mirror of
https://github.com/apache/nuttx.git
synced 2026-10-11 16:20:21 +00:00
f6ecf80ebb
The documentation grew one page at a time, so the tree follows the
history of who wrote what and not the shape of NuttX. Scheduling is
spread over three places, a driver page can sit above the subsystem
that owns it, and the front page lists everything at the same level.
That is a lot to face when all you want to know is where the scheduler
lives.
This change files every page under the code it describes. It is a move,
not a rewrite: outside the ten pages named below, every page keeps the
text that is already in master, and no page's text is deleted.
What it does:
* Groups the table of contents into nine chapters.
* Moves the OS subsystems under os/: scheduling, memory, drivers,
filesystem, networking, IPC, interrupts, libs, time.
* Renames the platform pages to the names the source tree uses, and
derives their tags from the tree instead of by hand.
* Splits guides/ by subject.
* Adds Documentation/redirects.py, with a rule for every page that left
its old path, so old URLs keep working. The redirect page also carries
a link's #anchor across to the new page.
Ten pages have text that is new or rewritten. Nine of them are the
landing page of a chapter, which has to exist for the new structure:
index the front page
os/index OS Design
os/scheduling/index Scheduling
os/interrupts/index Interrupts
os/ipc/index IPC
os/time/index Time and timers
about/index About
developing/index Developing NuttX
ReleaseNotes/index Release notes
The tenth is os/libs/libbuiltin, the only page here with technical
content: libs/libbuiltin/ had no page at all. Five SVG diagrams come
with these pages, hand-written XML with no editor metadata.
Nothing outside Documentation/ is touched.
How it was checked:
* Sphinx builds with -W: no warnings, and no document left outside a
toctree.
* A script, offered in the PR, proves the narrow claim this rests on.
For every page outside the ten named above it erases what a move
touches -- link target, path, tag line, toctree block, table border --
from the whole old text and the whole new text, and requires the two
to be byte for byte identical. It also requires every sentence of a
deleted page to turn up somewhere, and every page that left its old
path to have a redirect, from a URL that existed, to where its content
went. It exits non-zero and names the page if any of that is not true,
and it tests added pages too, so forgetting to declare one cannot make
it pass.
* An independent audit checked 133 factual claims on these ten pages
against the tree, one shell command per claim: 130 confirmed, 1
refuted and fixed here, 2 not checkable.
* tools/checkpatch.sh is clean over the range.
The diff is large because moving a page changes every link that points
to it. Most of it is pure renames, and board pages that gained one tag
line.
Assisted-by: Claude:claude-opus-5
125 lines
4.4 KiB
ReStructuredText
125 lines
4.4 KiB
ReStructuredText
=====================
|
|
Shared Memory Support
|
|
=====================
|
|
|
|
Prerequisites
|
|
-------------
|
|
|
|
These features must be enabled before shared memory support can be
|
|
provided:
|
|
|
|
- ``CONFIG_ARCH_ADDRENV=y`` - Support for per-task address environment using a
|
|
MMU.
|
|
- ``CONFIG_BUILD_KERNEL=y`` - Support for protected kernel-/user-space memory
|
|
regions must be provided by the MMU.
|
|
- ``CONFIG_GRAN=y`` - The granule allocation is the allocation underlying all
|
|
paged allocations.
|
|
- ``CONFIG_MM_PGALLOC=y`` - Enables the physical page allocator
|
|
- ``CONFIG_MM_PGSIZE`` - Determines the size of one page that can be mapped by
|
|
the MMU.
|
|
|
|
And then finally:
|
|
|
|
- ``CONFIG_MM_SHM=y`` - Enables shared memory support
|
|
- ``CONFIG_ARCH_SHM_VBASE`` - The virtual address of the beginning of the
|
|
shared memory region.
|
|
- ``CONFIG_ARCH_SHM_MAXREGIONS`` - The maximum number of regions that can
|
|
allocated for the shared memory space. This hard-coded value permits
|
|
static allocation of the shared memory data structures and serves no
|
|
other purpose. Default is 1.
|
|
- ``CONFIG_ARCH_SHM_NPAGES`` - The maximum number of pages that can allocated
|
|
for the shared memory region. Default is 1.
|
|
|
|
The size of the virtual shared memory address space is then determined by
|
|
the product of the maximum number of regions, the maximum number of pages
|
|
per region, and the configured size of each page.
|
|
|
|
Concepts
|
|
--------
|
|
|
|
Each process has a task group structure, struct task_group_s, that holds
|
|
information common to all threads in the group. If ``CONFIG_MM_SHM=y``, then
|
|
this includes data structures for the per-process shared memory virtual
|
|
page allocator.
|
|
|
|
A memory region is accessed using:
|
|
|
|
.. code-block:: C
|
|
|
|
int shmget(key_t key, size_t size, int shmflg);
|
|
|
|
by a lookup using internal shared memory data sets with key as the lookup
|
|
match value. On success, shmget returns the shared memory identifier for
|
|
the match -- in this implementation that identifier is simply the table
|
|
index of the match.
|
|
|
|
If the memory region does not exist, it may also be created by shmget (if
|
|
the IPC_CREAT bit is set in the shmflag). When a shared memory region is
|
|
created, the following things happen:
|
|
|
|
- A new entry is set aside in the internal data set. The key value is
|
|
assigned to the entry and the table index is the new shared memory
|
|
identifier.
|
|
|
|
- The requested size is rounded up to rounded up to full pages, each of
|
|
size ``CONFIG_MM_PGSIZE``.
|
|
|
|
- A set of physical pages are allocated and the physical address of
|
|
these pages is retained in the internal data set.
|
|
|
|
Now the key maps to and shared memory identifier (the table index) and
|
|
the table index provides access to the list of physical pages making up
|
|
the shared memory region.
|
|
|
|
NOTE: An improved implementation my perform a "lazy" back up of the
|
|
physical memory, i.e., do not allocate the physical memory until the
|
|
memory is required, for example, when a page fault occurs when a
|
|
application tries to allocate the memory.
|
|
|
|
A shared memory region is destroyed via:
|
|
|
|
.. code-block:: C
|
|
|
|
int shmctl(int shmid, int cmd, struct shmid_ds *buf);
|
|
|
|
In order for a process to make use of the memory region, it must be
|
|
"attached" the process using:
|
|
|
|
.. code-block:: C
|
|
|
|
FAR void *shmat(int shmid, FAR const void *shmaddr, int shmflg);
|
|
|
|
``shmat()`` returns the virtual address where the shared memory can be found
|
|
in the user process. Attaching the shared memory region involves the
|
|
following steps:
|
|
|
|
- Use the shmid as a table index to look up the mapping in the shared
|
|
memory internal data structures.
|
|
|
|
- Allocate a virtual address spaces of the same size as the physical
|
|
address space using the per-process virtual shared memory virtual
|
|
page allocator that can be found in the calling process's task group
|
|
structure.
|
|
|
|
- Use platform specific interfaces to mapy the physical memory to the
|
|
selected virtual address space, and
|
|
|
|
- Return the allocated virtual base address to the caller.
|
|
|
|
The memory region can be detached from the user process using:
|
|
|
|
.. code-block:: C
|
|
|
|
int shmdt(FAR const void *shmaddr);
|
|
|
|
Relevant header files
|
|
---------------------
|
|
|
|
- ``include/sys/shm.h`` - Shared memory interface declarations
|
|
- ``include/sys/ipc.h`` - Provides additional definitions used by the shared
|
|
memory interfaces
|
|
- ``include/nuttx/addrenv.h`` - Defines the virtual address space of the
|
|
process.
|
|
- ``include/nuttx/pgalloc.h`` - Page allocator interfaces
|
|
- ``mm/shm/shm.h`` - Internal shared memory definitions. This includes the
|
|
definitions of the internal shared memory data structures.
|