linux/Documentation/core-api/memory-allocation.rst
Vlastimil Babka 59bb47985c mm, sl[aou]b: guarantee natural alignment for kmalloc(power-of-two)
In most configurations, kmalloc() happens to return naturally aligned
(i.e.  aligned to the block size itself) blocks for power of two sizes.

That means some kmalloc() users might unknowingly rely on that
alignment, until stuff breaks when the kernel is built with e.g.
CONFIG_SLUB_DEBUG or CONFIG_SLOB, and blocks stop being aligned.  Then
developers have to devise workaround such as own kmem caches with
specified alignment [1], which is not always practical, as recently
evidenced in [2].

The topic has been discussed at LSF/MM 2019 [3].  Adding a
'kmalloc_aligned()' variant would not help with code unknowingly relying
on the implicit alignment.  For slab implementations it would either
require creating more kmalloc caches, or allocate a larger size and only
give back part of it.  That would be wasteful, especially with a generic
alignment parameter (in contrast with a fixed alignment to size).

Ideally we should provide to mm users what they need without difficult
workarounds or own reimplementations, so let's make the kmalloc()
alignment to size explicitly guaranteed for power-of-two sizes under all
configurations.  What this means for the three available allocators?

* SLAB object layout happens to be mostly unchanged by the patch.  The
  implicitly provided alignment could be compromised with
  CONFIG_DEBUG_SLAB due to redzoning, however SLAB disables redzoning for
  caches with alignment larger than unsigned long long.  Practically on at
  least x86 this includes kmalloc caches as they use cache line alignment,
  which is larger than that.  Still, this patch ensures alignment on all
  arches and cache sizes.

* SLUB layout is also unchanged unless redzoning is enabled through
  CONFIG_SLUB_DEBUG and boot parameter for the particular kmalloc cache.
  With this patch, explicit alignment is guaranteed with redzoning as
  well.  This will result in more memory being wasted, but that should be
  acceptable in a debugging scenario.

* SLOB has no implicit alignment so this patch adds it explicitly for
  kmalloc().  The potential downside is increased fragmentation.  While
  pathological allocation scenarios are certainly possible, in my testing,
  after booting a x86_64 kernel+userspace with virtme, around 16MB memory
  was consumed by slab pages both before and after the patch, with
  difference in the noise.

[1] https://lore.kernel.org/linux-btrfs/c3157c8e8e0e7588312b40c853f65c02fe6c957a.1566399731.git.christophe.leroy@c-s.fr/
[2] https://lore.kernel.org/linux-fsdevel/20190225040904.5557-1-ming.lei@redhat.com/
[3] https://lwn.net/Articles/787740/

[akpm@linux-foundation.org: documentation fixlet, per Matthew]
Link: http://lkml.kernel.org/r/20190826111627.7505-3-vbabka@suse.cz
Signed-off-by: Vlastimil Babka <vbabka@suse.cz>
Reviewed-by: Matthew Wilcox (Oracle) <willy@infradead.org>
Acked-by: Michal Hocko <mhocko@suse.com>
Acked-by: Kirill A. Shutemov <kirill.shutemov@linux.intel.com>
Acked-by: Christoph Hellwig <hch@lst.de>
Cc: David Sterba <dsterba@suse.cz>
Cc: Christoph Lameter <cl@linux.com>
Cc: Pekka Enberg <penberg@kernel.org>
Cc: David Rientjes <rientjes@google.com>
Cc: Ming Lei <ming.lei@redhat.com>
Cc: Dave Chinner <david@fromorbit.com>
Cc: "Darrick J . Wong" <darrick.wong@oracle.com>
Cc: Christoph Hellwig <hch@lst.de>
Cc: James Bottomley <James.Bottomley@HansenPartnership.com>
Cc: Vlastimil Babka <vbabka@suse.cz>
Cc: Joonsoo Kim <iamjoonsoo.kim@lge.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2019-10-07 15:47:20 -07:00

131 lines
6.1 KiB
ReStructuredText

.. _memory_allocation:
=======================
Memory Allocation Guide
=======================
Linux provides a variety of APIs for memory allocation. You can
allocate small chunks using `kmalloc` or `kmem_cache_alloc` families,
large virtually contiguous areas using `vmalloc` and its derivatives,
or you can directly request pages from the page allocator with
`alloc_pages`. It is also possible to use more specialized allocators,
for instance `cma_alloc` or `zs_malloc`.
Most of the memory allocation APIs use GFP flags to express how that
memory should be allocated. The GFP acronym stands for "get free
pages", the underlying memory allocation function.
Diversity of the allocation APIs combined with the numerous GFP flags
makes the question "How should I allocate memory?" not that easy to
answer, although very likely you should use
::
kzalloc(<size>, GFP_KERNEL);
Of course there are cases when other allocation APIs and different GFP
flags must be used.
Get Free Page flags
===================
The GFP flags control the allocators behavior. They tell what memory
zones can be used, how hard the allocator should try to find free
memory, whether the memory can be accessed by the userspace etc. The
:ref:`Documentation/core-api/mm-api.rst <mm-api-gfp-flags>` provides
reference documentation for the GFP flags and their combinations and
here we briefly outline their recommended usage:
* Most of the time ``GFP_KERNEL`` is what you need. Memory for the
kernel data structures, DMAable memory, inode cache, all these and
many other allocations types can use ``GFP_KERNEL``. Note, that
using ``GFP_KERNEL`` implies ``GFP_RECLAIM``, which means that
direct reclaim may be triggered under memory pressure; the calling
context must be allowed to sleep.
* If the allocation is performed from an atomic context, e.g interrupt
handler, use ``GFP_NOWAIT``. This flag prevents direct reclaim and
IO or filesystem operations. Consequently, under memory pressure
``GFP_NOWAIT`` allocation is likely to fail. Allocations which
have a reasonable fallback should be using ``GFP_NOWARN``.
* If you think that accessing memory reserves is justified and the kernel
will be stressed unless allocation succeeds, you may use ``GFP_ATOMIC``.
* Untrusted allocations triggered from userspace should be a subject
of kmem accounting and must have ``__GFP_ACCOUNT`` bit set. There
is the handy ``GFP_KERNEL_ACCOUNT`` shortcut for ``GFP_KERNEL``
allocations that should be accounted.
* Userspace allocations should use either of the ``GFP_USER``,
``GFP_HIGHUSER`` or ``GFP_HIGHUSER_MOVABLE`` flags. The longer
the flag name the less restrictive it is.
``GFP_HIGHUSER_MOVABLE`` does not require that allocated memory
will be directly accessible by the kernel and implies that the
data is movable.
``GFP_HIGHUSER`` means that the allocated memory is not movable,
but it is not required to be directly accessible by the kernel. An
example may be a hardware allocation that maps data directly into
userspace but has no addressing limitations.
``GFP_USER`` means that the allocated memory is not movable and it
must be directly accessible by the kernel.
You may notice that quite a few allocations in the existing code
specify ``GFP_NOIO`` or ``GFP_NOFS``. Historically, they were used to
prevent recursion deadlocks caused by direct memory reclaim calling
back into the FS or IO paths and blocking on already held
resources. Since 4.12 the preferred way to address this issue is to
use new scope APIs described in
:ref:`Documentation/core-api/gfp_mask-from-fs-io.rst <gfp_mask_from_fs_io>`.
Other legacy GFP flags are ``GFP_DMA`` and ``GFP_DMA32``. They are
used to ensure that the allocated memory is accessible by hardware
with limited addressing capabilities. So unless you are writing a
driver for a device with such restrictions, avoid using these flags.
And even with hardware with restrictions it is preferable to use
`dma_alloc*` APIs.
Selecting memory allocator
==========================
The most straightforward way to allocate memory is to use a function
from the :c:func:`kmalloc` family. And, to be on the safe size it's
best to use routines that set memory to zero, like
:c:func:`kzalloc`. If you need to allocate memory for an array, there
are :c:func:`kmalloc_array` and :c:func:`kcalloc` helpers.
The maximal size of a chunk that can be allocated with `kmalloc` is
limited. The actual limit depends on the hardware and the kernel
configuration, but it is a good practice to use `kmalloc` for objects
smaller than page size.
The address of a chunk allocated with `kmalloc` is aligned to at least
ARCH_KMALLOC_MINALIGN bytes. For sizes which are a power of two, the
alignment is also guaranteed to be at least the respective size.
For large allocations you can use :c:func:`vmalloc` and
:c:func:`vzalloc`, or directly request pages from the page
allocator. The memory allocated by `vmalloc` and related functions is
not physically contiguous.
If you are not sure whether the allocation size is too large for
`kmalloc`, it is possible to use :c:func:`kvmalloc` and its
derivatives. It will try to allocate memory with `kmalloc` and if the
allocation fails it will be retried with `vmalloc`. There are
restrictions on which GFP flags can be used with `kvmalloc`; please
see :c:func:`kvmalloc_node` reference documentation. Note that
`kvmalloc` may return memory that is not physically contiguous.
If you need to allocate many identical objects you can use the slab
cache allocator. The cache should be set up with
:c:func:`kmem_cache_create` or :c:func:`kmem_cache_create_usercopy`
before it can be used. The second function should be used if a part of
the cache might be copied to the userspace. After the cache is
created :c:func:`kmem_cache_alloc` and its convenience wrappers can
allocate memory from that cache.
When the allocated memory is no longer needed it must be freed. You
can use :c:func:`kvfree` for the memory allocated with `kmalloc`,
`vmalloc` and `kvmalloc`. The slab caches should be freed with
:c:func:`kmem_cache_free`. And don't forget to destroy the cache with
:c:func:`kmem_cache_destroy`.