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 .\" 4. 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 pagesize, the mapped region may extend past the
58 Any such extension beyond the end of the mapped object will be zero-filled.
62 is non-zero, it is used as a hint to the system.
63 (As a convenience to the system, the actual address of the region may differ
64 from the address supplied.)
67 is zero, an address will be selected by the system.
68 The actual starting address of the region is returned.
71 deletes any previous mapping in the allocated address range.
73 The protections (region accessibility) are specified in the
79 .Bl -tag -width PROT_WRITE -compact
81 Pages may not be accessed.
87 Pages may be executed.
92 argument specifies the type of the mapped object, mapping options and
93 whether modifications made to the mapped copy of the page are private
94 to the process or are to be shared with other references.
95 Sharing, mapping type and options are specified in the
100 .Bl -tag -width MAP_PREFAULT_READ
101 .It Dv MAP_ALIGNED Ns Pq Fa n
102 Align the region on a requested boundary.
103 If a suitable region cannot be found,
108 argument specifies the binary logarithm of the desired alignment.
109 .It Dv MAP_ALIGNED_SUPER
110 Align the region to maximize the potential use of large
113 If a suitable region cannot be found,
116 The system will choose a suitable page size based on the size of
118 The page size used as well as the alignment of the region may both be
119 affected by properties of the file being mapped.
121 the physical address of existing pages of a file may require a specific
123 The region is not guaranteed to be aligned on any specific boundary.
125 Map anonymous memory not associated with any specific file.
126 The file descriptor used for creating
133 .\"Mapped from a regular file or character-special device memory.
135 This flag is identical to
137 and is provided for compatibility.
139 Do not permit the system to select a different address than the one
141 If the specified address cannot be used,
148 must be a multiple of the pagesize.
151 request is successful, the mapping established by
153 replaces any previous mappings for the process' pages in the range from
159 Use of this option is discouraged.
160 .It Dv MAP_HASSEMAPHORE
161 Notify the kernel that the region may contain semaphores and that special
162 handling may be necessary.
164 This flag never operated as advertised and is no longer supported.
167 for further information.
169 Region is not included in a core file.
171 Causes data dirtied via this VM map to be flushed to physical media
172 only when necessary (usually by the pager) rather than gratuitously.
173 Typically this prevents the update daemons from flushing pages dirtied
174 through such maps and thus allows efficient sharing of memory across
175 unassociated processes using a file-backed shared memory map.
177 this option any VM pages you dirty may be flushed to disk every so often
178 (every 30-60 seconds usually) which can create performance problems if you
179 do not need that to occur (such as when you are using shared file-backed
180 mmap regions for IPC purposes).
181 Note that VM/file system coherency is
182 maintained whether you use
185 This option is not portable
188 platforms (yet), though some may implement the same behavior
192 Extending a file with
194 thus creating a big hole, and then filling the hole by modifying a shared
196 can lead to severe file fragmentation.
197 In order to avoid such fragmentation you should always pre-allocate the
198 file's backing store by
200 zero's into the newly extended area prior to modifying the area via your
202 The fragmentation problem is especially sensitive to
204 pages, because pages may be flushed to disk in a totally random order.
206 The same applies when using
208 to implement a file-based shared memory store.
209 It is recommended that you create the backing store by
211 zero's to the backing file rather than
214 You can test file fragmentation by observing the KB/t (kilobytes per
215 transfer) results from an
217 while reading a large file sequentially, e.g.\& using
218 .Dq Li dd if=filename of=/dev/null bs=32k .
222 system call will flush all dirty data and metadata associated with a file,
223 including dirty NOSYNC VM data, to physical media.
228 system call generally do not flush dirty NOSYNC VM data.
231 system call is usually not needed since
233 implements a coherent file system buffer cache.
235 used to associate dirty VM pages with file system buffers and thus cause
236 them to be flushed to physical media sooner rather than later.
237 .It Dv MAP_PREFAULT_READ
238 Immediately update the calling process's lowest-level virtual address
239 translation structures, such as its page table, so that every memory
240 resident page within the region is mapped for read access.
241 Ordinarily these structures are updated lazily.
242 The effect of this option is to eliminate any soft faults that would
243 otherwise occur on the initial read accesses to the region.
244 Although this option does not preclude
248 it does not eliminate soft faults on the initial write accesses to the
251 Modifications are private.
253 Modifications are shared.
266 must include at least
271 a memory region that grows to at most
273 bytes in size, starting from the stack top and growing down.
275 stack top is the starting address returned by the call, plus
278 The bottom of the stack at maximum growth is the starting
279 address returned by the call.
284 system call does not unmap pages, see
286 for further information.
288 The current design does not allow a process to specify the location of
290 In the future we may define an additional mapping type,
293 the file descriptor argument specifies a file or device to which swapping
296 Although this implementation does not impose any alignment restrictions on
299 argument, a portable program must only use page-aligned values.
301 Large page mappings require that the pages backing an object be
302 aligned in matching blocks in both the virtual address space and RAM.
303 The system will automatically attempt to use large page mappings when
304 mapping an object that is already backed by large pages in RAM by
305 aligning the mapping request in the virtual address space to match the
306 alignment of the large physical pages.
307 The system may also use large page mappings when mapping portions of an
308 object that are not yet backed by pages in RAM.
310 .Dv MAP_ALIGNED_SUPER
311 flag is an optimization that will align the mapping request to the
312 size of a large page similar to
314 except that the system will override this alignment if an object already
315 uses large pages so that the mapping will be consistent with the existing
317 This flag is mostly useful for maximizing the use of large pages on the
318 first mapping of objects that do not yet have pages present in RAM.
320 Upon successful completion,
322 returns a pointer to the mapped region.
323 Otherwise, a value of
327 is set to indicate the error.
337 was specified as part of the
341 was not open for reading.
346 were specified as part of the
352 was not open for writing.
357 is not a valid open file descriptor.
360 was specified and the
362 argument was not page aligned, or part of the desired address space
363 resides out of the valid address space for a user process.
371 was specified and the desired alignment was either larger than the
372 virtual address size of the machine or smaller than a page.
375 was specified and the
380 was specified and the
385 has not been specified and
387 did not reference a regular or character special file.
390 was specified and the
392 argument was not available.
394 was specified and insufficient memory was available.
411 is limited to the maximum file size or available userland address
413 Files may not be able to be made more than 1TB large on 32 bit systems
414 due to file systems restrictions and bugs, but address space is far more
416 Larger files may be possible on 64 bit systems.
418 The previous documented limit of 2GB was a documentation bug.
419 That limit has not existed since