man2/fallocate.2

   1 .\" Copyright (c) 2007 Silicon Graphics, Inc. All Rights Reserved
   2 .\" Written by Dave Chinner <dgc@sgi.com>
   3 .\"
   4 .\" %%%LICENSE_START(GPLv2_ONELINE)
   5 .\" May be distributed as per GNU General Public License version 2.
   6 .\" %%%LICENSE_END
   7 .\"
   8 .\" 2011-09-19: Added FALLOC_FL_PUNCH_HOLE
   9 .\" 2011-09-19: Substantial restructuring of the page
  10 .\"
  11 .TH FALLOCATE 2 2018-04-30 "Linux" "Linux Programmer's Manual"
  12 .SH NAME
  13 fallocate \- manipulate file space
  14 .SH SYNOPSIS
  15 .nf
  16 .BR "#define _GNU_SOURCE" "             /* See feature_test_macros(7) */"
  17 .B #include <fcntl.h>
  18 .PP
  19 .BI "int fallocate(int " fd ", int " mode ", off_t " offset \
  20 ", off_t " len ");
  21 .fi
  22 .SH DESCRIPTION
  23 This is a nonportable, Linux-specific system call.
  24 For the portable, POSIX.1-specified method of ensuring that space
  25 is allocated for a file, see
  26 .BR posix_fallocate (3).
  27 .PP
  28 .BR fallocate ()
  29 allows the caller to directly manipulate the allocated disk space
  30 for the file referred to by
  31 .I fd
  32 for the byte range starting at
  33 .I offset
  34 and continuing for
  35 .I len
  36 bytes.
  37 .PP
  38 The
  39 .I mode
  40 argument determines the operation to be performed on the given range.
  41 Details of the supported operations are given in the subsections below.
  42 .SS Allocating disk space
  43 The default operation (i.e.,
  44 .I mode
  45 is zero) of
  46 .BR fallocate ()
  47 allocates the disk space within the range specified by
  48 .I offset
  49 and
  50 .IR len .
  51 The file size (as reported by
  52 .BR stat (2))
  53 will be changed if
  54 .IR offset + len
  55 is greater than the file size.
  56 Any subregion within the range specified by
  57 .I offset
  58 and
  59 .IR len
  60 that did not contain data before the call will be initialized to zero.
  61 This default behavior closely resembles the behavior of the
  62 .BR posix_fallocate (3)
  63 library function,
  64 and is intended as a method of optimally implementing that function.
  65 .PP
  66 After a successful call, subsequent writes into the range specified by
  67 .IR offset
  68 and
  69 .IR len
  70 are guaranteed not to fail because of lack of disk space.
  71 .PP
  72 If the
  73 .B FALLOC_FL_KEEP_SIZE
  74 flag is specified in
  75 .IR mode ,
  76 the behavior of the call is similar,
  77 but the file size will not be changed even if
  78 .IR offset + len
  79 is greater than the file size.
  80 Preallocating zeroed blocks beyond the end of the file in this manner
  81 is useful for optimizing append workloads.
  82 .PP
  83 If the
  84 .B FALLOC_FL_UNSHARE
  85 flag is specified in
  86 .IR mode ,
  87 shared file data extents will be made private to the file to guarantee
  88 that a subsequent write will not fail due to lack of space.
  89 Typically, this will be done by performing a copy-on-write operation on
  90 all shared data in the file.
  91 This flag may not be supported by all filesystems.
  92 .PP
  93 Because allocation is done in block size chunks,
  94 .BR fallocate ()
  95 may allocate a larger range of disk space than was specified.
  96 .SS Deallocating file space
  97 Specifying the
  98 .BR FALLOC_FL_PUNCH_HOLE
  99 flag (available since Linux 2.6.38) in
 100 .I mode
 101 deallocates space (i.e., creates a hole)
 102 in the byte range starting at
 103 .I offset
 104 and continuing for
 105 .I len
 106 bytes.
 107 Within the specified range, partial filesystem blocks are zeroed,
 108 and whole filesystem blocks are removed from the file.
 109 After a successful call,
 110 subsequent reads from this range will return zeros.
 111 .PP
 112 The
 113 .BR FALLOC_FL_PUNCH_HOLE
 114 flag must be ORed with
 115 .BR FALLOC_FL_KEEP_SIZE
 116 in
 117 .IR mode ;
 118 in other words, even when punching off the end of the file, the file size
 119 (as reported by
 120 .BR stat (2))
 121 does not change.
 122 .PP
 123 Not all filesystems support
 124 .BR FALLOC_FL_PUNCH_HOLE ;
 125 if a filesystem doesn't support the operation, an error is returned.
 126 The operation is supported on at least the following filesystems:
 127 .IP * 3
 128 XFS (since Linux 2.6.38)
 129 .IP *
 130 ext4 (since Linux 3.0)
 131 .\" commit a4bb6b64e39abc0e41ca077725f2a72c868e7622
 132 .IP *
 133 Btrfs (since Linux 3.7)
 134 .IP *
 135 .BR tmpfs (5)
 136 (since Linux 3.5)"
 137 .\" commit 83e4fa9c16e4af7122e31be3eca5d57881d236fe
 138 .SS Collapsing file space
 139 .\" commit 00f5e61998dd17f5375d9dfc01331f104b83f841
 140 Specifying the
 141 .BR FALLOC_FL_COLLAPSE_RANGE
 142 flag (available since Linux 3.15) in
 143 .I mode
 144 removes a byte range from a file, without leaving a hole.
 145 The byte range to be collapsed starts at
 146 .I offset
 147 and continues for
 148 .I len
 149 bytes.
 150 At the completion of the operation,
 151 the contents of the file starting at the location
 152 .I offset+len
 153 will be appended at the location
 154 .IR offset ,
 155 and the file will be
 156 .I len
 157 bytes smaller.
 158 .PP
 159 A filesystem may place limitations on the granularity of the operation,
 160 in order to ensure efficient implementation.
 161 Typically,
 162 .I offset
 163 and
 164 .I len
 165 must be a multiple of the filesystem logical block size,
 166 which varies according to the filesystem type and configuration.
 167 If a filesystem has such a requirement,
 168 .BR fallocate ()
 169 fails with the error
 170 .BR EINVAL
 171 if this requirement is violated.
 172 .PP
 173 If the region specified by
 174 .I offset
 175 plus
 176 .I len
 177 reaches or passes the end of file, an error is returned;
 178 instead, use
 179 .BR ftruncate (2)
 180 to truncate a file.
 181 .PP
 182 No other flags may be specified in
 183 .IR mode
 184 in conjunction with
 185 .BR FALLOC_FL_COLLAPSE_RANGE .
 186 .PP
 187 As at Linux 3.15,
 188 .B FALLOC_FL_COLLAPSE_RANGE
 189 is supported by
 190 ext4 (only for extent-based files)
 191 .\" commit 9eb79482a97152930b113b51dff530aba9e28c8e
 192 and XFS.
 193 .\" commit e1d8fb88a64c1f8094b9f6c3b6d2d9e6719c970d
 194 .SS Zeroing file space
 195 Specifying the
 196 .BR FALLOC_FL_ZERO_RANGE
 197 flag (available since Linux 3.15)
 198 .\" commit 409332b65d3ed8cfa7a8030f1e9d52f372219642
 199 in
 200 .I mode
 201 zeros space in the byte range starting at
 202 .I offset
 203 and continuing for
 204 .I len
 205 bytes.
 206 Within the specified range, blocks are preallocated for the regions
 207 that span the holes in the file.
 208 After a successful call, subsequent
 209 reads from this range will return zeros.
 210 .PP
 211 Zeroing is done within the filesystem preferably by converting the range into
 212 unwritten extents.
 213 This approach means that the specified range will not be physically zeroed
 214 out on the device (except for partial blocks at the either end of the range),
 215 and I/O is (otherwise) required only to update metadata.
 216 .PP
 217 If the
 218 .B FALLOC_FL_KEEP_SIZE
 219 flag is additionally specified in
 220 .IR mode ,
 221 the behavior of the call is similar,
 222 but the file size will not be changed even if
 223 .IR offset + len
 224 is greater than the file size.
 225 This behavior is the same as when preallocating space with
 226 .B FALLOC_FL_KEEP_SIZE
 227 specified.
 228 .PP
 229 Not all filesystems support
 230 .BR FALLOC_FL_ZERO_RANGE ;
 231 if a filesystem doesn't support the operation, an error is returned.
 232 The operation is supported on at least the following filesystems:
 233 .IP * 3
 234 XFS (since Linux 3.15)
 235 .\" commit 376ba313147b4172f3e8cf620b9fb591f3e8cdfa
 236 .IP *
 237 ext4, for extent-based files (since Linux 3.15)
 238 .\" commit b8a8684502a0fc852afa0056c6bb2a9273f6fcc0
 239 .IP *
 240 SMB3 (since Linux 3.17)
 241 .\" commit 30175628bf7f521e9ee31ac98fa6d6fe7441a556
 242 .IP *
 243 Btrfs (since Linux 4.16)
 244 .\" commit f27451f229966874a8793995b8e6b74326d125df
 245 .SS Increasing file space
 246 Specifying the
 247 .BR FALLOC_FL_INSERT_RANGE
 248 flag
 249 (available since Linux 4.1)
 250 .\" commit dd46c787788d5bf5b974729d43e4c405814a4c7d
 251 in
 252 .I mode
 253 increases the file space by inserting a hole within the file size without
 254 overwriting any existing data.
 255 The hole will start at
 256 .I offset
 257 and continue for
 258 .I len
 259 bytes.
 260 When inserting the hole inside file, the contents of the file starting at
 261 .I offset
 262 will be shifted upward (i.e., to a higher file offset) by
 263 .I len
 264 bytes.
 265 Inserting a hole inside a file increases the file size by
 266 .I len
 267 bytes.
 268 .PP
 269 This mode has the same limitations as
 270 .BR FALLOC_FL_COLLAPSE_RANGE
 271 regarding the granularity of the operation.
 272 If the granularity requirements are not met,
 273 .BR fallocate ()
 274 fails with the error
 275 .BR EINVAL.
 276 If the
 277 .I offset
 278 is equal to or greater than the end of file, an error is returned.
 279 For such operations (i.e., inserting a hole at the end of file),
 280 .BR ftruncate (2)
 281 should be used.
 282 .PP
 283 No other flags may be specified in
 284 .IR mode
 285 in conjunction with
 286 .BR FALLOC_FL_INSERT_RANGE .
 287 .PP
 288 .B FALLOC_FL_INSERT_RANGE
 289 requires filesystem support.
 290 Filesystems that support this operation include
 291 XFS (since Linux 4.1)
 292 .\" commit a904b1ca5751faf5ece8600e18cd3b674afcca1b
 293 and ext4 (since Linux 4.2).
 294 .\" commit 331573febb6a224bc50322e3670da326cb7f4cfc
 295 .\" f2fs also has support since Linux 4.2
 296 .\"     commit f62185d0e283e9d311e3ac1020f159d95f0aab39
 297 .SH RETURN VALUE
 298 On success,
 299 .BR fallocate ()
 300 returns zero.
 301 On error, \-1 is returned and
 302 .I errno
 303 is set to indicate the error.
 304 .SH ERRORS
 305 .TP
 306 .B EBADF
 307 .I fd
 308 is not a valid file descriptor, or is not opened for writing.
 309 .TP
 310 .B EFBIG
 311 .IR offset + len
 312 exceeds the maximum file size.
 313 .TP
 314 .B EFBIG
 315 .I mode
 316 is
 317 .BR FALLOC_FL_INSERT_RANGE ,
 318 and the current file size+\fIlen\fP exceeds the maximum file size.
 319 .TP
 320 .B EINTR
 321 A signal was caught during execution; see
 322 .BR signal (7).
 323 .TP
 324 .B EINVAL
 325 .I offset
 326 was less than 0, or
 327 .I len
 328 .\" FIXME . (raise a kernel bug) Probably the len==0 case should be
 329 .\" a no-op, rather than an error. That would be consistent with
 330 .\" similar APIs for the len==0 case.
 331 .\" See "Re: [PATCH] fallocate.2: add FALLOC_FL_PUNCH_HOLE flag definition"
 332 .\" 21 Sep 2012
 333 .\" http://thread.gmane.org/gmane.linux.file-systems/48331/focus=1193526
 334 was less than or equal to 0.
 335 .TP
 336 .B EINVAL
 337 .I mode
 338 is
 339 .BR FALLOC_FL_COLLAPSE_RANGE
 340 and the range specified by
 341 .I offset
 342 plus
 343 .I len
 344 reaches or passes the end of the file.
 345 .TP
 346 .B EINVAL
 347 .I mode
 348 is
 349 .BR FALLOC_FL_INSERT_RANGE
 350 and the range specified by
 351 .I offset
 352 reaches or passes the end of the file.
 353 .TP
 354 .B EINVAL
 355 .I mode
 356 is
 357 .BR FALLOC_FL_COLLAPSE_RANGE
 358 or
 359 .BR FALLOC_FL_INSERT_RANGE ,
 360 but either
 361 .I offset
 362 or
 363 .I len
 364 is not a multiple of the filesystem block size.
 365 .TP
 366 .B EINVAL
 367 .I mode
 368 contains one of
 369 .B FALLOC_FL_COLLAPSE_RANGE
 370 or
 371 .B FALLOC_FL_INSERT_RANGE
 372 and also other flags;
 373 no other flags are permitted with
 374 .BR FALLOC_FL_COLLAPSE_RANGE
 375 or
 376 .BR FALLOC_FL_INSERT_RANGE .
 377 .TP
 378 .B EINVAL
 379 .I mode
 380 is
 381 .BR FALLOC_FL_COLLAPSE_RANGE
 382 or
 383 .BR FALLOC_FL_ZERO_RANGE
 384 or
 385 .BR FALLOC_FL_INSERT_RANGE ,
 386 but the file referred to by
 387 .I fd
 388 is not a regular file.
 389 .\" There was an inconsistency in 3.15-rc1, that should be resolved so that all
 390 .\" filesystems use this error for this case. (Tytso says ex4 will change.)
 391 .\" http://thread.gmane.org/gmane.comp.file-systems.xfs.general/60485/focus=5521
 392 .\" From: Michael Kerrisk (man-pages <mtk.manpages@...>
 393 .\" Subject: Re: [PATCH v5 10/10] manpage: update FALLOC_FL_COLLAPSE_RANGE flag in fallocate
 394 .\" Newsgroups: gmane.linux.man, gmane.linux.file-systems
 395 .\" Date: 2014-04-17 13:40:05 GMT
 396 .TP
 397 .B EIO
 398 An I/O error occurred while reading from or writing to a filesystem.
 399 .TP
 400 .B ENODEV
 401 .I fd
 402 does not refer to a regular file or a directory.
 403 (If
 404 .I fd
 405 is a pipe or FIFO, a different error results.)
 406 .TP
 407 .B ENOSPC
 408 There is not enough space left on the device containing the file
 409 referred to by
 410 .IR fd .
 411 .TP
 412 .B ENOSYS
 413 This kernel does not implement
 414 .BR fallocate ().
 415 .TP
 416 .B EOPNOTSUPP
 417 The filesystem containing the file referred to by
 418 .I fd
 419 does not support this operation;
 420 or the
 421 .I mode
 422 is not supported by the filesystem containing the file referred to by
 423 .IR fd .
 424 .TP
 425 .B EPERM
 426 The file referred to by
 427 .I fd
 428 is marked immutable (see
 429 .BR chattr (1)).
 430 .TP
 431 .B EPERM
 432 .I mode
 433 specifies
 434 .BR FALLOC_FL_PUNCH_HOLE
 435 or
 436 .BR FALLOC_FL_COLLAPSE_RANGE
 437 or
 438 .BR FALLOC_FL_INSERT_RANGE
 439 and
 440 the file referred to by
 441 .I fd
 442 is marked append-only
 443 (see
 444 .BR chattr (1)).
 445 .TP
 446 .B EPERM
 447 The operation was prevented by a file seal; see
 448 .BR fcntl (2).
 449 .TP
 450 .B ESPIPE
 451 .I fd
 452 refers to a pipe or FIFO.
 453 .TP
 454 .B ETXTBSY
 455 .I mode
 456 specifies
 457 .BR FALLOC_FL_COLLAPSE_RANGE
 458 or
 459 .BR FALLOC_FL_INSERT_RANGE ,
 460 but the file referred to by
 461 .IR fd
 462 is currently being executed.
 463 .SH VERSIONS
 464 .BR fallocate ()
 465 is available on Linux since kernel 2.6.23.
 466 Support is provided by glibc since version 2.10.
 467 The
 468 .BR FALLOC_FL_*
 469 flags are defined in glibc headers only since version 2.18.
 470 .\" See http://sourceware.org/bugzilla/show_bug.cgi?id=14964
 471 .SH CONFORMING TO
 472 .BR fallocate ()
 473 is Linux-specific.
 474 .SH SEE ALSO
 475 .BR fallocate (1),
 476 .BR ftruncate (2),
 477 .BR posix_fadvise (3),
 478 .BR posix_fallocate (3)