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 creates a grow-down stack area with the specified maximum 136stack size. This flag is no longer special-cased and will be converted 137to a normal anonymous mmap(), meaning that other mmap() calls cannot 138sub-map ungrown areas returned by prior MAP_STACK maps using TRYFIXED. 139The entire area is now applicable to the mapping. 140.Pp 141Note that the kernel itself can still create auto-grow areas but will 142do so for the user stack in order to maintain backwards compatibility 143with older code that might otherwise assume it can map below the user 144stack (in particular, older pthread libraries). This compatibility is 145deprecated and will be removed in a future release. 146.It Dv MAP_HASSEMAPHORE 147Notify the kernel that the region may contain semaphores and that special 148handling may be necessary. 149.It Dv MAP_NOCORE 150Region is not included in a core file. 151.It Dv MAP_NOSYNC 152Causes data dirtied via this VM map to be flushed to physical media 153only when necessary (usually by the pager) rather than gratuitously. 154Typically this prevents the update daemons from flushing pages dirtied 155through such maps and thus allows efficient sharing of memory across 156unassociated processes using a file-backed shared memory map. 157Without 158this option any VM pages you dirty may be flushed to disk every so often 159(every 30-60 seconds usually) which can create performance problems if you 160do not need that to occur (such as when you are using shared file-backed 161mmap regions for IPC purposes). 162Note that VM/filesystem coherency is maintained whether you use 163.Dv MAP_NOSYNC 164or not. 165This option is not portable across 166.Ux 167platforms (yet), though some may implement the same behavior 168by default. 169.Pp 170.Em WARNING ! 171Extending a file with 172.Xr ftruncate 2 , 173thus creating a big hole, and then filling the hole by modifying a shared 174.Fn mmap 175can lead to severe file fragmentation. 176In order to avoid such fragmentation you should always pre-allocate the 177file's backing store by 178.Fn write Ns ing 179zero's into the newly extended area prior to modifying the area via your 180.Fn mmap . 181The fragmentation problem is especially sensitive to 182.Dv MAP_NOSYNC 183pages, because pages may be flushed to disk in a totally random order. 184.Pp 185The same applies when using 186.Dv MAP_NOSYNC 187to implement a file-based shared memory store. 188It is recommended that you create the backing store by 189.Fn write Ns ing 190zero's to the backing file rather than 191.Fn ftruncate Ns ing 192it. 193You can test file fragmentation by observing the KB/t (kilobytes per 194transfer) results from an 195.Dq Li iostat 1 196while reading a large file sequentially, e.g.,\& using 197.Dq Li dd if=filename of=/dev/null bs=32k . 198.Pp 199The 200.Xr fsync 2 201function will flush all dirty data and metadata associated with a file, 202including dirty NOSYNC VM data, to physical media. 203The 204.Xr sync 8 205command and 206.Xr sync 2 207system call generally do not flush dirty NOSYNC VM data. 208The 209.Xr msync 2 210system call is obsolete since 211.Bx 212implements a coherent filesystem buffer cache. 213However, it may be 214used to associate dirty VM pages with filesystem buffers and thus cause 215them to be flushed to physical media sooner rather than later. 216.It Dv MAP_PRIVATE 217Modifications are private. 218.It Dv MAP_SHARED 219Modifications are shared. 220.It Dv MAP_STACK 221Map the area as a stack. 222.Dv MAP_ANON 223is implied. 224.Fa Offset 225should be 0, 226.Fa fd 227must be -1, and 228.Fa prot 229should include at least 230.Dv PROT_READ 231and 232.Dv PROT_WRITE . 233This option creates 234a memory region that grows to at most 235.Fa len 236bytes in size, starting from the stack top and growing down. 237The stack top is the starting address returned by the call, plus 238.Fa len 239bytes. 240The bottom of the stack at maximum growth is the starting 241address returned by the call. 242.Pp 243The entire area is reserved from the point of view of other 244.Fn mmap 245calls, even if not faulted in yet. 246.Pp 247Note that unless 248.Dv MAP_FIXED 249or 250.Dv MAP_TRYFIXED 251is used, you cannot count on the returned address matching the hint 252you have provided. 253.It Dv MAP_VPAGETABLE 254Memory accessed via this map is not linearly mapped and will be governed 255by a virtual page table. 256The base address of the virtual page table may be set using 257.Xr mcontrol 2 258with 259.Dv MADV_SETMAP . 260Virtual page tables work with anonymous memory but there 261is no way to populate the page table so for all intents and purposes 262.Dv MAP_VPAGETABLE 263can only be used when mapping file descriptors. 264Since the kernel will update the 265.Dv VPTE_M 266bit in the virtual page table, the mapping must R+W 267even though actual access to the memory will be properly governed by 268the virtual page table. 269.Pp 270Addressable backing store is limited by the range supported in the virtual 271page table entries. 272The kernel may implement a page table abstraction capable 273of addressing a larger range within the backing store then could otherwise 274be mapped into memory. 275.El 276.Pp 277The 278.Xr close 2 279function does not unmap pages, see 280.Xr munmap 2 281for further information. 282.Pp 283The current design does not allow a process to specify the location of 284swap space. 285In the future we may define an additional mapping type, 286.Dv MAP_SWAP , 287in which 288the file descriptor argument specifies a file or device to which swapping 289should be done. 290.Sh RETURN VALUES 291Upon successful completion, 292.Fn mmap 293returns a pointer to the mapped region. 294Otherwise, a value of 295.Dv MAP_FAILED 296is returned and 297.Va errno 298is set to indicate the error. 299.Sh ERRORS 300.Fn Mmap 301will fail if: 302.Bl -tag -width Er 303.It Bq Er EACCES 304The flag 305.Dv PROT_READ 306was specified as part of the 307.Fa prot 308parameter and 309.Fa fd 310was not open for reading. 311The flags 312.Dv MAP_SHARED 313and 314.Dv PROT_WRITE 315were specified as part of the 316.Fa flags 317and 318.Fa prot 319parameters and 320.Fa fd 321was not open for writing. 322.It Bq Er EBADF 323.Fa fd 324is not a valid open file descriptor. 325.It Bq Er EINVAL 326.Dv MAP_FIXED 327was specified and the 328.Fa addr 329parameter was not page aligned, or part of the desired address space 330resides out of the valid address space for a user process. 331.It Bq Er EINVAL 332.Fa Len 333was negative. 334.It Bq Er EINVAL 335.Dv MAP_ANON 336was specified and the 337.Fa fd 338parameter was not -1. 339.It Bq Er EINVAL 340.Dv MAP_ANON 341has not been specified and 342.Fa fd 343did not reference a regular or character special file. 344.It Bq Er EINVAL 345.Fa Offset 346was not page-aligned. 347.It Bq Er ENOMEM 348.Dv MAP_FIXED 349was specified and the 350.Fa addr 351parameter wasn't available. 352.Dv MAP_ANON 353was specified and insufficient memory was available. 354The system has reached the per-process mmap limit specified in the 355.Va vm.max_proc_mmap 356sysctl. 357.El 358.Sh SEE ALSO 359.Xr madvise 2 , 360.Xr mincore 2 , 361.Xr mlock 2 , 362.Xr mprotect 2 , 363.Xr msync 2 , 364.Xr munlock 2 , 365.Xr munmap 2 , 366.Xr getpagesize 3 367