€•+Þ      Œsphinx.addnodes”Œdocument”“”)”}”(Œ	rawsource”Œ ”Œchildren”]”(Œtranslations”ŒLanguagesNode”“”)”}”(hhh]”(h Œpending_xref”“”)”}”(hhh]”Œdocutils.nodes”ŒText”“”ŒChinese (Simplified)”…””}”Œparent”hsbaŒ
attributes”}”(Œids”]”Œclasses”]”Œnames”]”Œdupnames”]”Œbackrefs”]”Œ	refdomain”Œstd”Œreftype”Œdoc”Œ	reftarget”Œ-/translations/zh_CN/admin-guide/mm/nommu-mmap”Œmodname”NŒ	classname”NŒrefexplicit”ˆuŒtagname”hhhubh)”}”(hhh]”hŒChinese (Traditional)”…””}”hh2sbah}”(h]”h ]”h"]”h$]”h&]”Œ	refdomain”h)Œreftype”h+Œ	reftarget”Œ-/translations/zh_TW/admin-guide/mm/nommu-mmap”Œmodname”NŒ	classname”NŒrefexplicit”ˆuh1hhhubh)”}”(hhh]”hŒItalian”…””}”hhFsbah}”(h]”h ]”h"]”h$]”h&]”Œ	refdomain”h)Œreftype”h+Œ	reftarget”Œ-/translations/it_IT/admin-guide/mm/nommu-mmap”Œmodname”NŒ	classname”NŒrefexplicit”ˆuh1hhhubh)”}”(hhh]”hŒJapanese”…””}”hhZsbah}”(h]”h ]”h"]”h$]”h&]”Œ	refdomain”h)Œreftype”h+Œ	reftarget”Œ-/translations/ja_JP/admin-guide/mm/nommu-mmap”Œmodname”NŒ	classname”NŒrefexplicit”ˆuh1hhhubh)”}”(hhh]”hŒKorean”…””}”hhnsbah}”(h]”h ]”h"]”h$]”h&]”Œ	refdomain”h)Œreftype”h+Œ	reftarget”Œ-/translations/ko_KR/admin-guide/mm/nommu-mmap”Œmodname”NŒ	classname”NŒrefexplicit”ˆuh1hhhubh)”}”(hhh]”hŒSpanish”…””}”hh‚sbah}”(h]”h ]”h"]”h$]”h&]”Œ	refdomain”h)Œreftype”h+Œ	reftarget”Œ-/translations/sp_SP/admin-guide/mm/nommu-mmap”Œmodname”NŒ	classname”NŒrefexplicit”ˆuh1hhhubeh}”(h]”h ]”h"]”h$]”h&]”Œcurrent_language”ŒEnglish”uh1h
hhŒ	_document”hŒsource”NŒline”NubhŒsection”“”)”}”(hhh]”(hŒtitle”“”)”}”(hŒNo-MMU memory mapping support”h]”hŒNo-MMU memory mapping support”…””}”(hh¨hžhhŸNh Nubah}”(h]”h ]”h"]”h$]”h&]”uh1h¦hh£hžhhŸŒG/var/lib/git/docbuild/linux/Documentation/admin-guide/mm/nommu-mmap.rst”h KubhŒ	paragraph”“”)”}”(hX®  The kernel has limited support for memory mapping under no-MMU conditions, such
as are used in uClinux environments. From the userspace point of view, memory
mapping is made use of in conjunction with the mmap() system call, the shmat()
call and the execve() system call. From the kernel's point of view, execve()
mapping is actually performed by the binfmt drivers, which call back into the
mmap() routines to do the actual work.”h]”hX°  The kernel has limited support for memory mapping under no-MMU conditions, such
as are used in uClinux environments. From the userspace point of view, memory
mapping is made use of in conjunction with the mmap() system call, the shmat()
call and the execve() system call. From the kernelâ€™s point of view, execve()
mapping is actually performed by the binfmt drivers, which call back into the
mmap() routines to do the actual work.”…””}”(hh¹hžhhŸNh Nubah}”(h]”h ]”h"]”h$]”h&]”uh1h·hŸh¶h Khh£hžhubh¸)”}”(hŒ¬Memory mapping behaviour also involves the way fork(), vfork(), clone() and
ptrace() work. Under uClinux there is no fork(), and clone() must be supplied
the CLONE_VM flag.”h]”hŒ¬Memory mapping behaviour also involves the way fork(), vfork(), clone() and
ptrace() work. Under uClinux there is no fork(), and clone() must be supplied
the CLONE_VM flag.”…””}”(hhÇhžhhŸNh Nubah}”(h]”h ]”h"]”h$]”h&]”uh1h·hŸh¶h Khh£hžhubh¸)”}”(hŒ„The behaviour is similar between the MMU and no-MMU cases, but not identical;
and it's also much more restricted in the latter case:”h]”hŒ†The behaviour is similar between the MMU and no-MMU cases, but not identical;
and itâ€™s also much more restricted in the latter case:”…””}”(hhÕhžhhŸNh Nubah}”(h]”h ]”h"]”h$]”h&]”uh1h·hŸh¶h Khh£hžhubhŒblock_quote”“”)”}”(hX¦  (#) Anonymous mapping, MAP_PRIVATE

       In the MMU case: VM regions backed by arbitrary pages; copy-on-write
       across fork.

       In the no-MMU case: VM regions backed by arbitrary contiguous runs of
       pages.

(#) Anonymous mapping, MAP_SHARED

       These behave very much like private mappings, except that they're
       shared across fork() or clone() without CLONE_VM in the MMU case. Since
       the no-MMU case doesn't support these, behaviour is identical to
       MAP_PRIVATE there.

(#) File, MAP_PRIVATE, PROT_READ / PROT_EXEC, !PROT_WRITE

       In the MMU case: VM regions backed by pages read from file; changes to
       the underlying file are reflected in the mapping; copied across fork.

       In the no-MMU case:

        - If one exists, the kernel will re-use an existing mapping to the
          same segment of the same file if that has compatible permissions,
          even if this was created by another process.

        - If possible, the file mapping will be directly on the backing device
          if the backing device has the NOMMU_MAP_DIRECT capability and
          appropriate mapping protection capabilities. Ramfs, romfs, cramfs
          and mtd might all permit this.

        - If the backing device can't or won't permit direct sharing,
          but does have the NOMMU_MAP_COPY capability, then a copy of the
          appropriate bit of the file will be read into a contiguous bit of
          memory and any extraneous space beyond the EOF will be cleared

        - Writes to the file do not affect the mapping; writes to the mapping
          are visible in other processes (no MMU protection), but should not
          happen.

(#) File, MAP_PRIVATE, PROT_READ / PROT_EXEC, PROT_WRITE

       In the MMU case: like the non-PROT_WRITE case, except that the pages in
       question get copied before the write actually happens. From that point
       on writes to the file underneath that page no longer get reflected into
       the mapping's backing pages. The page is then backed by swap instead.

       In the no-MMU case: works much like the non-PROT_WRITE case, except
       that a copy is always taken and never shared.

(#) Regular file / blockdev, MAP_SHARED, PROT_READ / PROT_EXEC / PROT_WRITE

       In the MMU case: VM regions backed by pages read from file; changes to
       pages written back to file; writes to file reflected into pages backing
       mapping; shared across fork.

       In the no-MMU case: not supported.

(#) Memory backed regular file, MAP_SHARED, PROT_READ / PROT_EXEC / PROT_WRITE

       In the MMU case: As for ordinary regular files.

       In the no-MMU case: The filesystem providing the memory-backed file
       (such as ramfs or tmpfs) may choose to honour an open, truncate, mmap
       sequence by providing a contiguous sequence of pages to map. In that
       case, a shared-writable memory mapping will be possible. It will work
       as for the MMU case. If the filesystem does not provide any such
       support, then the mapping request will be denied.

(#) Memory backed blockdev, MAP_SHARED, PROT_READ / PROT_EXEC / PROT_WRITE

       In the MMU case: As for ordinary regular files.

       In the no-MMU case: As for memory backed regular files, but the
       blockdev must be able to provide a contiguous run of pages without
       truncate being called. The ramdisk driver could do this if it allocated
       all its memory as a contiguous array upfront.

(#) Memory backed chardev, MAP_SHARED, PROT_READ / PROT_EXEC / PROT_WRITE

       In the MMU case: As for ordinary regular files.

       In the no-MMU case: The character device driver may choose to honour
       the mmap() by providing direct access to the underlying device if it
       provides memory or quasi-memory that can be accessed directly. Examples
       of such are frame buffers and flash devices. If the driver does not
       provide any such support, then the mapping request will be denied.

”h]”hŒenumerated_list”“”)”}”(hhh]”(hŒ	list_item”“”)”}”(hŒÌAnonymous mapping, MAP_PRIVATE

   In the MMU case: VM regions backed by arbitrary pages; copy-on-write
   across fork.

   In the no-MMU case: VM regions backed by arbitrary contiguous runs of
   pages.
”h]”(h¸)”}”(hŒAnonymous mapping, MAP_PRIVATE”h]”hŒAnonymous mapping, MAP_PRIVATE”…””}”(hhôhžhhŸNh Nubah}”(h]”h ]”h"]”h$]”h&]”uh1h·hŸh¶h Khhðubhä)”}”(hŒ In the MMU case: VM regions backed by arbitrary pages; copy-on-write
across fork.

In the no-MMU case: VM regions backed by arbitrary contiguous runs of
pages.
”h]”(h¸)”}”(hŒQIn the MMU case: VM regions backed by arbitrary pages; copy-on-write
across fork.”h]”hŒQIn the MMU case: VM regions backed by arbitrary pages; copy-on-write
across fork.”…””}”(hj  hžhhŸNh Nubah}”(h]”h ]”h"]”h$]”h&]”uh1h·hŸh¶h Khj  ubh¸)”}”(hŒLIn the no-MMU case: VM regions backed by arbitrary contiguous runs of
pages.”h]”hŒLIn the no-MMU case: VM regions backed by arbitrary contiguous runs of
pages.”…””}”(hj  hžhhŸNh Nubah}”(h]”h ]”h"]”h$]”h&]”uh1h·hŸh¶h Khj  ubeh}”(h]”h ]”h"]”h$]”h&]”uh1hãhŸh¶h Khhðubeh}”(h]”h ]”h"]”h$]”h&]”uh1hîhhëubhï)”}”(hX	  Anonymous mapping, MAP_SHARED

   These behave very much like private mappings, except that they're
   shared across fork() or clone() without CLONE_VM in the MMU case. Since
   the no-MMU case doesn't support these, behaviour is identical to
   MAP_PRIVATE there.
”h]”(h¸)”}”(hŒAnonymous mapping, MAP_SHARED”h]”hŒAnonymous mapping, MAP_SHARED”…””}”(hj2  hžhhŸNh Nubah}”(h]”h ]”h"]”h$]”h&]”uh1h·hŸh¶h Khj.  ubhä)”}”(hŒÞThese behave very much like private mappings, except that they're
shared across fork() or clone() without CLONE_VM in the MMU case. Since
the no-MMU case doesn't support these, behaviour is identical to
MAP_PRIVATE there.
”h]”h¸)”}”(hŒÝThese behave very much like private mappings, except that they're
shared across fork() or clone() without CLONE_VM in the MMU case. Since
the no-MMU case doesn't support these, behaviour is identical to
MAP_PRIVATE there.”h]”hŒáThese behave very much like private mappings, except that theyâ€™re
shared across fork() or clone() without CLONE_VM in the MMU case. Since
the no-MMU case doesnâ€™t support these, behaviour is identical to
MAP_PRIVATE there.”…””}”(hjD  hžhhŸNh Nubah}”(h]”h ]”h"]”h$]”h&]”uh1h·hŸh¶h Khj@  ubah}”(h]”h ]”h"]”h$]”h&]”uh1hãhŸh¶h Khj.  ubeh}”(h]”h ]”h"]”h$]”h&]”uh1hîhhëubhï)”}”(hXZ  File, MAP_PRIVATE, PROT_READ / PROT_EXEC, !PROT_WRITE

   In the MMU case: VM regions backed by pages read from file; changes to
   the underlying file are reflected in the mapping; copied across fork.

   In the no-MMU case:

    - If one exists, the kernel will re-use an existing mapping to the
      same segment of the same file if that has compatible permissions,
      even if this was created by another process.

    - If possible, the file mapping will be directly on the backing device
      if the backing device has the NOMMU_MAP_DIRECT capability and
      appropriate mapping protection capabilities. Ramfs, romfs, cramfs
      and mtd might all permit this.

    - If the backing device can't or won't permit direct sharing,
      but does have the NOMMU_MAP_COPY capability, then a copy of the
      appropriate bit of the file will be read into a contiguous bit of
      memory and any extraneous space beyond the EOF will be cleared

    - Writes to the file do not affect the mapping; writes to the mapping
      are visible in other processes (no MMU protection), but should not
      happen.
”h]”(h¸)”}”(hŒ5File, MAP_PRIVATE, PROT_READ / PROT_EXEC, !PROT_WRITE”h]”hŒ5File, MAP_PRIVATE, PROT_READ / PROT_EXEC, !PROT_WRITE”…””}”(hjb  hžhhŸNh Nubah}”(h]”h ]”h"]”h$]”h&]”uh1h·hŸh¶h K"hj^  ubhä)”}”(hXð  In the MMU case: VM regions backed by pages read from file; changes to
the underlying file are reflected in the mapping; copied across fork.

In the no-MMU case:

 - If one exists, the kernel will re-use an existing mapping to the
   same segment of the same file if that has compatible permissions,
   even if this was created by another process.

 - If possible, the file mapping will be directly on the backing device
   if the backing device has the NOMMU_MAP_DIRECT capability and
   appropriate mapping protection capabilities. Ramfs, romfs, cramfs
   and mtd might all permit this.

 - If the backing device can't or won't permit direct sharing,
   but does have the NOMMU_MAP_COPY capability, then a copy of the
   appropriate bit of the file will be read into a contiguous bit of
   memory and any extraneous space beyond the EOF will be cleared

 - Writes to the file do not affect the mapping; writes to the mapping
   are visible in other processes (no MMU protection), but should not
   happen.
”h]”(h¸)”}”(hŒŒIn the MMU case: VM regions backed by pages read from file; changes to
the underlying file are reflected in the mapping; copied across fork.”h]”hŒŒIn the MMU case: VM regions backed by pages read from file; changes to
the underlying file are reflected in the mapping; copied across fork.”…””}”(hjt  hžhhŸNh Nubah}”(h]”h ]”h"]”h$]”h&]”uh1h·hŸh¶h K$hjp  ubh¸)”}”(hŒIn the no-MMU case:”h]”hŒIn the no-MMU case:”…””}”(hj‚  hžhhŸNh Nubah}”(h]”h ]”h"]”h$]”h&]”uh1h·hŸh¶h K'hjp  ubhä)”}”(hX?  - If one exists, the kernel will re-use an existing mapping to the
  same segment of the same file if that has compatible permissions,
  even if this was created by another process.

- If possible, the file mapping will be directly on the backing device
  if the backing device has the NOMMU_MAP_DIRECT capability and
  appropriate mapping protection capabilities. Ramfs, romfs, cramfs
  and mtd might all permit this.

- If the backing device can't or won't permit direct sharing,
  but does have the NOMMU_MAP_COPY capability, then a copy of the
  appropriate bit of the file will be read into a contiguous bit of
  memory and any extraneous space beyond the EOF will be cleared

- Writes to the file do not affect the mapping; writes to the mapping
  are visible in other processes (no MMU protection), but should not
  happen.
”h]”hŒbullet_list”“”)”}”(hhh]”(hï)”}”(hŒ°If one exists, the kernel will re-use an existing mapping to the
same segment of the same file if that has compatible permissions,
even if this was created by another process.
”h]”h¸)”}”(hŒ¯If one exists, the kernel will re-use an existing mapping to the
same segment of the same file if that has compatible permissions,
even if this was created by another process.”h]”hŒ¯If one exists, the kernel will re-use an existing mapping to the
same segment of the same file if that has compatible permissions,
even if this was created by another process.”…””}”(hj  hžhhŸNh Nubah}”(h]”h ]”h"]”h$]”h&]”uh1h·hŸh¶h K)hj™  ubah}”(h]”h ]”h"]”h$]”h&]”uh1hîhj–  ubhï)”}”(hŒäIf possible, the file mapping will be directly on the backing device
if the backing device has the NOMMU_MAP_DIRECT capability and
appropriate mapping protection capabilities. Ramfs, romfs, cramfs
and mtd might all permit this.
”h]”h¸)”}”(hŒãIf possible, the file mapping will be directly on the backing device
if the backing device has the NOMMU_MAP_DIRECT capability and
appropriate mapping protection capabilities. Ramfs, romfs, cramfs
and mtd might all permit this.”h]”hŒãIf possible, the file mapping will be directly on the backing device
if the backing device has the NOMMU_MAP_DIRECT capability and
appropriate mapping protection capabilities. Ramfs, romfs, cramfs
and mtd might all permit this.”…””}”(hjµ  hžhhŸNh Nubah}”(h]”h ]”h"]”h$]”h&]”uh1h·hŸh¶h K-hj±  ubah}”(h]”h ]”h"]”h$]”h&]”uh1hîhj–  ubhï)”}”(hŒýIf the backing device can't or won't permit direct sharing,
but does have the NOMMU_MAP_COPY capability, then a copy of the
appropriate bit of the file will be read into a contiguous bit of
memory and any extraneous space beyond the EOF will be cleared
”h]”h¸)”}”(hŒüIf the backing device can't or won't permit direct sharing,
but does have the NOMMU_MAP_COPY capability, then a copy of the
appropriate bit of the file will be read into a contiguous bit of
memory and any extraneous space beyond the EOF will be cleared”h]”hX   If the backing device canâ€™t or wonâ€™t permit direct sharing,
but does have the NOMMU_MAP_COPY capability, then a copy of the
appropriate bit of the file will be read into a contiguous bit of
memory and any extraneous space beyond the EOF will be cleared”…””}”(hjÍ  hžhhŸNh Nubah}”(h]”h ]”h"]”h$]”h&]”uh1h·hŸh¶h K2hjÉ  ubah}”(h]”h ]”h"]”h$]”h&]”uh1hîhj–  ubhï)”}”(hŒWrites to the file do not affect the mapping; writes to the mapping
are visible in other processes (no MMU protection), but should not
happen.
”h]”h¸)”}”(hŒŽWrites to the file do not affect the mapping; writes to the mapping
are visible in other processes (no MMU protection), but should not
happen.”h]”hŒŽWrites to the file do not affect the mapping; writes to the mapping
are visible in other processes (no MMU protection), but should not
happen.”…””}”(hjå  hžhhŸNh Nubah}”(h]”h ]”h"]”h$]”h&]”uh1h·hŸh¶h K7hjá  ubah}”(h]”h ]”h"]”h$]”h&]”uh1hîhj–  ubeh}”(h]”h ]”h"]”h$]”h&]”Œbullet”Œ-”uh1j”  hŸh¶h K)hj  ubah}”(h]”h ]”h"]”h$]”h&]”uh1hãhŸh¶h K)hjp  ubeh}”(h]”h ]”h"]”h$]”h&]”uh1hãhŸh¶h K$hj^  ubeh}”(h]”h ]”h"]”h$]”h&]”uh1hîhhëubhï)”}”(hXØ  File, MAP_PRIVATE, PROT_READ / PROT_EXEC, PROT_WRITE

   In the MMU case: like the non-PROT_WRITE case, except that the pages in
   question get copied before the write actually happens. From that point
   on writes to the file underneath that page no longer get reflected into
   the mapping's backing pages. The page is then backed by swap instead.

   In the no-MMU case: works much like the non-PROT_WRITE case, except
   that a copy is always taken and never shared.
”h]”(h¸)”}”(hŒ4File, MAP_PRIVATE, PROT_READ / PROT_EXEC, PROT_WRITE”h]”hŒ4File, MAP_PRIVATE, PROT_READ / PROT_EXEC, PROT_WRITE”…””}”(hj  hžhhŸNh Nubah}”(h]”h ]”h"]”h$]”h&]”uh1h·hŸh¶h K;hj  ubhä)”}”(hX  In the MMU case: like the non-PROT_WRITE case, except that the pages in
question get copied before the write actually happens. From that point
on writes to the file underneath that page no longer get reflected into
the mapping's backing pages. The page is then backed by swap instead.

In the no-MMU case: works much like the non-PROT_WRITE case, except
that a copy is always taken and never shared.
”h]”(h¸)”}”(hX  In the MMU case: like the non-PROT_WRITE case, except that the pages in
question get copied before the write actually happens. From that point
on writes to the file underneath that page no longer get reflected into
the mapping's backing pages. The page is then backed by swap instead.”h]”hX  In the MMU case: like the non-PROT_WRITE case, except that the pages in
question get copied before the write actually happens. From that point
on writes to the file underneath that page no longer get reflected into
the mappingâ€™s backing pages. The page is then backed by swap instead.”…””}”(hj)  hžhhŸNh Nubah}”(h]”h ]”h"]”h$]”h&]”uh1h·hŸh¶h K=hj%  ubh¸)”}”(hŒqIn the no-MMU case: works much like the non-PROT_WRITE case, except
that a copy is always taken and never shared.”h]”hŒqIn the no-MMU case: works much like the non-PROT_WRITE case, except
that a copy is always taken and never shared.”…””}”(hj7  hžhhŸNh Nubah}”(h]”h ]”h"]”h$]”h&]”uh1h·hŸh¶h KBhj%  ubeh}”(h]”h ]”h"]”h$]”h&]”uh1hãhŸh¶h K=hj  ubeh}”(h]”h ]”h"]”h$]”h&]”uh1hîhhëubhï)”}”(hX%  Regular file / blockdev, MAP_SHARED, PROT_READ / PROT_EXEC / PROT_WRITE

   In the MMU case: VM regions backed by pages read from file; changes to
   pages written back to file; writes to file reflected into pages backing
   mapping; shared across fork.

   In the no-MMU case: not supported.
”h]”(h¸)”}”(hŒGRegular file / blockdev, MAP_SHARED, PROT_READ / PROT_EXEC / PROT_WRITE”h]”hŒGRegular file / blockdev, MAP_SHARED, PROT_READ / PROT_EXEC / PROT_WRITE”…””}”(hjU  hžhhŸNh Nubah}”(h]”h ]”h"]”h$]”h&]”uh1h·hŸh¶h KEhjQ  ubhä)”}”(hŒÐIn the MMU case: VM regions backed by pages read from file; changes to
pages written back to file; writes to file reflected into pages backing
mapping; shared across fork.

In the no-MMU case: not supported.
”h]”(h¸)”}”(hŒ«In the MMU case: VM regions backed by pages read from file; changes to
pages written back to file; writes to file reflected into pages backing
mapping; shared across fork.”h]”hŒ«In the MMU case: VM regions backed by pages read from file; changes to
pages written back to file; writes to file reflected into pages backing
mapping; shared across fork.”…””}”(hjg  hžhhŸNh Nubah}”(h]”h ]”h"]”h$]”h&]”uh1h·hŸh¶h KGhjc  ubh¸)”}”(hŒ"In the no-MMU case: not supported.”h]”hŒ"In the no-MMU case: not supported.”…””}”(hju  hžhhŸNh Nubah}”(h]”h ]”h"]”h$]”h&]”uh1h·hŸh¶h KKhjc  ubeh}”(h]”h ]”h"]”h$]”h&]”uh1hãhŸh¶h KGhjQ  ubeh}”(h]”h ]”h"]”h$]”h&]”uh1hîhhëubhï)”}”(hX  Memory backed regular file, MAP_SHARED, PROT_READ / PROT_EXEC / PROT_WRITE

   In the MMU case: As for ordinary regular files.

   In the no-MMU case: The filesystem providing the memory-backed file
   (such as ramfs or tmpfs) may choose to honour an open, truncate, mmap
   sequence by providing a contiguous sequence of pages to map. In that
   case, a shared-writable memory mapping will be possible. It will work
   as for the MMU case. If the filesystem does not provide any such
   support, then the mapping request will be denied.
”h]”(h¸)”}”(hŒJMemory backed regular file, MAP_SHARED, PROT_READ / PROT_EXEC / PROT_WRITE”h]”hŒJMemory backed regular file, MAP_SHARED, PROT_READ / PROT_EXEC / PROT_WRITE”…””}”(hj“  hžhhŸNh Nubah}”(h]”h ]”h"]”h$]”h&]”uh1h·hŸh¶h KMhj  ubhä)”}”(hX¹  In the MMU case: As for ordinary regular files.

In the no-MMU case: The filesystem providing the memory-backed file
(such as ramfs or tmpfs) may choose to honour an open, truncate, mmap
sequence by providing a contiguous sequence of pages to map. In that
case, a shared-writable memory mapping will be possible. It will work
as for the MMU case. If the filesystem does not provide any such
support, then the mapping request will be denied.
”h]”(h¸)”}”(hŒ/In the MMU case: As for ordinary regular files.”h]”hŒ/In the MMU case: As for ordinary regular files.”…””}”(hj¥  hžhhŸNh Nubah}”(h]”h ]”h"]”h$]”h&]”uh1h·hŸh¶h KOhj¡  ubh¸)”}”(hX‡  In the no-MMU case: The filesystem providing the memory-backed file
(such as ramfs or tmpfs) may choose to honour an open, truncate, mmap
sequence by providing a contiguous sequence of pages to map. In that
case, a shared-writable memory mapping will be possible. It will work
as for the MMU case. If the filesystem does not provide any such
support, then the mapping request will be denied.”h]”hX‡  In the no-MMU case: The filesystem providing the memory-backed file
(such as ramfs or tmpfs) may choose to honour an open, truncate, mmap
sequence by providing a contiguous sequence of pages to map. In that
case, a shared-writable memory mapping will be possible. It will work
as for the MMU case. If the filesystem does not provide any such
support, then the mapping request will be denied.”…””}”(hj³  hžhhŸNh Nubah}”(h]”h ]”h"]”h$]”h&]”uh1h·hŸh¶h KQhj¡  ubeh}”(h]”h ]”h"]”h$]”h&]”uh1hãhŸh¶h KOhj  ubeh}”(h]”h ]”h"]”h$]”h&]”uh1hîhhëubhï)”}”(hX  Memory backed blockdev, MAP_SHARED, PROT_READ / PROT_EXEC / PROT_WRITE

   In the MMU case: As for ordinary regular files.

   In the no-MMU case: As for memory backed regular files, but the
   blockdev must be able to provide a contiguous run of pages without
   truncate being called. The ramdisk driver could do this if it allocated
   all its memory as a contiguous array upfront.
”h]”(h¸)”}”(hŒFMemory backed blockdev, MAP_SHARED, PROT_READ / PROT_EXEC / PROT_WRITE”h]”hŒFMemory backed blockdev, MAP_SHARED, PROT_READ / PROT_EXEC / PROT_WRITE”…””}”(hjÑ  hžhhŸNh Nubah}”(h]”h ]”h"]”h$]”h&]”uh1h·hŸh¶h KXhjÍ  ubhä)”}”(hX*  In the MMU case: As for ordinary regular files.

In the no-MMU case: As for memory backed regular files, but the
blockdev must be able to provide a contiguous run of pages without
truncate being called. The ramdisk driver could do this if it allocated
all its memory as a contiguous array upfront.
”h]”(h¸)”}”(hŒ/In the MMU case: As for ordinary regular files.”h]”hŒ/In the MMU case: As for ordinary regular files.”…””}”(hjã  hžhhŸNh Nubah}”(h]”h ]”h"]”h$]”h&]”uh1h·hŸh¶h KZhjß  ubh¸)”}”(hŒøIn the no-MMU case: As for memory backed regular files, but the
blockdev must be able to provide a contiguous run of pages without
truncate being called. The ramdisk driver could do this if it allocated
all its memory as a contiguous array upfront.”h]”hŒøIn the no-MMU case: As for memory backed regular files, but the
blockdev must be able to provide a contiguous run of pages without
truncate being called. The ramdisk driver could do this if it allocated
all its memory as a contiguous array upfront.”…””}”(hjñ  hžhhŸNh Nubah}”(h]”h ]”h"]”h$]”h&]”uh1h·hŸh¶h K\hjß  ubeh}”(h]”h ]”h"]”h$]”h&]”uh1hãhŸh¶h KZhjÍ  ubeh}”(h]”h ]”h"]”h$]”h&]”uh1hîhhëubhï)”}”(hXä  Memory backed chardev, MAP_SHARED, PROT_READ / PROT_EXEC / PROT_WRITE

   In the MMU case: As for ordinary regular files.

   In the no-MMU case: The character device driver may choose to honour
   the mmap() by providing direct access to the underlying device if it
   provides memory or quasi-memory that can be accessed directly. Examples
   of such are frame buffers and flash devices. If the driver does not
   provide any such support, then the mapping request will be denied.

”h]”(h¸)”}”(hŒEMemory backed chardev, MAP_SHARED, PROT_READ / PROT_EXEC / PROT_WRITE”h]”hŒEMemory backed chardev, MAP_SHARED, PROT_READ / PROT_EXEC / PROT_WRITE”…””}”(hj  hžhhŸNh Nubah}”(h]”h ]”h"]”h$]”h&]”uh1h·hŸh¶h Kahj  ubhä)”}”(hX‹  In the MMU case: As for ordinary regular files.

In the no-MMU case: The character device driver may choose to honour
the mmap() by providing direct access to the underlying device if it
provides memory or quasi-memory that can be accessed directly. Examples
of such are frame buffers and flash devices. If the driver does not
provide any such support, then the mapping request will be denied.

”h]”(h¸)”}”(hŒ/In the MMU case: As for ordinary regular files.”h]”hŒ/In the MMU case: As for ordinary regular files.”…””}”(hj!  hžhhŸNh Nubah}”(h]”h ]”h"]”h$]”h&]”uh1h·hŸh¶h Kchj  ubh¸)”}”(hXX  In the no-MMU case: The character device driver may choose to honour
the mmap() by providing direct access to the underlying device if it
provides memory or quasi-memory that can be accessed directly. Examples
of such are frame buffers and flash devices. If the driver does not
provide any such support, then the mapping request will be denied.”h]”hXX  In the no-MMU case: The character device driver may choose to honour
the mmap() by providing direct access to the underlying device if it
provides memory or quasi-memory that can be accessed directly. Examples
of such are frame buffers and flash devices. If the driver does not
provide any such support, then the mapping request will be denied.”…””}”(hj/  hžhhŸNh Nubah}”(h]”h ]”h"]”h$]”h&]”uh1h·hŸh¶h Kehj  ubeh}”(h]”h ]”h"]”h$]”h&]”uh1hãhŸh¶h Kchj  ubeh}”(h]”h ]”h"]”h$]”h&]”uh1hîhhëubeh}”(h]”h ]”h"]”h$]”h&]”Œenumtype”Œarabic”Œprefix”Œ(”Œsuffix”Œ)”uh1héhhåubah}”(h]”h ]”h"]”h$]”h&]”uh1hãhŸh¶h Khh£hžhubh¢)”}”(hhh]”(h§)”}”(hŒFurther notes on no-MMU MMAP”h]”hŒFurther notes on no-MMU MMAP”…””}”(hj^  hžhhŸNh Nubah}”(h]”h ]”h"]”h$]”h&]”uh1h¦hj[  hžhhŸh¶h Kmubhä)”}”(hXs
  (#) A request for a private mapping of a file may return a buffer that is not
    page-aligned.  This is because XIP may take place, and the data may not be
    paged aligned in the backing store.

(#) A request for an anonymous mapping will always be page aligned.  If
    possible the size of the request should be a power of two otherwise some
    of the space may be wasted as the kernel must allocate a power-of-2
    granule but will only discard the excess if appropriately configured as
    this has an effect on fragmentation.

(#) The memory allocated by a request for an anonymous mapping will normally
    be cleared by the kernel before being returned in accordance with the
    Linux man pages (ver 2.22 or later).

    In the MMU case this can be achieved with reasonable performance as
    regions are backed by virtual pages, with the contents only being mapped
    to cleared physical pages when a write happens on that specific page
    (prior to which, the pages are effectively mapped to the global zero page
    from which reads can take place).  This spreads out the time it takes to
    initialize the contents of a page - depending on the write-usage of the
    mapping.

    In the no-MMU case, however, anonymous mappings are backed by physical
    pages, and the entire map is cleared at allocation time.  This can cause
    significant delays during a userspace malloc() as the C library does an
    anonymous mapping and the kernel then does a memset for the entire map.

    However, for memory that isn't required to be precleared - such as that
    returned by malloc() - mmap() can take a MAP_UNINITIALIZED flag to
    indicate to the kernel that it shouldn't bother clearing the memory before
    returning it.  Note that CONFIG_MMAP_ALLOW_UNINITIALIZED must be enabled
    to permit this, otherwise the flag will be ignored.

    uClibc uses this to speed up malloc(), and the ELF-FDPIC binfmt uses this
    to allocate the brk and stack region.

(#) A list of all the private copy and anonymous mappings on the system is
    visible through /proc/maps in no-MMU mode.

(#) A list of all the mappings in use by a process is visible through
    /proc/<pid>/maps in no-MMU mode.

(#) Supplying MAP_FIXED or a requesting a particular mapping address will
    result in an error.

(#) Files mapped privately usually have to have a read method provided by the
    driver or filesystem so that the contents can be read into the memory
    allocated if mmap() chooses not to map the backing device directly. An
    error will result if they don't. This is most likely to be encountered
    with character device files, pipes, fifos and sockets.

”h]”hê)”}”(hhh]”(hï)”}”(hŒ¹A request for a private mapping of a file may return a buffer that is not
page-aligned.  This is because XIP may take place, and the data may not be
paged aligned in the backing store.
”h]”h¸)”}”(hŒ¸A request for a private mapping of a file may return a buffer that is not
page-aligned.  This is because XIP may take place, and the data may not be
paged aligned in the backing store.”h]”hŒ¸A request for a private mapping of a file may return a buffer that is not
page-aligned.  This is because XIP may take place, and the data may not be
paged aligned in the backing store.”…””}”(hjw  hžhhŸNh Nubah}”(h]”h ]”h"]”h$]”h&]”uh1h·hŸh¶h Kohjs  ubah}”(h]”h ]”h"]”h$]”h&]”uh1hîhjp  ubhï)”}”(hX>  A request for an anonymous mapping will always be page aligned.  If
possible the size of the request should be a power of two otherwise some
of the space may be wasted as the kernel must allocate a power-of-2
granule but will only discard the excess if appropriately configured as
this has an effect on fragmentation.
”h]”h¸)”}”(hX=  A request for an anonymous mapping will always be page aligned.  If
possible the size of the request should be a power of two otherwise some
of the space may be wasted as the kernel must allocate a power-of-2
granule but will only discard the excess if appropriately configured as
this has an effect on fragmentation.”h]”hX=  A request for an anonymous mapping will always be page aligned.  If
possible the size of the request should be a power of two otherwise some
of the space may be wasted as the kernel must allocate a power-of-2
granule but will only discard the excess if appropriately configured as
this has an effect on fragmentation.”…””}”(hj  hžhhŸNh Nubah}”(h]”h ]”h"]”h$]”h&]”uh1h·hŸh¶h Kshj‹  ubah}”(h]”h ]”h"]”h$]”h&]”uh1hîhjp  ubhï)”}”(hXQ  The memory allocated by a request for an anonymous mapping will normally
be cleared by the kernel before being returned in accordance with the
Linux man pages (ver 2.22 or later).

In the MMU case this can be achieved with reasonable performance as
regions are backed by virtual pages, with the contents only being mapped
to cleared physical pages when a write happens on that specific page
(prior to which, the pages are effectively mapped to the global zero page
from which reads can take place).  This spreads out the time it takes to
initialize the contents of a page - depending on the write-usage of the
mapping.

In the no-MMU case, however, anonymous mappings are backed by physical
pages, and the entire map is cleared at allocation time.  This can cause
significant delays during a userspace malloc() as the C library does an
anonymous mapping and the kernel then does a memset for the entire map.

However, for memory that isn't required to be precleared - such as that
returned by malloc() - mmap() can take a MAP_UNINITIALIZED flag to
indicate to the kernel that it shouldn't bother clearing the memory before
returning it.  Note that CONFIG_MMAP_ALLOW_UNINITIALIZED must be enabled
to permit this, otherwise the flag will be ignored.

uClibc uses this to speed up malloc(), and the ELF-FDPIC binfmt uses this
to allocate the brk and stack region.
”h]”(h¸)”}”(hŒ³The memory allocated by a request for an anonymous mapping will normally
be cleared by the kernel before being returned in accordance with the
Linux man pages (ver 2.22 or later).”h]”hŒ³The memory allocated by a request for an anonymous mapping will normally
be cleared by the kernel before being returned in accordance with the
Linux man pages (ver 2.22 or later).”…””}”(hj§  hžhhŸNh Nubah}”(h]”h ]”h"]”h$]”h&]”uh1h·hŸh¶h Kyhj£  ubh¸)”}”(hXµ  In the MMU case this can be achieved with reasonable performance as
regions are backed by virtual pages, with the contents only being mapped
to cleared physical pages when a write happens on that specific page
(prior to which, the pages are effectively mapped to the global zero page
from which reads can take place).  This spreads out the time it takes to
initialize the contents of a page - depending on the write-usage of the
mapping.”h]”hXµ  In the MMU case this can be achieved with reasonable performance as
regions are backed by virtual pages, with the contents only being mapped
to cleared physical pages when a write happens on that specific page
(prior to which, the pages are effectively mapped to the global zero page
from which reads can take place).  This spreads out the time it takes to
initialize the contents of a page - depending on the write-usage of the
mapping.”…””}”(hjµ  hžhhŸNh Nubah}”(h]”h ]”h"]”h$]”h&]”uh1h·hŸh¶h K}hj£  ubh¸)”}”(hX  In the no-MMU case, however, anonymous mappings are backed by physical
pages, and the entire map is cleared at allocation time.  This can cause
significant delays during a userspace malloc() as the C library does an
anonymous mapping and the kernel then does a memset for the entire map.”h]”hX  In the no-MMU case, however, anonymous mappings are backed by physical
pages, and the entire map is cleared at allocation time.  This can cause
significant delays during a userspace malloc() as the C library does an
anonymous mapping and the kernel then does a memset for the entire map.”…””}”(hjÃ  hžhhŸNh Nubah}”(h]”h ]”h"]”h$]”h&]”uh1h·hŸh¶h K…hj£  ubh¸)”}”(hXR  However, for memory that isn't required to be precleared - such as that
returned by malloc() - mmap() can take a MAP_UNINITIALIZED flag to
indicate to the kernel that it shouldn't bother clearing the memory before
returning it.  Note that CONFIG_MMAP_ALLOW_UNINITIALIZED must be enabled
to permit this, otherwise the flag will be ignored.”h]”hXV  However, for memory that isnâ€™t required to be precleared - such as that
returned by malloc() - mmap() can take a MAP_UNINITIALIZED flag to
indicate to the kernel that it shouldnâ€™t bother clearing the memory before
returning it.  Note that CONFIG_MMAP_ALLOW_UNINITIALIZED must be enabled
to permit this, otherwise the flag will be ignored.”…””}”(hjÑ  hžhhŸNh Nubah}”(h]”h ]”h"]”h$]”h&]”uh1h·hŸh¶h KŠhj£  ubh¸)”}”(hŒouClibc uses this to speed up malloc(), and the ELF-FDPIC binfmt uses this
to allocate the brk and stack region.”h]”hŒouClibc uses this to speed up malloc(), and the ELF-FDPIC binfmt uses this
to allocate the brk and stack region.”…””}”(hjß  hžhhŸNh Nubah}”(h]”h ]”h"]”h$]”h&]”uh1h·hŸh¶h Khj£  ubeh}”(h]”h ]”h"]”h$]”h&]”uh1hîhjp  ubhï)”}”(hŒrA list of all the private copy and anonymous mappings on the system is
visible through /proc/maps in no-MMU mode.
”h]”h¸)”}”(hŒqA list of all the private copy and anonymous mappings on the system is
visible through /proc/maps in no-MMU mode.”h]”hŒqA list of all the private copy and anonymous mappings on the system is
visible through /proc/maps in no-MMU mode.”…””}”(hj÷  hžhhŸNh Nubah}”(h]”h ]”h"]”h$]”h&]”uh1h·hŸh¶h K“hjó  ubah}”(h]”h ]”h"]”h$]”h&]”uh1hîhjp  ubhï)”}”(hŒcA list of all the mappings in use by a process is visible through
/proc/<pid>/maps in no-MMU mode.
”h]”h¸)”}”(hŒbA list of all the mappings in use by a process is visible through
/proc/<pid>/maps in no-MMU mode.”h]”hŒbA list of all the mappings in use by a process is visible through
/proc/<pid>/maps in no-MMU mode.”…””}”(hj  hžhhŸNh Nubah}”(h]”h ]”h"]”h$]”h&]”uh1h·hŸh¶h K–hj  ubah}”(h]”h ]”h"]”h$]”h&]”uh1hîhjp  ubhï)”}”(hŒZSupplying MAP_FIXED or a requesting a particular mapping address will
result in an error.
”h]”h¸)”}”(hŒYSupplying MAP_FIXED or a requesting a particular mapping address will
result in an error.”h]”hŒYSupplying MAP_FIXED or a requesting a particular mapping address will
result in an error.”…””}”(hj'  hžhhŸNh Nubah}”(h]”h ]”h"]”h$]”h&]”uh1h·hŸh¶h K™hj#  ubah}”(h]”h ]”h"]”h$]”h&]”uh1hîhjp  ubhï)”}”(hXV  Files mapped privately usually have to have a read method provided by the
driver or filesystem so that the contents can be read into the memory
allocated if mmap() chooses not to map the backing device directly. An
error will result if they don't. This is most likely to be encountered
with character device files, pipes, fifos and sockets.

”h]”h¸)”}”(hXT  Files mapped privately usually have to have a read method provided by the
driver or filesystem so that the contents can be read into the memory
allocated if mmap() chooses not to map the backing device directly. An
error will result if they don't. This is most likely to be encountered
with character device files, pipes, fifos and sockets.”h]”hXV  Files mapped privately usually have to have a read method provided by the
driver or filesystem so that the contents can be read into the memory
allocated if mmap() chooses not to map the backing device directly. An
error will result if they donâ€™t. This is most likely to be encountered
with character device files, pipes, fifos and sockets.”…””}”(hj?  hžhhŸNh Nubah}”(h]”h ]”h"]”h$]”h&]”uh1h·hŸh¶h Kœhj;  ubah}”(h]”h ]”h"]”h$]”h&]”uh1hîhjp  ubeh}”(h]”h ]”h"]”h$]”h&]”jO  jP  jQ  jR  jS  jT  uh1héhjl  ubah}”(h]”h ]”h"]”h$]”h&]”uh1hãhŸh¶h Kohj[  hžhubeh}”(h]”Œfurther-notes-on-no-mmu-mmap”ah ]”h"]”Œfurther notes on no-mmu mmap”ah$]”h&]”uh1h¡hh£hžhhŸh¶h Kmubh¢)”}”(hhh]”(h§)”}”(hŒInterprocess shared memory”h]”hŒInterprocess shared memory”…””}”(hjj  hžhhŸNh Nubah}”(h]”h ]”h"]”h$]”h&]”uh1h¦hjg  hžhhŸh¶h K¤ubh¸)”}”(hŒ·Both SYSV IPC SHM shared memory and POSIX shared memory is supported in NOMMU
mode.  The former through the usual mechanism, the latter through files created
on ramfs or tmpfs mounts.”h]”hŒ·Both SYSV IPC SHM shared memory and POSIX shared memory is supported in NOMMU
mode.  The former through the usual mechanism, the latter through files created
on ramfs or tmpfs mounts.”…””}”(hjx  hžhhŸNh Nubah}”(h]”h ]”h"]”h$]”h&]”uh1h·hŸh¶h K¦hjg  hžhubeh}”(h]”Œinterprocess-shared-memory”ah ]”h"]”Œinterprocess shared memory”ah$]”h&]”uh1h¡hh£hžhhŸh¶h K¤ubh¢)”}”(hhh]”(h§)”}”(hŒFutexes”h]”hŒFutexes”…””}”(hj‘  hžhhŸNh Nubah}”(h]”h ]”h"]”h$]”h&]”uh1h¦hjŽ  hžhhŸh¶h K¬ubh¸)”}”(hX  Futexes are supported in NOMMU mode if the arch supports them.  An error will
be given if an address passed to the futex system call lies outside the
mappings made by a process or if the mapping in which the address lies does not
support futexes (such as an I/O chardev mapping).”h]”hX  Futexes are supported in NOMMU mode if the arch supports them.  An error will
be given if an address passed to the futex system call lies outside the
mappings made by a process or if the mapping in which the address lies does not
support futexes (such as an I/O chardev mapping).”…””}”(hjŸ  hžhhŸNh Nubah}”(h]”h ]”h"]”h$]”h&]”uh1h·hŸh¶h K®hjŽ  hžhubeh}”(h]”Œfutexes”ah ]”h"]”Œfutexes”ah$]”h&]”uh1h¡hh£hžhhŸh¶h K¬ubh¢)”}”(hhh]”(h§)”}”(hŒNo-MMU mremap”h]”hŒNo-MMU mremap”…””}”(hj¸  hžhhŸNh Nubah}”(h]”h ]”h"]”h$]”h&]”uh1h¦hjµ  hžhhŸh¶h Kµubh¸)”}”(hX7  The mremap() function is partially supported.  It may change the size of a
mapping, and may move it [#]_ if MREMAP_MAYMOVE is specified and if the new size
of the mapping exceeds the size of the slab object currently occupied by the
memory to which the mapping refers, or if a smaller slab object could be used.”h]”(hŒdThe mremap() function is partially supported.  It may change the size of a
mapping, and may move it ”…””}”(hjÆ  hžhhŸNh NubhŒfootnote_reference”“”)”}”(hŒ[#]_”h]”hŒ1”…””}”(hjÐ  hžhhŸNh Nubah}”(h]”Œid1”ah ]”h"]”h$]”h&]”Œauto”KŒrefid”Œid2”Œdocname”Œadmin-guide/mm/nommu-mmap”uh1jÎ  hjÆ  Œresolved”KubhŒÏ if MREMAP_MAYMOVE is specified and if the new size
of the mapping exceeds the size of the slab object currently occupied by the
memory to which the mapping refers, or if a smaller slab object could be used.”…””}”(hjÆ  hžhhŸNh Nubeh}”(h]”h ]”h"]”h$]”h&]”uh1h·hŸh¶h K·hjµ  hžhubh¸)”}”(hŒ}MREMAP_FIXED is not supported, though it is ignored if there's no change of
address and the object does not need to be moved.”h]”hŒMREMAP_FIXED is not supported, though it is ignored if thereâ€™s no change of
address and the object does not need to be moved.”…””}”(hjï  hžhhŸNh Nubah}”(h]”h ]”h"]”h$]”h&]”uh1h·hŸh¶h K¼hjµ  hžhubh¸)”}”(hŒuShared mappings may not be moved.  Shareable mappings may not be moved either,
even if they are not currently shared.”h]”hŒuShared mappings may not be moved.  Shareable mappings may not be moved either,
even if they are not currently shared.”…””}”(hjý  hžhhŸNh Nubah}”(h]”h ]”h"]”h$]”h&]”uh1h·hŸh¶h K¿hjµ  hžhubh¸)”}”(hX  The mremap() function must be given an exact match for base address and size of
a previously mapped object.  It may not be used to create holes in existing
mappings, move parts of existing mappings or resize parts of mappings.  It must
act on a complete mapping.”h]”hX  The mremap() function must be given an exact match for base address and size of
a previously mapped object.  It may not be used to create holes in existing
mappings, move parts of existing mappings or resize parts of mappings.  It must
act on a complete mapping.”…””}”(hj  hžhhŸNh Nubah}”(h]”h ]”h"]”h$]”h&]”uh1h·hŸh¶h KÂhjµ  hžhubhŒfootnote”“”)”}”(hŒNot currently supported.

”h]”(hŒlabel”“”)”}”(hhh]”hŒ1”…””}”(hj!  hžhhŸNh Nubah}”(h]”h ]”h"]”h$]”h&]”uh1j  hj  hžhhŸNh Nubh¸)”}”(hŒNot currently supported.”h]”hŒNot currently supported.”…””}”(hj.  hžhhŸNh Nubah}”(h]”h ]”h"]”h$]”h&]”uh1h·hŸh¶h KÇhj  ubeh}”(h]”já  ah ]”h"]”Œ1”ah$]”h&]”jÚ  ajß  Kjâ  jã  uh1j  hŸh¶h KÇhjµ  hžhubeh}”(h]”Œno-mmu-mremap”ah ]”h"]”Œno-mmu mremap”ah$]”h&]”uh1h¡hh£hžhhŸh¶h Kµubh¢)”}”(hhh]”(h§)”}”(hŒ,Providing shareable character device support”h]”hŒ,Providing shareable character device support”…””}”(hjN  hžhhŸNh Nubah}”(h]”h ]”h"]”h$]”h&]”uh1h¦hjK  hžhhŸh¶h KËubh¸)”}”(hXd  To provide shareable character device support, a driver must provide a
file->f_op->get_unmapped_area() operation. The mmap() routines will call this
to get a proposed address for the mapping. This may return an error if it
doesn't wish to honour the mapping because it's too long, at a weird offset,
under some unsupported combination of flags or whatever.”h]”hXh  To provide shareable character device support, a driver must provide a
file->f_op->get_unmapped_area() operation. The mmap() routines will call this
to get a proposed address for the mapping. This may return an error if it
doesnâ€™t wish to honour the mapping because itâ€™s too long, at a weird offset,
under some unsupported combination of flags or whatever.”…””}”(hj\  hžhhŸNh Nubah}”(h]”h ]”h"]”h$]”h&]”uh1h·hŸh¶h KÍhjK  hžhubh¸)”}”(hŒþThe driver should also provide backing device information with capabilities set
to indicate the permitted types of mapping on such devices. The default is
assumed to be readable and writable, not executable, and only shareable
directly (can't be copied).”h]”hX   The driver should also provide backing device information with capabilities set
to indicate the permitted types of mapping on such devices. The default is
assumed to be readable and writable, not executable, and only shareable
directly (canâ€™t be copied).”…””}”(hjj  hžhhŸNh Nubah}”(h]”h ]”h"]”h$]”h&]”uh1h·hŸh¶h KÓhjK  hžhubh¸)”}”(hŒÜThe file->f_op->mmap() operation will be called to actually inaugurate the
mapping. It can be rejected at that point. Returning the ENOSYS error will
cause the mapping to be copied instead if NOMMU_MAP_COPY is specified.”h]”hŒÜThe file->f_op->mmap() operation will be called to actually inaugurate the
mapping. It can be rejected at that point. Returning the ENOSYS error will
cause the mapping to be copied instead if NOMMU_MAP_COPY is specified.”…””}”(hjx  hžhhŸNh Nubah}”(h]”h ]”h"]”h$]”h&]”uh1h·hŸh¶h KØhjK  hžhubh¸)”}”(hŒºThe vm_ops->close() routine will be invoked when the last mapping on a chardev
is removed. An existing mapping will be shared, partially or not, if possible
without notifying the driver.”h]”hŒºThe vm_ops->close() routine will be invoked when the last mapping on a chardev
is removed. An existing mapping will be shared, partially or not, if possible
without notifying the driver.”…””}”(hj†  hžhhŸNh Nubah}”(h]”h ]”h"]”h$]”h&]”uh1h·hŸh¶h KÜhjK  hžhubh¸)”}”(hX&  It is permitted also for the file->f_op->get_unmapped_area() operation to
return -ENOSYS. This will be taken to mean that this operation just doesn't
want to handle it, despite the fact it's got an operation. For instance, it
might try directing the call to a secondary driver which turns out not to
implement it. Such is the case for the framebuffer driver which attempts to
direct the call to the device-specific driver. Under such circumstances, the
mapping request will be rejected if NOMMU_MAP_COPY is not specified, and a
copy mapped otherwise.”h]”hX*  It is permitted also for the file->f_op->get_unmapped_area() operation to
return -ENOSYS. This will be taken to mean that this operation just doesnâ€™t
want to handle it, despite the fact itâ€™s got an operation. For instance, it
might try directing the call to a secondary driver which turns out not to
implement it. Such is the case for the framebuffer driver which attempts to
direct the call to the device-specific driver. Under such circumstances, the
mapping request will be rejected if NOMMU_MAP_COPY is not specified, and a
copy mapped otherwise.”…””}”(hj”  hžhhŸNh Nubah}”(h]”h ]”h"]”h$]”h&]”uh1h·hŸh¶h KàhjK  hžhubhŒ	important”“”)”}”(hX  Some types of device may present a different appearance to anyone
looking at them in certain modes. Flash chips can be like this; for
instance if they're in programming or erase mode, you might see the
status reflected in the mapping, instead of the data.

In such a case, care must be taken lest userspace see a shared or a
private mapping showing such information when the driver is busy
controlling the device. Remember especially: private executable
mappings may still be mapped directly off the device under some
circumstances!”h]”(h¸)”}”(hŒÿSome types of device may present a different appearance to anyone
looking at them in certain modes. Flash chips can be like this; for
instance if they're in programming or erase mode, you might see the
status reflected in the mapping, instead of the data.”h]”hX  Some types of device may present a different appearance to anyone
looking at them in certain modes. Flash chips can be like this; for
instance if theyâ€™re in programming or erase mode, you might see the
status reflected in the mapping, instead of the data.”…””}”(hj¨  hžhhŸNh Nubah}”(h]”h ]”h"]”h$]”h&]”uh1h·hŸh¶h Këhj¤  ubh¸)”}”(hX  In such a case, care must be taken lest userspace see a shared or a
private mapping showing such information when the driver is busy
controlling the device. Remember especially: private executable
mappings may still be mapped directly off the device under some
circumstances!”h]”hX  In such a case, care must be taken lest userspace see a shared or a
private mapping showing such information when the driver is busy
controlling the device. Remember especially: private executable
mappings may still be mapped directly off the device under some
circumstances!”…””}”(hj¶  hžhhŸNh Nubah}”(h]”h ]”h"]”h$]”h&]”uh1h·hŸh¶h Kðhj¤  ubeh}”(h]”h ]”h"]”h$]”h&]”uh1j¢  hjK  hžhhŸh¶h Nubeh}”(h]”Œ,providing-shareable-character-device-support”ah ]”h"]”Œ,providing shareable character device support”ah$]”h&]”uh1h¡hh£hžhhŸh¶h KËubh¢)”}”(hhh]”(h§)”}”(hŒ.Providing shareable memory-backed file support”h]”hŒ.Providing shareable memory-backed file support”…””}”(hjÕ  hžhhŸNh Nubah}”(h]”h ]”h"]”h$]”h&]”uh1h¦hjÒ  hžhhŸh¶h Køubh¸)”}”(hX   Provision of shared mappings on memory backed files is similar to the provision
of support for shared mapped character devices. The main difference is that the
filesystem providing the service will probably allocate a contiguous collection
of pages and permit mappings to be made on that.”h]”hX   Provision of shared mappings on memory backed files is similar to the provision
of support for shared mapped character devices. The main difference is that the
filesystem providing the service will probably allocate a contiguous collection
of pages and permit mappings to be made on that.”…””}”(hjã  hžhhŸNh Nubah}”(h]”h ]”h"]”h$]”h&]”uh1h·hŸh¶h KúhjÒ  hžhubh¸)”}”(hŒêIt is recommended that a truncate operation applied to such a file that
increases the file size, if that file is empty, be taken as a request to gather
enough pages to honour a mapping. This is required to support POSIX shared
memory.”h]”hŒêIt is recommended that a truncate operation applied to such a file that
increases the file size, if that file is empty, be taken as a request to gather
enough pages to honour a mapping. This is required to support POSIX shared
memory.”…””}”(hjñ  hžhhŸNh Nubah}”(h]”h ]”h"]”h$]”h&]”uh1h·hŸh¶h KÿhjÒ  hžhubh¸)”}”(hŒkMemory backed devices are indicated by the mapping's backing device info having
the memory_backed flag set.”h]”hŒmMemory backed devices are indicated by the mappingâ€™s backing device info having
the memory_backed flag set.”…””}”(hjÿ  hžhhŸNh Nubah}”(h]”h ]”h"]”h$]”h&]”uh1h·hŸh¶h MhjÒ  hžhubeh}”(h]”Œ.providing-shareable-memory-backed-file-support”ah ]”h"]”Œ.providing shareable memory-backed file support”ah$]”h&]”uh1h¡hh£hžhhŸh¶h Køubh¢)”}”(hhh]”(h§)”}”(hŒ(Providing shareable block device support”h]”hŒ(Providing shareable block device support”…””}”(hj  hžhhŸNh Nubah}”(h]”h ]”h"]”h$]”h&]”uh1h¦hj  hžhhŸh¶h M	ubh¸)”}”(hŒçProvision of shared mappings on block device files is exactly the same as for
character devices. If there isn't a real device underneath, then the driver
should allocate sufficient contiguous memory to honour any supported mapping.”h]”hŒéProvision of shared mappings on block device files is exactly the same as for
character devices. If there isnâ€™t a real device underneath, then the driver
should allocate sufficient contiguous memory to honour any supported mapping.”…””}”(hj&  hžhhŸNh Nubah}”(h]”h ]”h"]”h$]”h&]”uh1h·hŸh¶h Mhj  hžhubeh}”(h]”Œ(providing-shareable-block-device-support”ah ]”h"]”Œ(providing shareable block device support”ah$]”h&]”uh1h¡hh£hžhhŸh¶h M	ubh¢)”}”(hhh]”(h§)”}”(hŒ!Adjusting page trimming behaviour”h]”hŒ!Adjusting page trimming behaviour”…””}”(hj?  hžhhŸNh Nubah}”(h]”h ]”h"]”h$]”h&]”uh1h¦hj<  hžhhŸh¶h Mubh¸)”}”(hXì  NOMMU mmap automatically rounds up to the nearest power-of-2 number of pages
when performing an allocation.  This can have adverse effects on memory
fragmentation, and as such, is left configurable.  The default behaviour is to
aggressively trim allocations and discard any excess pages back in to the page
allocator.  In order to retain finer-grained control over fragmentation, this
behaviour can either be disabled completely, or bumped up to a higher page
watermark where trimming begins.”h]”hXì  NOMMU mmap automatically rounds up to the nearest power-of-2 number of pages
when performing an allocation.  This can have adverse effects on memory
fragmentation, and as such, is left configurable.  The default behaviour is to
aggressively trim allocations and discard any excess pages back in to the page
allocator.  In order to retain finer-grained control over fragmentation, this
behaviour can either be disabled completely, or bumped up to a higher page
watermark where trimming begins.”…””}”(hjM  hžhhŸNh Nubah}”(h]”h ]”h"]”h$]”h&]”uh1h·hŸh¶h Mhj<  hžhubh¸)”}”(hŒLPage trimming behaviour is configurable via the sysctl ``vm.nr_trim_pages``.”h]”(hŒ7Page trimming behaviour is configurable via the sysctl ”…””}”(hj[  hžhhŸNh NubhŒliteral”“”)”}”(hŒ``vm.nr_trim_pages``”h]”hŒvm.nr_trim_pages”…””}”(hje  hžhhŸNh Nubah}”(h]”h ]”h"]”h$]”h&]”uh1jc  hj[  ubhŒ.”…””}”(hj[  hžhhŸNh Nubeh}”(h]”h ]”h"]”h$]”h&]”uh1h·hŸh¶h Mhj<  hžhubeh}”(h]”Œ!adjusting-page-trimming-behaviour”ah ]”h"]”Œ!adjusting page trimming behaviour”ah$]”h&]”uh1h¡hh£hžhhŸh¶h Mubeh}”(h]”Œno-mmu-memory-mapping-support”ah ]”h"]”Œno-mmu memory mapping support”ah$]”h&]”uh1h¡hhhžhhŸh¶h Kubeh}”(h]”h ]”h"]”h$]”h&]”Œsource”h¶uh1hŒcurrent_source”NŒcurrent_line”NŒsettings”Œdocutils.frontend”ŒValues”“”)”}”(h¦NŒ	generator”NŒ	datestamp”NŒsource_link”NŒ
source_url”NŒtoc_backlinks”Œentry”Œfootnote_backlinks”KŒsectnum_xform”KŒstrip_comments”NŒstrip_elements_with_classes”NŒstrip_classes”NŒreport_level”KŒ
halt_level”KŒexit_status_level”KŒdebug”NŒwarning_stream”NŒ	traceback”ˆŒinput_encoding”Œ	utf-8-sig”Œinput_encoding_error_handler”Œstrict”Œoutput_encoding”Œutf-8”Œoutput_encoding_error_handler”j°  Œerror_encoding”Œutf-8”Œerror_encoding_error_handler”Œbackslashreplace”Œlanguage_code”Œen”Œrecord_dependencies”NŒconfig”NŒ	id_prefix”hŒauto_id_prefix”Œid”Œdump_settings”NŒdump_internals”NŒdump_transforms”NŒdump_pseudo_xml”NŒexpose_internals”NŒstrict_visitor”NŒ_disable_config”NŒ_source”h¶Œ_destination”NŒ_config_files”]”Œ7/var/lib/git/docbuild/linux/Documentation/docutils.conf”aŒfile_insertion_enabled”ˆŒraw_enabled”KŒline_length_limit”M'Œpep_references”NŒpep_base_url”Œhttps://peps.python.org/”Œpep_file_url_template”Œpep-%04d”Œrfc_references”NŒrfc_base_url”Œ&https://datatracker.ietf.org/doc/html/”Œ	tab_width”KŒtrim_footnote_reference_space”‰Œsyntax_highlight”Œlong”Œsmart_quotes”ˆŒsmartquotes_locales”]”Œcharacter_level_inline_markup”‰Œdoctitle_xform”‰Œdocinfo_xform”KŒsectsubtitle_xform”‰Œimage_loading”Œlink”Œembed_stylesheet”‰Œcloak_email_addresses”ˆŒsection_self_link”‰Œenv”NubŒreporter”NŒindirect_targets”]”Œsubstitution_defs”}”Œsubstitution_names”}”Œrefnames”}”Œrefids”}”já  ]”jÐ  asŒnameids”}”(jŠ  j‡  jd  ja  j‹  jˆ  j²  j¯  jH  jE  jÏ  jÌ  j  j  j9  j6  j‚  j  j@  já  uŒ	nametypes”}”(jŠ  ‰jd  ‰j‹  ‰j²  ‰jH  ‰jÏ  ‰j  ‰j9  ‰j‚  ‰j@  ˆuh}”(j‡  h£ja  j[  jˆ  jg  j¯  jŽ  jE  jµ  jÚ  jÐ  já  j  jÌ  jK  j  jÒ  j6  j  j  j<  uŒfootnote_refs”}”Œcitation_refs”}”Œautofootnotes”]”j  aŒautofootnote_refs”]”jÐ  aŒsymbol_footnotes”]”Œsymbol_footnote_refs”]”Œ	footnotes”]”Œ	citations”]”Œautofootnote_start”KŒsymbol_footnote_start”K Œ
id_counter”Œcollections”ŒCounter”“”}”j¾  Ks…”R”Œparse_messages”]”Œtransform_messages”]”Œtransformer”NŒinclude_log”]”Œ
decoration”Nhžhub.