 0x012fd16d /* Linux 2.0 and earlier */
126 Most of these MAGIC constants are defined in /usr/in‐
128 clude/linux/magic.h, and some are hardcoded in kernel sources.
129 The f_flags field is a bit mask indicating mount options for the
131 filesystem. It contains zero or more of the following bits:
132 ST_MANDLOCK
134 Mandatory locking is permitted on the filesystem (see fcntl(2)).
135 ST_NOATIME
137 Do not update access times; see mount(2).
138 ST_NODEV
140 Disallow access to device special files on this filesystem.
141 ST_NODIRATIME
143 Do not update directory access times; see mount(2).
144 ST_NOEXEC
146 Execution of programs is disallowed on this filesystem.
147 ST_NOSUID
149 The set-user-ID and set-group-ID bits are ignored by exec(3) for
150 executable files on this filesystem
151 ST_RDONLY
153 This filesystem is mounted read-only.
154 ST_RELATIME
156 Update atime relative to mtime/ctime; see mount(2).
157 ST_SYNCHRONOUS
159 Writes are synched to the filesystem immediately (see the de‐
160 scription of O_SYNC in open(2)).
161 ST_NOSYMFOLLOW (since Linux 5.10)
163 Symbolic links are not followed when resolving paths; see
164 mount(2).
165 Nobody knows what f_fsid is supposed to contain (but see below).
167 Fields that are undefined for a particular filesystem are set to 0.
169 fstatfs() returns the same information about an open file referenced by
171 descriptor fd.
172
174 On success, zero is returned. On error, -1 is returned, and errno is
175 set to indicate the error.
176
178 EACCES (statfs()) Search permission is denied for a component of the
179 path prefix of path. (See also path_resolution(7).)
180 EBADF (fstatfs()) fd is not a valid open file descriptor.
182 EFAULT buf or path points to an invalid address.
184 EINTR The call was interrupted by a signal; see signal(7).
186 EIO An I/O error occurred while reading from the filesystem.
188 ELOOP (statfs()) Too many symbolic links were encountered in translat‐
190 ing path.
191 ENAMETOOLONG
193 (statfs()) path is too long.
194 ENOENT (statfs()) The file referred to by path does not exist.
196 ENOMEM Insufficient kernel memory was available.
198 ENOSYS The filesystem does not support this call.
200 ENOTDIR
202 (statfs()) A component of the path prefix of path is not a di‐
203 rectory.
204 EOVERFLOW
206 Some values were too large to be represented in the returned
207 struct.
208
210 Linux-specific. The Linux statfs() was inspired by the 4.4BSD one (but
211 they do not use the same structure).
212
214 The __fsword_t type used for various fields in the statfs structure
215 definition is a glibc internal type, not intended for public use. This
216 leaves the programmer in a bit of a conundrum when trying to copy or
217 compare these fields to local variables in a program. Using un‐
218 signed int for such variables suffices on most systems.
219 The original Linux statfs() and fstatfs() system calls were not de‐
221 signed with extremely large file sizes in mind. Subsequently, Linux
222 2.6 added new statfs64() and fstatfs64() system calls that employ a new
223 structure, statfs64. The new structure contains the same fields as the
224 original statfs structure, but the sizes of various fields are in‐
225 creased, to accommodate large file sizes. The glibc statfs() and fs‐
226 tatfs() wrapper functions transparently deal with the kernel differ‐
227 ences.
228 Some systems have only <sys/vfs.h>, other systems also have
230 <sys/statfs.h>, where the former includes the latter. So it seems in‐
231 cluding the former is the best choice.
232 LSB has deprecated the library calls statfs() and fstatfs() and tells
234 us to use statvfs(3) and fstatvfs(3) instead.
235 The f_fsid field
237 Solaris, Irix, and POSIX have a system call statvfs(2) that returns a
238 struct statvfs (defined in <sys/statvfs.h>) containing an unsigned long
239 f_fsid. Linux, SunOS, HP-UX, 4.4BSD have a system call statfs() that
240 returns a struct statfs (defined in <sys/vfs.h>) containing a fsid_t
241 f_fsid, where fsid_t is defined as struct { int val[2]; }. The same
242 holds for FreeBSD, except that it uses the include file <sys/mount.h>.
243 The general idea is that f_fsid contains some random stuff such that
245 the pair (f_fsid,ino) uniquely determines a file. Some operating sys‐
246 tems use (a variation on) the device number, or the device number com‐
247 bined with the filesystem type. Several operating systems restrict
248 giving out the f_fsid field to the superuser only (and zero it for un‐
249 privileged users), because this field is used in the filehandle of the
250 filesystem when NFS-exported, and giving it out is a security concern.
251 Under some operating systems, the fsid can be used as the second argu‐
253 ment to the sysfs(2) system call.
254
256 From Linux 2.6.38 up to and including Linux 3.1, fstatfs() failed with
257 the error ENOSYS for file descriptors created by pipe(2).
258
260 stat(2), statvfs(3), path_resolution(7)
261
263 This page is part of release 5.12 of the Linux man-pages project. A
264 description of the project, information about reporting bugs, and the
265 latest version of this page, can be found at
266 https://www.kernel.org/doc/man-pages/.
267Linux 2021-03-22 STATFS(2)