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  15 .\" manual page may be incorrect or out-of-date.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 (e.g.,
 195 .BR chmod (2)),
 196 timestamps (e.g.,
 197 .BR utimensat (2)),
 198 extended attributes
 199 .RB ( setxattr (2)),
 200 link count (since Linux 2.6.25; e.g.,
 201 for the target of
 202 .BR link (2)
 203 and for
 204 .BR unlink (2)),
 205 and user/group ID (e.g.,
 206 .BR chown (2)).
 207 .TP
 208 .BR IN_CLOSE_WRITE " (*)"
 209 File opened for writing was closed.
 210 .TP
 211 .BR IN_CLOSE_NOWRITE " (*)"
 212 File not opened for writing was closed.
 213 .TP
 214 .BR IN_CREATE " (*)"
 215 File/directory created in watched directory (e.g.,
 216 .BR open (2)
 217 .BR O_CREAT ,
 218 .BR mkdir (2),
 219 .BR link (2),
 220 .BR bind (2)
 221 on a UNIX domain socket).
 222 .TP
 223 .BR IN_DELETE " (*)"
 224 File/directory deleted from watched directory.
 225 .TP
 226 .B IN_DELETE_SELF
 227 Watched file/directory was itself deleted.
 228 (This event also occurs if an object is moved to another filesystem,
 229 since
 230 .BR mv (1)
 231 in effect copies the file to the other filesystem and
 232 then deletes it from the original filesystem.)
 233 .TP
 234 .BR IN_MODIFY " (*)"
 235 File was modified (e.g.,
 236 .BR write (2),
 237 .BR truncate (2)).
 238 .TP
 239 .B IN_MOVE_SELF
 240 Watched file/directory was itself moved.
 241 .TP
 242 .BR IN_MOVED_FROM " (*)"
 243 Generated for the directory containing the old filename
 244 when a file is renamed.
 245 .TP
 246 .BR IN_MOVED_TO " (*)"
 247 Generated for the directory containing the new filename
 248 when a file is renamed.
 249 .TP
 250 .BR IN_OPEN " (*)"
 251 File was opened.
 252 .RE
 253 .PP
 254 When monitoring a directory,
 255 the events marked with an asterisk (*) above can occur for
 256 files in the directory, in which case the
 257 .I name
 258 field in the returned
 259 .I inotify_event
 260 structure identifies the name of the file within the directory.
 261 .PP
 262 The
 263 .B IN_ALL_EVENTS
 264 macro is defined as a bit mask of all of the above events.
 265 This macro can be used as the
 266 .I mask
 267 argument when calling
 268 .BR inotify_add_watch (2).
 269
 270 Two additional convenience macros are defined:
 271 .RS 4
 272 .TP
 273 .BR IN_MOVE
 274 Equates to
 275 .BR "IN_MOVED_FROM | IN_MOVED_TO" .
 276 .TP
 277 .BR IN_CLOSE
 278 Equates to
 279 .BR "IN_CLOSE_WRITE | IN_CLOSE_NOWRITE" .
 280 .RE
 281 .PP
 282 The following further bits can be specified in
 283 .I mask
 284 when calling
 285 .BR inotify_add_watch (2):
 286 .RS 4
 287 .TP
 288 .BR IN_DONT_FOLLOW " (since Linux 2.6.15)"
 289 Don't dereference
 290 .I pathname
 291 if it is a symbolic link.
 292 .TP
 293 .BR IN_EXCL_UNLINK " (since Linux 2.6.36)"
 294 .\" commit 8c1934c8d70b22ca8333b216aec6c7d09fdbd6a6
 295 By default, when watching events on the children of a directory,
 296 events are generated for children even after they have been unlinked
 297 from the directory.
 298 This can result in large numbers of uninteresting events for
 299 some applications (e.g., if watching
 300 .IR /tmp ,
 301 in which many applications create temporary files whose
 302 names are immediately unlinked).
 303 Specifying
 304 .B IN_EXCL_UNLINK
 305 changes the default behavior,
 306 so that events are not generated for children after
 307 they have been unlinked from the watched directory.
 308 .TP
 309 .B IN_MASK_ADD
 310 Add (OR) events to watch mask for this pathname if
 311 it already exists (instead of replacing mask).
 312 .TP
 313 .B IN_ONESHOT
 314 Monitor
 315 .I pathname
 316 for one event, then remove from
 317 watch list.
 318 .TP
 319 .BR IN_ONLYDIR " (since Linux 2.6.15)"
 320 Only watch
 321 .I pathname
 322 if it is a directory.
 323 .RE
 324 .PP
 325 The following bits may be set in the
 326 .I mask
 327 field returned by
 328 .BR read (2):
 329 .RS 4
 330 .TP
 331 .B IN_IGNORED
 332 Watch was removed explicitly
 333 .RB ( inotify_rm_watch (2))
 334 or automatically (file was deleted, or filesystem was unmounted).
 335 See also BUGS.
 336 .TP
 337 .B IN_ISDIR
 338 Subject of this event is a directory.
 339 .TP
 340 .B IN_Q_OVERFLOW
 341 Event queue overflowed
 342 .RI ( wd
 343 is \-1 for this event).
 344 .TP
 345 .B IN_UNMOUNT
 346 Filesystem containing watched object was unmounted.
 347 .RE
 348 .SS /proc interfaces
 349 The following interfaces can be used to limit the amount of
 350 kernel memory consumed by inotify:
 351 .TP
 352 .I /proc/sys/fs/inotify/max_queued_events
 353 The value in this file is used when an application calls
 354 .BR inotify_init (2)
 355 to set an upper limit on the number of events that can be
 356 queued to the corresponding inotify instance.
 357 Events in excess of this limit are dropped, but an
 358 .B IN_Q_OVERFLOW
 359 event is always generated.
 360 .TP
 361 .I /proc/sys/fs/inotify/max_user_instances
 362 This specifies an upper limit on the number of inotify instances
 363 that can be created per real user ID.
 364 .TP
 365 .I /proc/sys/fs/inotify/max_user_watches
 366 This specifies an upper limit on the number of watches
 367 that can be created per real user ID.
 368 .SH VERSIONS
 369 Inotify was merged into the 2.6.13 Linux kernel.
 370 The required library interfaces were added to glibc in version 2.4.
 371 .RB ( IN_DONT_FOLLOW ,
 372 .BR IN_MASK_ADD ,
 373 and
 374 .B IN_ONLYDIR
 375 were added in glibc version 2.5.)
 376 .SH CONFORMING TO
 377 The inotify API is Linux-specific.
 378 .SH NOTES
 379 Inotify file descriptors can be monitored using
 380 .BR select (2),
 381 .BR poll (2),
 382 and
 383 .BR epoll (7).
 384 When an event is available, the file descriptor indicates as readable.
 385
 386 Since Linux 2.6.25,
 387 signal-driven I/O notification is available for inotify file descriptors;
 388 see the discussion of
 389 .B F_SETFL
 390 (for setting the
 391 .B O_ASYNC
 392 flag),
 393 .BR F_SETOWN ,
 394 and
 395 .B F_SETSIG
 396 in
 397 .BR fcntl (2).
 398 The
 399 .I siginfo_t
 400 structure (described in
 401 .BR sigaction (2))
 402 that is passed to the signal handler has the following fields set:
 403 .IR si_fd
 404 is set to the inotify file descriptor number;
 405 .IR si_signo
 406 is set to the signal number;
 407 .IR si_code
 408 is set to
 409 .BR POLL_IN ;
 410 and
 411 .B POLLIN
 412 is set in
 413 .IR si_band .
 414
 415 If successive output inotify events produced on the
 416 inotify file descriptor are identical (same
 417 .IR wd ,
 418 .IR mask ,
 419 .IR cookie ,
 420 and
 421 .IR name ),
 422 then they are coalesced into a single event if the
 423 older event has not yet been read (but see BUGS).
 424 This reduces the amount of kernel memory required for the event queue,
 425 but also means that an application can't use inotify to reliably count
 426 file events.
 427
 428 The events returned by reading from an inotify file descriptor
 429 form an ordered queue.
 430 Thus, for example, it is guaranteed that when renaming from
 431 one directory to another, events will be produced in the
 432 correct order on the inotify file descriptor.
 433
 434 The
 435 .B FIONREAD
 436 .BR ioctl (2)
 437 returns the number of bytes available to read from an
 438 inotify file descriptor.
 439 .SS Limitations and caveats
 440 Inotify monitoring of directories is not recursive:
 441 to monitor subdirectories under a directory,
 442 additional watches must be created.
 443 This can take a significant amount time for large directory trees.
 444
 445 The inotify API provides no information about the user or process that
 446 triggered the inotify event.
 447 In particular, there is no easy
 448 way for a process that is monitoring events via inotify
 449 to distinguish events that it triggers
 450 itself from those that are triggered by other processes.
 451
 452 Note that the event queue can overflow.
 453 In this case, events are lost.
 454 Robust applications should handle the possibility of
 455 lost events gracefully.
 456
 457 The inotify API identifies affected files by filename.
 458 However, by the time an application processes an inotify event,
 459 the filename may already have been deleted or renamed.
 460
 461 If monitoring an entire directory subtree,
 462 and a new subdirectory is created in that tree or an existing directory
 463 is renamed into that tree,
 464 be aware that by the time you create a watch for the new subdirectory,
 465 new files (and subdirectories) may already exist inside the subdirectory.
 466 Therefore, you may want to scan the contents of the subdirectory
 467 immediately after adding the watch (and, if desired,
 468 recursively add watches for any subdirectories that it contains).
 469
 470 The inotify applications identifies events via watch descriptors.
 471 It is the application's responsibility to cache a mapping
 472 (if one is needed) between watch descriptors and pathnames.
 473 Be aware that directory renamings may affect multiple cached pathnames.
 474 .SH BUGS
 475 .\" FIXME kernel commit 611da04f7a31b2208e838be55a42c7a1310ae321
 476 .\" implies that unmount events were buggy 2.6.11 to 2.6.36
 477 .\"
 478 In kernels before 2.6.16, the
 479 .B IN_ONESHOT
 480 .I mask
 481 flag does not work.
 482
 483 As originally designed and implemented, the
 484 .B IN_ONESHOT
 485 flag did not cause an
 486 .B IN_IGNORED
 487 event to be generated when the watch was dropped after one event.
 488 However, as an unintended effect of other changes,
 489 since Linux 2.6.36, an
 490 .B IN_IGNORED
 491 event is generated in this case.
 492
 493 Before kernel 2.6.25,
 494 .\" commit 1c17d18e3775485bf1e0ce79575eb637a94494a2
 495 the kernel code that was intended to coalesce successive identical events
 496 (i.e., the two most recent events could potentially be coalesced
 497 if the older had not yet been read)
 498 instead checked if the most recent event could be coalesced with the
 499 .I oldest
 500 unread event.
 501 .SH SEE ALSO
 502 .BR inotifywait (1),
 503 .BR inotifywatch (1),
 504 .BR inotify_add_watch (2),
 505 .BR inotify_init (2),
 506 .BR inotify_init1 (2),
 507 .BR inotify_rm_watch (2),
 508 .BR read (2),
 509 .BR stat (2)
 510
 511 .IR Documentation/filesystems/inotify.txt
 512 in the Linux kernel source tree