1flock(2) System Calls Manual flock(2)
2
6 flock - apply or remove an advisory lock on an open file
7
9 Standard C library (libc, -lc)
10
12 #include <sys/file.h>
13 int flock(int fd, int operation);
15
17 Apply or remove an advisory lock on the open file specified by fd. The
18 argument operation is one of the following:
19 LOCK_SH Place a shared lock. More than one process may hold a
21 shared lock for a given file at a given time.
22 LOCK_EX Place an exclusive lock. Only one process may hold an ex‐
24 clusive lock for a given file at a given time.
25 LOCK_UN Remove an existing lock held by this process.
27 A call to flock() may block if an incompatible lock is held by another
29 process. To make a nonblocking request, include LOCK_NB (by ORing)
30 with any of the above operations.
31 A single file may not simultaneously have both shared and exclusive
33 locks.
34 Locks created by flock() are associated with an open file description
36 (see open(2)). This means that duplicate file descriptors (created by,
37 for example, fork(2) or dup(2)) refer to the same lock, and this lock
38 may be modified or released using any of these file descriptors. Fur‐
39 thermore, the lock is released either by an explicit LOCK_UN operation
40 on any of these duplicate file descriptors, or when all such file de‐
41 scriptors have been closed.
42 If a process uses open(2) (or similar) to obtain more than one file de‐
44 scriptor for the same file, these file descriptors are treated indepen‐
45 dently by flock(). An attempt to lock the file using one of these file
46 descriptors may be denied by a lock that the calling process has al‐
47 ready placed via another file descriptor.
48 A process may hold only one type of lock (shared or exclusive) on a
50 file. Subsequent flock() calls on an already locked file will convert
51 an existing lock to the new lock mode.
52 Locks created by flock() are preserved across an execve(2).
54 A shared or exclusive lock can be placed on a file regardless of the
56 mode in which the file was opened.
57
59 On success, zero is returned. On error, -1 is returned, and errno is
60 set to indicate the error.
61
63 EBADF fd is not an open file descriptor.
64 EINTR While waiting to acquire a lock, the call was interrupted by de‐
66 livery of a signal caught by a handler; see signal(7).
67 EINVAL operation is invalid.
69 ENOLCK The kernel ran out of memory for allocating lock records.
71 EWOULDBLOCK
73 The file is locked and the LOCK_NB flag was selected.
74
76 Since Linux 2.0, flock() is implemented as a system call in its own
77 right rather than being emulated in the GNU C library as a call to fc‐
78 ntl(2). With this implementation, there is no interaction between the
79 types of lock placed by flock() and fcntl(2), and flock() does not de‐
80 tect deadlock. (Note, however, that on some systems, such as the mod‐
81 ern BSDs, flock() and fcntl(2) locks do interact with one another.)
82 CIFS details
84 Up to Linux 5.4, flock() is not propagated over SMB. A file with such
85 locks will not appear locked for remote clients.
86 Since Linux 5.5, flock() locks are emulated with SMB byte-range locks
88 on the entire file. Similarly to NFS, this means that fcntl(2) and
89 flock() locks interact with one another. Another important side-effect
90 is that the locks are not advisory anymore: any IO on a locked file
91 will always fail with EACCES when done from a separate file descriptor.
92 This difference originates from the design of locks in the SMB proto‐
93 col, which provides mandatory locking semantics.
94 Remote and mandatory locking semantics may vary with SMB protocol,
96 mount options and server type. See mount.cifs(8) for additional infor‐
97 mation.
98
100 BSD.
101
103 4.4BSD (the flock() call first appeared in 4.2BSD). A version of
104 flock(), possibly implemented in terms of fcntl(2), appears on most
105 UNIX systems.
106 NFS details
108 Up to Linux 2.6.11, flock() does not lock files over NFS (i.e., the
109 scope of locks was limited to the local system). Instead, one could
110 use fcntl(2) byte-range locking, which does work over NFS, given a suf‐
111 ficiently recent version of Linux and a server which supports locking.
112 Since Linux 2.6.12, NFS clients support flock() locks by emulating them
114 as fcntl(2) byte-range locks on the entire file. This means that fc‐
115 ntl(2) and flock() locks do interact with one another over NFS. It
116 also means that in order to place an exclusive lock, the file must be
117 opened for writing.
118 Since Linux 2.6.37, the kernel supports a compatibility mode that al‐
120 lows flock() locks (and also fcntl(2) byte region locks) to be treated
121 as local; see the discussion of the local_lock option in nfs(5).
122
124 flock() places advisory locks only; given suitable permissions on a
125 file, a process is free to ignore the use of flock() and perform I/O on
126 the file.
127 flock() and fcntl(2) locks have different semantics with respect to
129 forked processes and dup(2). On systems that implement flock() using
130 fcntl(2), the semantics of flock() will be different from those de‐
131 scribed in this manual page.
132 Converting a lock (shared to exclusive, or vice versa) is not guaran‐
134 teed to be atomic: the existing lock is first removed, and then a new
135 lock is established. Between these two steps, a pending lock request
136 by another process may be granted, with the result that the conversion
137 either blocks, or fails if LOCK_NB was specified. (This is the origi‐
138 nal BSD behavior, and occurs on many other implementations.)
139
141 flock(1), close(2), dup(2), execve(2), fcntl(2), fork(2), open(2),
142 lockf(3), lslocks(8)
143 Documentation/filesystems/locks.txt in the Linux kernel source tree
145 (Documentation/locks.txt in older kernels)
146Linux man-pages 6.04 2023-03-30 flock(2)