1 .\" Copyright (c) 1991, 1993
2 .\" The Regents of the University of California. All rights reserved.
4 .\" Redistribution and use in source and binary forms, with or without
5 .\" modification, are permitted provided that the following conditions
7 .\" 1. Redistributions of source code must retain the above copyright
8 .\" notice, this list of conditions and the following disclaimer.
9 .\" 2. Redistributions in binary form must reproduce the above copyright
10 .\" notice, this list of conditions and the following disclaimer in the
11 .\" documentation and/or other materials provided with the distribution.
12 .\" 3. Neither the name of the University nor the names of its contributors
13 .\" may be used to endorse or promote products derived from this software
14 .\" without specific prior written permission.
16 .\" THIS SOFTWARE IS PROVIDED BY THE REGENTS AND CONTRIBUTORS ``AS IS'' AND
17 .\" ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE
18 .\" IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE
19 .\" ARE DISCLAIMED. IN NO EVENT SHALL THE REGENTS OR CONTRIBUTORS BE LIABLE
20 .\" FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL
21 .\" DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS
22 .\" OR SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION)
23 .\" HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT
24 .\" LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY
25 .\" OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF
28 .\" @(#)mmap.2 8.4 (Berkeley) 5/11/95
36 .Nd allocate memory, or map files or devices into memory
42 .Fn mmap "void *addr" "size_t len" "int prot" "int flags" "int fd" "off_t offset"
46 system call causes the pages starting at
48 and continuing for at most
50 bytes to be mapped from the object described by
52 starting at byte offset
56 is not a multiple of the page size, the mapped region may extend past the
58 Any such extension beyond the end of the mapped object will be zero-filled.
62 references a regular file or a shared memory object, the range of
67 bytes must be legitimate for the possible (not necessarily
68 current) offsets in the object.
71 value cannot be negative.
72 If the object is truncated and the process later accesses a page that
73 is wholly within the truncated region, the access is aborted and a
75 signal is delivered to the process.
79 references a device file, the interpretation of the
81 value is device specific and defined by the device driver.
82 The virtual memory subsystem does not impose any restrictions on the
84 value in this case, passing it unchanged to the driver.
88 is non-zero, it is used as a hint to the system.
89 (As a convenience to the system, the actual address of the region may differ
90 from the address supplied.)
93 is zero, an address will be selected by the system.
94 The actual starting address of the region is returned.
97 deletes any previous mapping in the allocated address range.
99 The protections (region accessibility) are specified in the
103 the following values:
105 .Bl -tag -width PROT_WRITE -compact
107 Pages may not be accessed.
111 Pages may be written.
113 Pages may be executed.
116 In addition to these protection flags,
118 provides the ability to set the maximum protection of a region allocated by
122 This is accomplished by
126 values wrapped in the
134 argument specifies the type of the mapped object, mapping options and
135 whether modifications made to the mapped copy of the page are private
136 to the process or are to be shared with other references.
137 Sharing, mapping type and options are specified in the
141 the following values:
142 .Bl -tag -width MAP_PREFAULT_READ
144 Request a region in the first 2GB of the current process's address space.
145 If a suitable region cannot be found,
148 This flag is only available on 64-bit platforms.
149 .It Dv MAP_ALIGNED Ns Pq Fa n
150 Align the region on a requested boundary.
151 If a suitable region cannot be found,
156 argument specifies the binary logarithm of the desired alignment.
157 .It Dv MAP_ALIGNED_SUPER
158 Align the region to maximize the potential use of large
161 If a suitable region cannot be found,
164 The system will choose a suitable page size based on the size of
166 The page size used as well as the alignment of the region may both be
167 affected by properties of the file being mapped.
169 the physical address of existing pages of a file may require a specific
171 The region is not guaranteed to be aligned on any specific boundary.
173 Map anonymous memory not associated with any specific file.
174 The file descriptor used for creating
181 .\"Mapped from a regular file or character-special device memory.
183 This flag is identical to
185 and is provided for compatibility.
187 This flag can only be used in combination with
189 Please see the definition of
191 for the description of its effect.
193 Do not permit the system to select a different address than the one
195 If the specified address cannot be used,
202 must be a multiple of the page size.
205 is not specified, a successful
207 request replaces any previous mappings for the process'
208 pages in the range from
216 is specified, the request will fail if a mapping
217 already exists within the range.
219 Instead of a mapping, create a guard of the specified size.
220 Guards allow a process to create reservations in its address space,
221 which can later be replaced by actual mappings.
224 will not create mappings in the address range of a guard unless
225 the request specifies
227 Guards can be destroyed with
229 Any memory access by a thread to the guarded range results
232 signal to that thread.
234 Region is not included in a core file.
236 Causes data dirtied via this VM map to be flushed to physical media
237 only when necessary (usually by the pager) rather than gratuitously.
238 Typically this prevents the update daemons from flushing pages dirtied
239 through such maps and thus allows efficient sharing of memory across
240 unassociated processes using a file-backed shared memory map.
242 this option any VM pages you dirty may be flushed to disk every so often
243 (every 30-60 seconds usually) which can create performance problems if you
244 do not need that to occur (such as when you are using shared file-backed
245 mmap regions for IPC purposes).
246 Dirty data will be flushed automatically when all mappings of an object are
247 removed and all descriptors referencing the object are closed.
248 Note that VM/file system coherency is
249 maintained whether you use
252 This option is not portable
255 platforms (yet), though some may implement the same behavior
259 Extending a file with
261 thus creating a big hole, and then filling the hole by modifying a shared
263 can lead to severe file fragmentation.
264 In order to avoid such fragmentation you should always pre-allocate the
265 file's backing store by
267 zero's into the newly extended area prior to modifying the area via your
269 The fragmentation problem is especially sensitive to
271 pages, because pages may be flushed to disk in a totally random order.
273 The same applies when using
275 to implement a file-based shared memory store.
276 It is recommended that you create the backing store by
278 zero's to the backing file rather than
281 You can test file fragmentation by observing the KB/t (kilobytes per
282 transfer) results from an
284 while reading a large file sequentially, e.g.,\& using
285 .Dq Li dd if=filename of=/dev/null bs=32k .
289 system call will flush all dirty data and metadata associated with a file,
290 including dirty NOSYNC VM data, to physical media.
295 system call generally do not flush dirty NOSYNC VM data.
298 system call is usually not needed since
300 implements a coherent file system buffer cache.
302 used to associate dirty VM pages with file system buffers and thus cause
303 them to be flushed to physical media sooner rather than later.
304 .It Dv MAP_PREFAULT_READ
305 Immediately update the calling process's lowest-level virtual address
306 translation structures, such as its page table, so that every memory
307 resident page within the region is mapped for read access.
308 Ordinarily these structures are updated lazily.
309 The effect of this option is to eliminate any soft faults that would
310 otherwise occur on the initial read accesses to the region.
311 Although this option does not preclude
315 it does not eliminate soft faults on the initial write accesses to the
318 Modifications are private.
320 Modifications are shared.
322 Creates both a mapped region that grows downward on demand and an
323 adjoining guard that both reserves address space for the mapped region
324 to grow into and limits the mapped region's growth.
325 Together, the mapped region and the guard occupy
327 bytes of the address space.
328 The guard starts at the returned address, and the mapped region ends at
329 the returned address plus
332 Upon access to the guard, the mapped region automatically grows in size,
333 and the guard shrinks by an equal amount.
334 Essentially, the boundary between the guard and the mapped region moves
335 downward so that the access falls within the enlarged mapped region.
336 However, the guard will never shrink to less than the number of pages
337 specified by the sysctl
338 .Dv security.bsd.stack_guard_page ,
339 thereby ensuring that a gap for detecting stack overflow always exists
340 between the downward growing mapped region and the closest mapped region
354 must include at least
358 The size of the guard, in pages, is specified by sysctl
359 .Dv security.bsd.stack_guard_page .
364 system call does not unmap pages, see
366 for further information.
368 Although this implementation does not impose any alignment restrictions on
371 argument, a portable program must only use page-aligned values.
373 Large page mappings require that the pages backing an object be
374 aligned in matching blocks in both the virtual address space and RAM.
375 The system will automatically attempt to use large page mappings when
376 mapping an object that is already backed by large pages in RAM by
377 aligning the mapping request in the virtual address space to match the
378 alignment of the large physical pages.
379 The system may also use large page mappings when mapping portions of an
380 object that are not yet backed by pages in RAM.
382 .Dv MAP_ALIGNED_SUPER
383 flag is an optimization that will align the mapping request to the
384 size of a large page similar to
386 except that the system will override this alignment if an object already
387 uses large pages so that the mapping will be consistent with the existing
389 This flag is mostly useful for maximizing the use of large pages on the
390 first mapping of objects that do not yet have pages present in RAM.
392 Upon successful completion,
394 returns a pointer to the mapped region.
395 Otherwise, a value of
399 is set to indicate the error.
409 was specified as part of the
413 was not open for reading.
418 were specified as part of the
424 was not open for writing.
429 is not a valid open file descriptor.
431 An invalid (negative) value was passed in the
435 referenced a regular file or shared memory.
437 An invalid value was passed in the
441 An undefined option was set in the
459 At least one of these flags must be included.
464 is less than or equal to the guard size.
467 was specified and the
469 argument was not page aligned, or part of the desired address space
470 resides out of the valid address space for a user process.
476 were specified and part of the desired address space resides outside
477 of the first 2GB of user address space.
485 was specified and the desired alignment was either larger than the
486 virtual address size of the machine or smaller than a page.
489 was specified and the
494 was specified and the
502 were specified, but the requested region is already used by a mapping.
510 was specified, but the
512 argument was not zero, the
514 argument was not -1, or the
520 was specified together with one of the flags
523 .Dv MAP_PREFAULT_READ ,
529 has not been specified and
531 did not reference a regular or character special file.
534 was specified and the
536 argument was not available.
538 was specified and insufficient memory was available.
542 argument contains protections which are not a subset of the specified
559 system call was first documented in
563 .\" XXX: lots of missing history of FreeBSD additions.
567 functionality was introduced in