1.\" Copyright (c) 1991, 1993 2.\" The Regents of the University of California. All rights reserved. 3.\" 4.\" Redistribution and use in source and binary forms, with or without 5.\" modification, are permitted provided that the following conditions 6.\" are met: 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. 15.\" 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 26.\" SUCH DAMAGE. 27.\" 28.\" @(#)mmap.2 8.4 (Berkeley) 5/11/95 29.\" $FreeBSD: src/lib/libc/sys/mmap.2,v 1.22.2.12 2002/02/27 03:40:13 dd Exp $ 30.\" 31.Dd January 18, 2015 32.Dt MMAP 2 33.Os 34.Sh NAME 35.Nm mmap 36.Nd allocate memory, or map files or devices into memory 37.Sh LIBRARY 38.Lb libc 39.Sh SYNOPSIS 40.In sys/types.h 41.In sys/mman.h 42.Ft void * 43.Fn mmap "void *addr" "size_t len" "int prot" "int flags" "int fd" "off_t offset" 44.Sh DESCRIPTION 45The 46.Fn mmap 47function causes the pages starting at 48.Fa addr 49and continuing for at most 50.Fa len 51bytes to be mapped from the object described by 52.Fa fd , 53starting at byte offset 54.Fa offset . 55If 56.Fa len 57is not a multiple of the pagesize, the mapped region may extend past the 58specified range. 59Any such extension beyond the end of the mapped object will be zero-filled. 60.Pp 61If 62.Fa addr 63is non-zero, it is used as a hint to the system. 64(As a convenience to the system, the actual address of the region may differ 65from the address supplied.) 66If 67.Fa addr 68is zero, an address will be selected by the system. 69The actual starting address of the region is returned. 70A successful 71.Fa mmap 72deletes any previous mapping in the allocated address range. 73.Pp 74The protections (region accessibility) are specified in the 75.Fa prot 76argument by 77.Em or Ns 'ing 78the following values: 79.Pp 80.Bl -tag -width ".Dv PROT_WRITE" -compact 81.It Dv PROT_NONE 82Pages may not be accessed. 83.It Dv PROT_READ 84Pages may be read. 85.It Dv PROT_WRITE 86Pages may be written. 87.It Dv PROT_EXEC 88Pages may be executed. 89.El 90.Pp 91The 92.Fa flags 93parameter specifies the type of the mapped object, mapping options and 94whether modifications made to the mapped copy of the page are private 95to the process or are to be shared with other references. 96Sharing, mapping type and options are specified in the 97.Fa flags 98argument by 99.Em or Ns 'ing 100the following values: 101.Bl -tag -width ".Dv MAP_HASSEMAPHORE" 102.It Dv MAP_ANON 103Map anonymous memory not associated with any specific file. 104The file descriptor used for creating 105.Dv MAP_ANON 106must be \-1. 107The 108.Fa offset 109parameter is ignored. 110.It Dv MAP_ANONYMOUS 111This flag is an alias for 112.Dv MAP_ANON 113and is provided for compatibility. 114.\".It Dv MAP_FILE 115.\"Mapped from a regular file or character-special device memory. 116.It Dv MAP_FIXED 117Do not permit the system to select a different address than the one 118specified. 119If the specified address contains other mappings those mappings will 120be replaced. 121If the specified address cannot otherwise be used, 122.Fn mmap 123will fail. 124If 125.Dv MAP_FIXED 126is specified, 127.Fa addr 128must be a multiple of the pagesize. 129.It Dv MAP_TRYFIXED 130Try to do a fixed mapping but fail if another mapping already exists in 131the space instead of overwriting the mapping. 132.Pp 133When used with 134.Dv MAP_STACK 135this flag allows one 136.Dv MAP_STACK 137mapping to be 138made within another (typically the master user stack), as long as 139no pages have been faulted in the area requested. 140.It Dv MAP_HASSEMAPHORE 141Notify the kernel that the region may contain semaphores and that special 142handling may be necessary. 143.It Dv MAP_NOCORE 144Region is not included in a core file. 145.It Dv MAP_NOSYNC 146Causes data dirtied via this VM map to be flushed to physical media 147only when necessary (usually by the pager) rather than gratuitously. 148Typically this prevents the update daemons from flushing pages dirtied 149through such maps and thus allows efficient sharing of memory across 150unassociated processes using a file-backed shared memory map. 151Without 152this option any VM pages you dirty may be flushed to disk every so often 153(every 30-60 seconds usually) which can create performance problems if you 154do not need that to occur (such as when you are using shared file-backed 155mmap regions for IPC purposes). 156Note that VM/filesystem coherency is maintained whether you use 157.Dv MAP_NOSYNC 158or not. 159This option is not portable across 160.Ux 161platforms (yet), though some may implement the same behavior 162by default. 163.Pp 164.Em WARNING ! 165Extending a file with 166.Xr ftruncate 2 , 167thus creating a big hole, and then filling the hole by modifying a shared 168.Fn mmap 169can lead to severe file fragmentation. 170In order to avoid such fragmentation you should always pre-allocate the 171file's backing store by 172.Fn write Ns ing 173zero's into the newly extended area prior to modifying the area via your 174.Fn mmap . 175The fragmentation problem is especially sensitive to 176.Dv MAP_NOSYNC 177pages, because pages may be flushed to disk in a totally random order. 178.Pp 179The same applies when using 180.Dv MAP_NOSYNC 181to implement a file-based shared memory store. 182It is recommended that you create the backing store by 183.Fn write Ns ing 184zero's to the backing file rather than 185.Fn ftruncate Ns ing 186it. 187You can test file fragmentation by observing the KB/t (kilobytes per 188transfer) results from an 189.Dq Li iostat 1 190while reading a large file sequentially, e.g.,\& using 191.Dq Li dd if=filename of=/dev/null bs=32k . 192.Pp 193The 194.Xr fsync 2 195function will flush all dirty data and metadata associated with a file, 196including dirty NOSYNC VM data, to physical media. 197The 198.Xr sync 8 199command and 200.Xr sync 2 201system call generally do not flush dirty NOSYNC VM data. 202The 203.Xr msync 2 204system call is obsolete since 205.Bx 206implements a coherent filesystem buffer cache. 207However, it may be 208used to associate dirty VM pages with filesystem buffers and thus cause 209them to be flushed to physical media sooner rather than later. 210.It Dv MAP_PRIVATE 211Modifications are private. 212.It Dv MAP_SHARED 213Modifications are shared. 214.It Dv MAP_STACK 215Map the area as a stack. 216.Dv MAP_ANON 217is implied. 218.Fa Offset 219should be 0, 220.Fa fd 221must be -1, and 222.Fa prot 223should include at least 224.Dv PROT_READ 225and 226.Dv PROT_WRITE . 227This option creates 228a memory region that grows to at most 229.Fa len 230bytes in size, starting from the stack top and growing down. 231The stack top is the starting address returned by the call, plus 232.Fa len 233bytes. 234The bottom of the stack at maximum growth is the starting 235address returned by the call. 236.Pp 237The entire area is reserved from the point of view of other 238.Fn mmap 239calls, even if not faulted in yet. 240.Pp 241.Em WARNING ! 242We currently allow 243.Dv MAP_STACK 244mappings to provide a hint that points within an existing 245.Dv MAP_STACK 246mapping's space, and this will succeed as long as no page have been 247faulted in the area specified, but this behavior is no longer supported 248unless you also specify the 249.Dv MAP_TRYFIXED 250flag. 251.Pp 252Note that unless 253.Dv MAP_FIXED 254or 255.Dv MAP_TRYFIXED 256is used, you cannot count on the returned address matching the hint 257you have provided. 258.It Dv MAP_VPAGETABLE 259Memory accessed via this map is not linearly mapped and will be governed 260by a virtual page table. 261The base address of the virtual page table may be set using 262.Xr mcontrol 2 263with 264.Dv MADV_SETMAP . 265Virtual page tables work with anonymous memory but there 266is no way to populate the page table so for all intents and purposes 267.Dv MAP_VPAGETABLE 268can only be used when mapping file descriptors. 269Since the kernel will update the 270.Dv VPTE_M 271bit in the virtual page table, the mapping must R+W 272even though actual access to the memory will be properly governed by 273the virtual page table. 274.Pp 275Addressable backing store is limited by the range supported in the virtual 276page table entries. 277The kernel may implement a page table abstraction capable 278of addressing a larger range within the backing store then could otherwise 279be mapped into memory. 280.El 281.Pp 282The 283.Xr close 2 284function does not unmap pages, see 285.Xr munmap 2 286for further information. 287.Pp 288The current design does not allow a process to specify the location of 289swap space. 290In the future we may define an additional mapping type, 291.Dv MAP_SWAP , 292in which 293the file descriptor argument specifies a file or device to which swapping 294should be done. 295.Sh RETURN VALUES 296Upon successful completion, 297.Fn mmap 298returns a pointer to the mapped region. 299Otherwise, a value of 300.Dv MAP_FAILED 301is returned and 302.Va errno 303is set to indicate the error. 304.Sh ERRORS 305.Fn Mmap 306will fail if: 307.Bl -tag -width Er 308.It Bq Er EACCES 309The flag 310.Dv PROT_READ 311was specified as part of the 312.Fa prot 313parameter and 314.Fa fd 315was not open for reading. 316The flags 317.Dv MAP_SHARED 318and 319.Dv PROT_WRITE 320were specified as part of the 321.Fa flags 322and 323.Fa prot 324parameters and 325.Fa fd 326was not open for writing. 327.It Bq Er EBADF 328.Fa fd 329is not a valid open file descriptor. 330.It Bq Er EINVAL 331.Dv MAP_FIXED 332was specified and the 333.Fa addr 334parameter was not page aligned, or part of the desired address space 335resides out of the valid address space for a user process. 336.It Bq Er EINVAL 337.Fa Len 338was negative. 339.It Bq Er EINVAL 340.Dv MAP_ANON 341was specified and the 342.Fa fd 343parameter was not -1. 344.It Bq Er EINVAL 345.Dv MAP_ANON 346has not been specified and 347.Fa fd 348did not reference a regular or character special file. 349.It Bq Er EINVAL 350.Fa Offset 351was not page-aligned. 352(See 353.Sx BUGS 354below.) 355.It Bq Er ENOMEM 356.Dv MAP_FIXED 357was specified and the 358.Fa addr 359parameter wasn't available. 360.Dv MAP_ANON 361was specified and insufficient memory was available. 362The system has reached the per-process mmap limit specified in the 363.Va vm.max_proc_mmap 364sysctl. 365.El 366.Sh SEE ALSO 367.Xr madvise 2 , 368.Xr mincore 2 , 369.Xr mlock 2 , 370.Xr mprotect 2 , 371.Xr msync 2 , 372.Xr munlock 2 , 373.Xr munmap 2 , 374.Xr getpagesize 3 375