man7/inotify.7

   1 '\" t
   2 .\" Copyright (C) 2006 Michael Kerrisk <mtk.manpages@gmail.com>
   3 .\"
   4 .\" %%%LICENSE_START(VERBATIM)
   5 .\" Permission is granted to make and distribute verbatim copies of this
   6 .\" manual provided the copyright notice and this permission notice are
   7 .\" preserved on all copies.
   8 .\"
   9 .\" Permission is granted to copy and distribute modified versions of this
  10 .\" manual under the conditions for verbatim copying, provided that the
  11 .\" entire resulting derived work is distributed under the terms of a
  12 .\" permission notice identical to this one.
  13 .\"
  14 .\" Since the Linux kernel and libraries are constantly changing, this
  15 .\" manual page may be incorrect or out-of-date.  The author(s) assume no
  16 .\" responsibility for errors or omissions, or for damages resulting from
  17 .\" the use of the information contained herein.  The author(s) may not
  18 .\" have taken the same level of care in the production of this manual,
  19 .\" which is licensed free of charge, as they might when working
  20 .\" professionally.
  21 .\"
  22 .\" Formatted or processed versions of this manual, if unaccompanied by
  23 .\" the source, must acknowledge the copyright and authors of this work.
  24 .\" %%%LICENSE_END
  25 .\"
  26 .TH INOTIFY 7 2014-03-28 "Linux" "Linux Programmer's Manual"
  27 .SH NAME
  28 inotify \- monitoring filesystem events
  29 .SH DESCRIPTION
  30 The
  31 .I inotify
  32 API provides a mechanism for monitoring filesystem events.
  33 Inotify can be used to monitor individual files,
  34 or to monitor directories.
  35 When a directory is monitored, inotify will return events
  36 for the directory itself, and for files inside the directory.
  37
  38 The following system calls are used with this API:
  39 .BR inotify_init (2)
  40 (or
  41 .BR inotify_init1 (2)),
  42 .BR inotify_add_watch (2),
  43 .BR inotify_rm_watch (2),
  44 .BR read (2),
  45 and
  46 .BR close (2).
  47
  48 .BR inotify_init (2)
  49 creates an inotify instance and returns a file descriptor
  50 referring to the inotify instance.
  51 The more recent
  52 .BR inotify_init1 (2)
  53 is like
  54 .BR inotify_init (2),
  55 but provides some extra functionality.
  56
  57 .BR inotify_add_watch (2)
  58 manipulates the "watch list" associated with an inotify instance.
  59 Each item ("watch") in the watch list specifies the pathname of
  60 a file or directory,
  61 along with some set of events that the kernel should monitor for the
  62 file referred to by that pathname.
  63 .BR inotify_add_watch (2)
  64 either creates a new watch item, or modifies an existing watch.
  65 Each watch has a unique "watch descriptor", an integer
  66 returned by
  67 .BR inotify_add_watch (2)
  68 when the watch is created.
  69
  70 .BR inotify_rm_watch (2)
  71 removes an item from an inotify watch list.
  72
  73 When all file descriptors referring to an inotify
  74 instance have been closed,
  75 the underlying object and its resources are
  76 freed for reuse by the kernel;
  77 all associated watches are automatically freed.
  78
  79 To determine what events have occurred, an application
  80 .BR read (2)s
  81 from the inotify file descriptor.
  82 If no events have so far occurred, then,
  83 assuming a blocking file descriptor,
  84 .BR read (2)
  85 will block until at least one event occurs
  86 (unless interrupted by a signal,
  87 in which case the call fails with the error
  88 .BR EINTR ;
  89 see
  90 .BR signal (7)).
  91
  92 Each successful
  93 .BR read (2)
  94 returns a buffer containing one or more of the following structures:
  95 .in +4n
  96 .nf
  97
  98 struct inotify_event {
  99     int      wd;       /* Watch descriptor */
 100 .\" FIXME . The type of the 'wd' field should probably be "int32_t".
 101 .\" I submitted a patch to fix this.  See the LKML thread
 102 .\" "[patch] Fix type errors in inotify interfaces", 18 Nov 2008
 103 .\" Glibc bug filed: http://sources.redhat.com/bugzilla/show_bug.cgi?id=7040
 104     uint32_t mask;     /* Mask of events */
 105     uint32_t cookie;   /* Unique cookie associating related
 106                           events (for rename(2)) */
 107     uint32_t len;      /* Size of \fIname\fP field */
 108     char     name[];   /* Optional null-terminated name */
 109 };
 110 .fi
 111 .in
 112
 113 .I wd
 114 identifies the watch for which this event occurs.
 115 It is one of the watch descriptors returned by a previous call to
 116 .BR inotify_add_watch (2).
 117
 118 .I mask
 119 contains bits that describe the event that occurred (see below).
 120
 121 .I cookie
 122 is a unique integer that connects related events.
 123 Currently this is used only for rename events, and
 124 allows the resulting pair of
 125 .B IN_MOVED_FROM
 126 and
 127 .B IN_MOVED_TO
 128 events to be connected by the application.
 129 For all other event types,
 130 .I cookie
 131 is set to 0.
 132
 133 The
 134 .I name
 135 field is present only when an event is returned
 136 for a file inside a watched directory;
 137 it identifies the file pathname relative to the watched directory.
 138 This pathname is null-terminated,
 139 and may include further null bytes (\(aq\\0\(aq) to align subsequent reads to a
 140 suitable address boundary.
 141
 142 The
 143 .I len
 144 field counts all of the bytes in
 145 .IR name ,
 146 including the null bytes;
 147 the length of each
 148 .I inotify_event
 149 structure is thus
 150 .IR "sizeof(struct inotify_event)+len" .
 151
 152 The behavior when the buffer given to
 153 .BR read (2)
 154 is too small to return information about the next event depends
 155 on the kernel version: in kernels before 2.6.21,
 156 .BR read (2)
 157 returns 0; since kernel 2.6.21,
 158 .BR read (2)
 159 fails with the error
 160 .BR EINVAL .
 161 Specifying a buffer of size
 162
 163     sizeof(struct inotify_event) + NAME_MAX + 1
 164
 165 will be sufficient to read at least one event.
 166 .SS inotify events
 167 The
 168 .BR inotify_add_watch (2)
 169 .I mask
 170 argument and the
 171 .I mask
 172 field of the
 173 .I inotify_event
 174 structure returned when
 175 .BR read (2)ing
 176 an inotify file descriptor are both bit masks identifying
 177 inotify events.
 178 The following bits can be specified in
 179 .I mask
 180 when calling
 181 .BR inotify_add_watch (2)
 182 and may be returned in the
 183 .I mask
 184 field returned by
 185 .BR read (2):
 186 .RS 4
 187 .TP
 188 .BR IN_ACCESS " (*)"
 189 File was accessed (e.g.,
 190 .BR read (2),
 191 .BR execve (2)).
 192 .TP
 193 .BR IN_ATTRIB " (*)"
 194 Metadata changed\(emfor example, permissions, timestamps, extended attributes,
 195 link count (since Linux 2.6.25), user ID, or group ID.
 196 .TP
 197 .BR IN_CLOSE_WRITE " (*)"
 198 File opened for writing was closed.
 199 .TP
 200 .BR IN_CLOSE_NOWRITE " (*)"
 201 File not opened for writing was closed.
 202 .TP
 203 .BR IN_CREATE " (*)"
 204 File/directory created in watched directory.
 205 .TP
 206 .BR IN_DELETE " (*)"
 207 File/directory deleted from watched directory.
 208 .TP
 209 .B IN_DELETE_SELF
 210 Watched file/directory was itself deleted.
 211 (This event also occurs if an object is moved to another filesystem,
 212 since
 213 .BR mv (1)
 214 in effect copies the file to the other filesystem and
 215 then deletes it from the original filesystem.)
 216 .TP
 217 .BR IN_MODIFY " (*)"
 218 File was modified (e.g.,
 219 .BR write (2),
 220 .BR truncate (2)).
 221 .TP
 222 .B IN_MOVE_SELF
 223 Watched file/directory was itself moved.
 224 .TP
 225 .BR IN_MOVED_FROM " (*)"
 226 Generated for the directory containing the old filename
 227 when a file is renamed.
 228 .TP
 229 .BR IN_MOVED_TO " (*)"
 230 Generated for the directory containing the new filename
 231 when a file is renamed.
 232 .TP
 233 .BR IN_OPEN " (*)"
 234 File was opened.
 235 .RE
 236 .PP
 237 When monitoring a directory,
 238 the events marked with an asterisk (*) above can occur for
 239 files in the directory, in which case the
 240 .I name
 241 field in the returned
 242 .I inotify_event
 243 structure identifies the name of the file within the directory.
 244 .PP
 245 The
 246 .B IN_ALL_EVENTS
 247 macro is defined as a bit mask of all of the above events.
 248 This macro can be used as the
 249 .I mask
 250 argument when calling
 251 .BR inotify_add_watch (2).
 252
 253 Two additional convenience macros are defined:
 254 .RS 4
 255 .TP
 256 .BR IN_MOVE
 257 Equates to
 258 .BR "IN_MOVED_FROM | IN_MOVED_TO" .
 259 .TP
 260 .BR IN_CLOSE
 261 Equates to
 262 .BR "IN_CLOSE_WRITE | IN_CLOSE_NOWRITE" .
 263 .RE
 264 .PP
 265 The following further bits can be specified in
 266 .I mask
 267 when calling
 268 .BR inotify_add_watch (2):
 269 .RS 4
 270 .TP
 271 .BR IN_DONT_FOLLOW " (since Linux 2.6.15)"
 272 Don't dereference
 273 .I pathname
 274 if it is a symbolic link.
 275 .TP
 276 .BR IN_EXCL_UNLINK " (since Linux 2.6.36)"
 277 .\" commit 8c1934c8d70b22ca8333b216aec6c7d09fdbd6a6
 278 By default, when watching events on the children of a directory,
 279 events are generated for children even after they have been unlinked
 280 from the directory.
 281 This can result in large numbers of uninteresting events for
 282 some applications (e.g., if watching
 283 .IR /tmp ,
 284 in which many applications create temporary files whose
 285 names are immediately unlinked).
 286 Specifying
 287 .B IN_EXCL_UNLINK
 288 changes the default behavior,
 289 so that events are not generated for children after
 290 they have been unlinked from the watched directory.
 291 .TP
 292 .B IN_MASK_ADD
 293 Add (OR) events to watch mask for this pathname if
 294 it already exists (instead of replacing mask).
 295 .TP
 296 .B IN_ONESHOT
 297 Monitor
 298 .I pathname
 299 for one event, then remove from
 300 watch list.
 301 .TP
 302 .BR IN_ONLYDIR " (since Linux 2.6.15)"
 303 Only watch
 304 .I pathname
 305 if it is a directory.
 306 .RE
 307 .PP
 308 The following bits may be set in the
 309 .I mask
 310 field returned by
 311 .BR read (2):
 312 .RS 4
 313 .TP
 314 .B IN_IGNORED
 315 Watch was removed explicitly
 316 .RB ( inotify_rm_watch (2))
 317 or automatically (file was deleted, or filesystem was unmounted).
 318 See also BUGS.
 319 .TP
 320 .B IN_ISDIR
 321 Subject of this event is a directory.
 322 .TP
 323 .B IN_Q_OVERFLOW
 324 Event queue overflowed
 325 .RI ( wd
 326 is \-1 for this event).
 327 .TP
 328 .B IN_UNMOUNT
 329 Filesystem containing watched object was unmounted.
 330 .RE
 331 .SS /proc interfaces
 332 The following interfaces can be used to limit the amount of
 333 kernel memory consumed by inotify:
 334 .TP
 335 .I /proc/sys/fs/inotify/max_queued_events
 336 The value in this file is used when an application calls
 337 .BR inotify_init (2)
 338 to set an upper limit on the number of events that can be
 339 queued to the corresponding inotify instance.
 340 Events in excess of this limit are dropped, but an
 341 .B IN_Q_OVERFLOW
 342 event is always generated.
 343 .TP
 344 .I /proc/sys/fs/inotify/max_user_instances
 345 This specifies an upper limit on the number of inotify instances
 346 that can be created per real user ID.
 347 .TP
 348 .I /proc/sys/fs/inotify/max_user_watches
 349 This specifies an upper limit on the number of watches
 350 that can be created per real user ID.
 351 .SH VERSIONS
 352 Inotify was merged into the 2.6.13 Linux kernel.
 353 The required library interfaces were added to glibc in version 2.4.
 354 .RB ( IN_DONT_FOLLOW ,
 355 .BR IN_MASK_ADD ,
 356 and
 357 .B IN_ONLYDIR
 358 were added in glibc version 2.5.)
 359 .SH CONFORMING TO
 360 The inotify API is Linux-specific.
 361 .SH NOTES
 362 Inotify file descriptors can be monitored using
 363 .BR select (2),
 364 .BR poll (2),
 365 and
 366 .BR epoll (7).
 367 When an event is available, the file descriptor indicates as readable.
 368
 369 Since Linux 2.6.25,
 370 signal-driven I/O notification is available for inotify file descriptors;
 371 see the discussion of
 372 .B F_SETFL
 373 (for setting the
 374 .B O_ASYNC
 375 flag),
 376 .BR F_SETOWN ,
 377 and
 378 .B F_SETSIG
 379 in
 380 .BR fcntl (2).
 381 The
 382 .I siginfo_t
 383 structure (described in
 384 .BR sigaction (2))
 385 that is passed to the signal handler has the following fields set:
 386 .IR si_fd
 387 is set to the inotify file descriptor number;
 388 .IR si_signo
 389 is set to the signal number;
 390 .IR si_code
 391 is set to
 392 .BR POLL_IN ;
 393 and
 394 .B POLLIN
 395 is set in
 396 .IR si_band .
 397
 398 If successive output inotify events produced on the
 399 inotify file descriptor are identical (same
 400 .IR wd ,
 401 .IR mask ,
 402 .IR cookie ,
 403 and
 404 .IR name ),
 405 then they are coalesced into a single event if the
 406 older event has not yet been read (but see BUGS).
 407 This reduces the amount of kernel memory required for the event queue,
 408 but also means that an application can't use inotify to reliably count
 409 file events.
 410
 411 The events returned by reading from an inotify file descriptor
 412 form an ordered queue.
 413 Thus, for example, it is guaranteed that when renaming from
 414 one directory to another, events will be produced in the
 415 correct order on the inotify file descriptor.
 416
 417 The
 418 .B FIONREAD
 419 .BR ioctl (2)
 420 returns the number of bytes available to read from an
 421 inotify file descriptor.
 422 .SS Limitations and caveats
 423 Inotify monitoring of directories is not recursive:
 424 to monitor subdirectories under a directory,
 425 additional watches must be created.
 426 This can take a significant amount time for large directory trees.
 427
 428 The inotify API provides no information about the user or process that
 429 triggered the inotify event.
 430 In particular, there is no easy
 431 way for a process that is monitoring events via inotify
 432 to distinguish events that it triggers
 433 itself from those that are triggered by other processes.
 434
 435 Note that the event queue can overflow.
 436 In this case, events are lost.
 437 Robust applications should handle the possibility of
 438 lost events gracefully.
 439
 440 The inotify API identifies affected files by filename.
 441 However, by the time an application processes an inotify event,
 442 the filename may already have been deleted or renamed.
 443
 444 If monitoring an entire directory subtree,
 445 and a new subdirectory is created in that tree or an existing directory
 446 is renamed into that tree,
 447 be aware that by the time you create a watch for the new subdirectory,
 448 new files (and subdirectories) may already exist inside the subdirectory.
 449 Therefore, you may want to scan the contents of the subdirectory
 450 immediately after adding the watch (and, if desired,
 451 recursively add watches for any subdirectories that it contains).
 452
 453 The inotify applications identifies events via watch descriptors.
 454 It is the application's responsibility to cache a mapping
 455 (if one is needed) between watch descriptors and pathnames.
 456 Be aware that directory renamings may affect multiple cached pathnames.
 457 .SH BUGS
 458 .\" FIXME kernel commit 611da04f7a31b2208e838be55a42c7a1310ae321
 459 .\" implies that unmount events were buggy 2.6.11 to 2.6.36
 460 .\"
 461 In kernels before 2.6.16, the
 462 .B IN_ONESHOT
 463 .I mask
 464 flag does not work.
 465
 466 As originally designed and implemented, the
 467 .B IN_ONESHOT
 468 flag did not cause an
 469 .B IN_IGNORED
 470 event to be generated when the watch was dropped after one event.
 471 However, as an unintended effect of other changes,
 472 since Linux 2.6.36, an
 473 .B IN_IGNORED
 474 event is generated in this case.
 475
 476 Before kernel 2.6.25,
 477 .\" commit 1c17d18e3775485bf1e0ce79575eb637a94494a2
 478 the kernel code that was intended to coalesce successive identical events
 479 (i.e., the two most recent events could potentially be coalesced
 480 if the older had not yet been read)
 481 instead checked if the most recent event could be coalesced with the
 482 .I oldest
 483 unread event.
 484 .SH SEE ALSO
 485 .BR inotifywait (1),
 486 .BR inotifywatch (1),
 487 .BR inotify_add_watch (2),
 488 .BR inotify_init (2),
 489 .BR inotify_init1 (2),
 490 .BR inotify_rm_watch (2),
 491 .BR read (2),
 492 .BR stat (2)
 493
 494 .IR Documentation/filesystems/inotify.txt
 495 in the Linux kernel source tree