Manpage of MMAP


Section: Linux Programmer's Manual (2)
Updated: 2016-12-12


mmap, munmap - map or unmap files or devices into memory  


#include <sys/mman.h>void *mmap(void *addr, size_t length, int prot, int flags,           int fd, off_t offset);int munmap(void *addr, size_t length);

See NOTES for information on feature test macro requirements.  


mmap() creates a new mapping in the virtual address space of the calling process. The starting address for the new mapping is specified in addr. The lengthargument specifies the length of the mapping.

If addris NULL, then the kernel chooses the address at which to create the mapping; this is the most portable method of creating a new mapping. If addris not NULL, then the kernel takes it as a hint about where to place the mapping; on Linux, the mapping will be created at a nearby page boundary. The address of the new mapping is returned as the result of the call.

The contents of a file mapping (as opposed to an anonymous mapping; see MAP_ANONYMOUSbelow), are initialized using lengthbytes starting at offset offsetin the file (or other object) referred to by the file descriptor fd. offsetmust be a multiple of the page size as returned by sysconf(_SC_PAGE_SIZE).

The protargument describes the desired memory protection of the mapping (and must not conflict with the open mode of the file). It is either PROT_NONEor the bitwise OR of one or more of the following flags:

Pages may be executed.
Pages may be read.
Pages may be written.
Pages may not be accessed.

The flagsargument determines whether updates to the mapping are visible to other processes mapping the same region, and whether updates are carried through to the underlying file. This behavior is determined by including exactly one of the following values in flags:

Share this mapping. Updates to the mapping are visible to other processes mapping the same region, and (in the case of file-backed mappings) are carried through to the underlying file. (To precisely control when updates are carried through to the underlying file requires the use of msync(2).)
Create a private copy-on-write mapping. Updates to the mapping are not visible to other processes mapping the same file, and are not carried through to the underlying file. It is unspecified whether changes made to the file after the mmap() call are visible in the mapped region.

Both of these flags are described in POSIX.1-2001 and POSIX.1-2008.

In addition, zero or more of the following values can be ORed in flags:

MAP_32BIT (since Linux 2.4.20, 2.6)
Put the mapping into the first 2 Gigabytes of the process address space. This flag is supported only on x86-64, for 64-bit programs. It was added to allow thread stacks to be allocated somewhere in the first 2GB of memory, so as to improve context-switch performance on some early 64-bit processors. Modern x86-64 processors no longer have this performance problem, so use of this flag is not required on those systems. The MAP_32BITflag is ignored when MAP_FIXEDis set.
Synonym for MAP_ANONYMOUS. Deprecated.
The mapping is not backed by any file; its contents are initialized to zero. The fdargument is ignored; however, some implementations require fdto be -1 if MAP_ANONYMOUS(or MAP_ANON) is specified, and portable applications should ensure this. The offsetargument should be zero. The use of MAP_ANONYMOUSin conjunction with MAP_SHAREDis supported on Linux only since kernel 2.4.
This flag is ignored. (Long ago, it signaled that attempts to write to the underlying file should fail with ETXTBUSY. But this was a source of denial-of-service attacks.)
This flag is ignored.
Compatibility flag. Ignored.
Don't interpret addras a hint: place the mapping at exactly that address. addrmust be a multiple of the page size. If the memory region specified by addrand lenoverlaps pages of any existing mapping(s), then the overlapped part of the existing mapping(s) will be discarded. If the specified address cannot be used, mmap() will fail. Because requiring a fixed address for a mapping is less portable, the use of this option is discouraged.
This flag is used for stacks. It indicates to the kernel virtual memory system that the mapping should extend downward in memory. The return address is one page lower than the memory area that is actually created in the process's virtual address space. Touching an address in the "guard" page below the mapping will cause the mapping to grow by a page. This growth can be repeated until the mapping grows to within a page of the high end of the next lower mapping, at which point touching the "guard" page will result in a SIGSEGVsignal.
MAP_HUGETLB (since Linux 2.6.32)
Allocate the mapping using "huge pages." See the Linux kernel source file Documentation/vm/hugetlbpage.txtfor further information, as well as NOTES, below.
MAP_HUGE_2MB, MAP_HUGE_1GB (since Linux 3.8)
Used in conjunction with MAP_HUGETLBto select alternative hugetlb page sizes (respectively, 2 MB and 1 GB) on systems that support multiple hugetlb page sizes.

More generally, the desired huge page size can be configured by encoding the base-2 logarithm of the desired page size in the six bits at the offset MAP_HUGE_SHIFT. (A value of zero in this bit field provides the default huge page size; the default huge page size can be discovered vie the Hugepagesizefield exposed by /proc/meminfo.) Thus, the above two constants are defined as:

#define MAP_HUGE_2MB    (21 << MAP_HUGE_SHIFT)
#define MAP_HUGE_1GB    (30 << MAP_HUGE_SHIFT)

The range of huge page sizes that are supported by the system can be discovered by listing the subdirectories in /sys/kernel/mm/hugepages.

MAP_LOCKED (since Linux 2.5.37)
Mark the mmaped region to be locked in the same way as mlock(2). This implementation will try to populate (prefault) the whole range but the mmap call doesn't fail with ENOMEMif this fails. Therefore major faults might happen later on. So the semantic is not as strong as mlock(2). One should use mmap() plus mlock(2) when major faults are not acceptable after the initialization of the mapping. The MAP_LOCKEDflag is ignored in older kernels.
MAP_NONBLOCK (since Linux 2.5.46)
This flag is meaningful only in conjunction with MAP_POPULATE. Don't perform read-ahead: create page tables entries only for pages that are already present in RAM. Since Linux 2.6.23, this flag causes MAP_POPULATEto do nothing. One day, the combination of MAP_POPULATEand MAP_NONBLOCKmay be reimplemented.
Do not reserve swap space for this mapping. When swap space is reserved, one has the guarantee that it is possible to modify the mapping. When swap space is not reserved one might get SIGSEGVupon a write if no physical memory is available. See also the discussion of the file /proc/sys/vm/overcommit_memoryin proc(5). In kernels before 2.6, this flag had effect only for private writable mappings.
MAP_POPULATE (since Linux 2.5.46)
Populate (prefault) page tables for a mapping. For a file mapping, this causes read-ahead on the file. This will help to reduce blocking on page faults later. MAP_POPULATEis supported for private mappings only since Linux 2.6.23.
MAP_STACK (since Linux 2.6.27)
Allocate the mapping at an address suitable for a process or thread stack. This flag is currently a no-op, but is used in the glibc threading implementation so that if some architectures require special treatment for stack allocations, support can later be transparently implemented for glibc.
MAP_UNINITIALIZED (since Linux 2.6.33)
Don't clear anonymous pages. This flag is intended to improve performance on embedded devices. This flag is honored only if the kernel was configured with the CONFIG_MMAP_ALLOW_UNINITIALIZEDoption. Because of the security implications, that option is normally enabled only on embedded devices (i.e., devices where one has complete control of the contents of user memory).

Of the above flags, only MAP_FIXEDis specified in POSIX.1-2001 and POSIX.1-2008. However, most systems also support MAP_ANONYMOUS(or its synonym MAP_ANON).

Some systems document the additional flags MAP_AUTOGROW, MAP_AUTORESRV, MAP_COPY, and MAP_LOCAL.

Memory mapped by mmap() is preserved across fork(2), with the same attributes.

A file is mapped in multiples of the page size. For a file that is not a multiple of the page size, the remaining memory is zeroed when mapped, and writes to that region are not written out to the file. The effect of changing the size of the underlying file of a mapping on the pages that correspond to added or removed regions of the file is unspecified.  


The munmap() system call deletes the mappings for the specified address range, and causes further references to addresses within the range to generate invalid memory references. The region is also automatically unmapped when the process is terminated. On the other hand, closing the file descriptor does not unmap the region.

The address addrmust be a multiple of the page size (but lengthneed not be). All pages containing a part of the indicated range are unmapped, and subsequent references to these pages will generate SIGSEGV. It is not an error if the indicated range does not contain any mapped pages.  


On success, mmap() returns a pointer to the mapped area. On error, the value MAP_FAILED(that is, (void *) -1) is returned, and errnois set to indicate the cause of the error.

On success, munmap() returns 0. On failure, it returns -1, and errnois set to indicate the cause of the error (probably to EINVAL).  


A file descriptor refers to a non-regular file. Or a file mapping was requested, but fdis not open for reading. Or MAP_SHAREDwas requested and PROT_WRITEis set, but fdis not open in read/write (O_RDWR) mode. Or PROT_WRITEis set, but the file is append-only.
The file has been locked, or too much memory has been locked (see setrlimit(2)).
fdis not a valid file descriptor (and MAP_ANONYMOUSwas not set).
We don't like addr, length, or offset(e.g., they are too large, or not aligned on a page boundary).
(since Linux 2.6.12) lengthwas 0.
flagscontained neither MAP_PRIVATEor MAP_SHARED, or contained both of these values.
The system-wide limit on the total number of open files has been reached.
The underlying filesystem of the specified file does not support memory mapping.
No memory is available.
The process's maximum number of mappings would have been exceeded. This error can also occur for munmap(), when unmapping a region in the middle of an existing mapping, since this results in two smaller mappings on either side of the region being unmapped.
On 32-bit architecture together with the large file extension (i.e., using 64-bit off_t): the number of pages used for lengthplus number of pages used for offsetwould overflow unsigned long(32 bits).
The protargument asks for PROT_EXECbut the mapped area belongs to a file on a filesystem that was mounted no-exec.
The operation was prevented by a file seal; see fcntl(2).
MAP_DENYWRITEwas set but the object specified by fdis open for writing.

Use of a mapped region can result in these signals:

Attempted write into a region mapped as read-only.
Attempted access to a portion of the buffer that does not correspond to the file (for example, beyond the end of the file, including the case where another process has truncated the file).


For an explanation of the terms used in this section, see attributes(7).
mmap(), munmap() Thread safetyMT-Safe


POSIX.1-2001, POSIX.1-2008, SVr4, 4.4BSD.  


On POSIX systems on which mmap(), msync(2), and munmap() are available, _POSIX_MAPPED_FILESis defined in <unistd.h> to a value greater than 0. (See also sysconf(3).)  


On some hardware architectures (e.g., i386), PROT_WRITEimplies PROT_READ. It is architecture dependent whether PROT_READimplies PROT_EXECor not. Portable programs should always set PROT_EXECif they intend to execute code in the new mapping.

The portable way to create a mapping is to specify addras 0 (NULL), and omit MAP_FIXEDfrom flags. In this case, the system chooses the address for the mapping; the address is chosen so as not to conflict with any existing mapping, and will not be 0. If the MAP_FIXEDflag is specified, and addris 0 (NULL), then the mapped address will be 0 (NULL).

Certain flagsconstants are defined only if suitable feature test macros are defined (possibly by default): _DEFAULT_SOURCEwith glibc 2.19 or later; or _BSD_SOURCEor _SVID_SOURCEin glibc 2.19 and earlier. (Employing _GNU_SOURCEalso suffices, and requiring that macro specifically would have been more logical, since these flags are all Linux-specific.) The relevant flags are: MAP_32BIT, MAP_ANONYMOUS(and the synonym MAP_ANON), MAP_DENYWRITE, MAP_EXECUTABLE, MAP_FILE, MAP_GROWSDOWN, MAP_HUGETLB, MAP_LOCKED, MAP_NONBLOCK, MAP_NORESERVE, MAP_POPULATE, and MAP_STACK.

An application can determine which pages of a mapping are currently resident in the buffer/page cache using mincore(2).  

Timestamps changes for file-backed mappings

For file-backed mappings, the st_atimefield for the mapped file may be updated at any time between the mmap() and the corresponding unmapping; the first reference to a mapped page will update the field if it has not been already.

The st_ctimeand st_mtimefield for a file mapped with PROT_WRITEand MAP_SHAREDwill be updated after a write to the mapped region, and before a subsequent msync(2) with the MS_SYNCor MS_ASYNCflag, if one occurs.  

Huge page (Huge TLB) mappings

For mappings that employ huge pages, the requirements for the arguments of mmap() and munmap() differ somewhat from the requirements for mappings that use the native system page size.

For mmap(), offsetmust be a multiple of the underlying huge page size. The system automatically aligns lengthto be a multiple of the underlying huge page size.

For munmap(), addrand lengthmust both be a multiple of the underlying huge page size.  

C library/kernel differences

This page describes the interface provided by the glibc mmap() wrapper function. Originally, this function invoked a system call of the same name. Since kernel 2.4, that system call has been superseded by mmap2(2), and nowadays the glibc mmap() wrapper function invokes mmap2(2) with a suitably adjusted value for offset.  


On Linux, there are no guarantees like those suggested above under MAP_NORESERVE. By default, any process can be killed at any moment when the system runs out of memory.

In kernels before 2.6.7, the MAP_POPULATEflag has effect only if protis specified as PROT_NONE.

SUSv3 specifies that mmap() should fail if lengthis 0. However, in kernels before 2.6.12, mmap() succeeded in this case: no mapping was created and the call returned addr. Since kernel 2.6.12, mmap() fails with the error EINVALfor this case.

POSIX specifies that the system shall always zero fill any partial page at the end of the object and that system will never write any modification of the object beyond its end. On Linux, when you write data to such partial page after the end of the object, the data stays in the page cache even after the file is closed and unmapped and even though the data is never written to the file itself, subsequent mappings may see the modified content. In some cases, this could be fixed by calling msync(2) before the unmap takes place; however, this doesn't work on tmpfs(5) (for example, when using POSIX shared memory interface documented in shm_overview(7)).  


The following program prints part of the file specified in its first command-line argument to standard output. The range of bytes to be printed is specified via offset and length values in the second and third command-line arguments. The program creates a memory mapping of the required pages of the file and then uses write(2) to output the desired bytes.  

Program source

#include <sys/mman.h>
#include <sys/stat.h>
#include <fcntl.h>
#include <stdio.h>
#include <stdlib.h>
#include <unistd.h>

#define handle_error(msg) \
    do { perror(msg); exit(EXIT_FAILURE); } while (0)

main(int argc, char *argv[])
    char *addr;
    int fd;
    struct stat sb;
    off_t offset, pa_offset;
    size_t length;
    ssize_t s;

    if (argc < 3 || argc > 4) {
        fprintf(stderr, "%s file offset [length]\n", argv[0]);

    fd = open(argv[1], O_RDONLY);
    if (fd == -1)

    if (fstat(fd, &sb) == -1)           /* To obtain file size */

    offset = atoi(argv[2]);
    pa_offset = offset & ~(sysconf(_SC_PAGE_SIZE) - 1);
        /* offset for mmap() must be page aligned */

    if (offset >= sb.st_size) {
        fprintf(stderr, "offset is past end of file\n");

    if (argc == 4) {
        length = atoi(argv[3]);
        if (offset + length > sb.st_size)
            length = sb.st_size - offset;
                /* Canaqt display bytes past end of file */

    } else {    /* No length arg ==> display to end of file */
        length = sb.st_size - offset;

    addr = mmap(NULL, length + offset - pa_offset, PROT_READ,
                MAP_PRIVATE, fd, pa_offset);
    if (addr == MAP_FAILED)

    s = write(STDOUT_FILENO, addr + offset - pa_offset, length);
    if (s != length) {
        if (s == -1)

        fprintf(stderr, "partial write");

    munmap(addr, length + offset - pa_offset);



getpagesize(2), memfd_create(2), mincore(2), mlock(2), mmap2(2), mprotect(2), mremap(2), msync(2), remap_file_pages(2), setrlimit(2), shmat(2), shm_open(3), shm_overview(7)

The descriptions of the following files in proc(5): /proc/[pid]/maps, /proc/[pid]/map_files, and /proc/[pid]/smaps.

B.O. Gallmeister, POSIX.4, O'Reilly, pp. 128-129 and 389-391.



Timestamps changes for file-backed mappings
Huge page (Huge TLB) mappings
C library/kernel differences
Program source

This document was created by man2html, using the manual pages.
Time: 16:30:27 GMT, March 14, 2017 Click Here!