python-peps/pep-0445.txt

PEP: 445
Title: Add new APIs to customize memory allocators
Version: $Revision$
Last-Modified: $Date$
Author: Victor Stinner <victor.stinner@gmail.com>
Status: Draft
Type: Standards Track
Content-Type: text/x-rst
Created: 15-june-2013
Python-Version: 3.4

Abstract
========

Add new APIs to customize memory allocators.


Rationale
=========

Use cases:

* Application embedding Python may want to isolate Python memory from the
  memory of the application, or may want to different memory allocator
  optimized for its Python usage
* Python running on embedded devices with low memory and slow CPU.
  A custom memory allocator may be required to use efficiently the memory
  and/or to be able to use all memory of the device.
* Debug tool to:

  - track memory leaks
  - get the Python filename and line number where an object was allocated
  - detect buffer underflow, buffer overflow and detect misuse of Python
    allocator APIs (builtin Python debug hooks)
  - force allocation to fail to test handling of ``MemoryError`` exception

API:

* Setup a custom memory allocator for all memory allocated by Python
* Hook memory allocator functions to call extra code before and/or after
  the underlying allocator function


Proposal
========

API changes
-----------

* Add a new ``PyMemAllocators`` structure

* Add new GIL-free memory allocator functions:

  - ``void* PyMem_RawMalloc(size_t size)``
  - ``void* PyMem_RawRealloc(void *ptr, size_t new_size)``
  - ``void PyMem_RawFree(void *ptr)``

* Add new functions to get and set memory allocators:

  - ``void PyMem_GetRawAllocators(PyMemAllocators *allocators)``
  - ``void PyMem_SetRawAllocators(PyMemAllocators *allocators)``
  - ``void PyMem_GetAllocators(PyMemAllocators *allocators)``
  - ``void PyMem_SetAllocators(PyMemAllocators *allocators)``
  - ``void PyObject_GetAllocators(PyMemAllocators *allocators)``
  - ``void PyObject_SetAllocators(PyMemAllocators *allocators)``
  - ``void _PyObject_GetArenaAllocators(void **ctx_p, void* (**malloc_p) (void *ctx, size_t size), void (**free_p) (void *ctx, void *ptr, size_t size))``
  - ``void _PyObject_SetArenaAllocators(void *ctx, void* (*malloc) (void *ctx, size_t size), void (*free) (void *ctx, void *ptr, size_t size))``

* Add a new function to setup Python builtin debug hooks when memory
  allocators are replaced:

  - ``void PyMem_SetupDebugHooks(void)``


Use these new APIs
------------------

* ``PyMem_Malloc()`` and ``PyMem_Realloc()`` always call ``malloc()`` and
  ``realloc()``, instead of calling ``PyObject_Malloc()`` and
  ``PyObject_Realloc()`` in debug mode

* ``PyObject_Malloc()`` falls back on ``PyMem_Malloc()`` instead of
  ``malloc()`` if size is greater or equal than ``SMALL_REQUEST_THRESHOLD``
  (512 bytes), and ``PyObject_Realloc()`` falls back on ``PyMem_Realloc()``
  instead of ``realloc()``

* Replace direct calls to ``malloc()`` with ``PyMem_Malloc()``, or
  ``PyMem_RawMalloc()`` if the GIL is not held

* Configure external libraries like zlib or OpenSSL to allocate memory using
  ``PyMem_RawMalloc()``


Examples
========

Use case 1: Replace Memory Allocators, keep pymalloc
----------------------------------------------------

Setup your custom memory allocators, keeping pymalloc. Dummy example wasting 2
bytes per allocation, and 10 bytes per arena::

    #include <stdlib.h>

    int alloc_padding = 2;
    int arena_padding = 10;

    void* my_malloc(void *ctx, size_t size)
    {
        int padding = *(int *)ctx;
        return malloc(size + padding);
    }

    void* my_realloc(void *ctx, void *ptr, size_t new_size)
    {
        int padding = *(int *)ctx;
        return realloc(ptr, new_size + padding);
    }

    void my_free(void *ctx, void *ptr)
    {
        free(ptr);
    }

    void* my_alloc_arena(void *ctx, size_t size)
    {
        int padding = *(int *)ctx;
        return malloc(size + padding);
    }

    void my_free_arena(void *ctx, void *ptr, size_t size)
    {
        free(ptr);
    }

    void setup_custom_allocators(void)
    {
        PyMemAllocators alloc;

        alloc.ctx = &alloc_padding;
        alloc.malloc = my_malloc;
        alloc.realloc = my_realloc;
        alloc.free = my_free;

        PyMem_SetRawAllocators(&alloc);
        PyMem_SetAllocators(&alloc);

        _PyObject_SetArenaAllocators(&arena_padding,
                                     my_alloc_arena, my_free_arena);

        PyMem_SetupDebugHooks();
    }

.. warning::
   Remove the call ``PyMem_SetRawAllocators(&alloc)`` if the new allocators
   are not thread-safe.


Use case 2: Replace Memory Allocators, overriding pymalloc
----------------------------------------------------------

If your allocator is optimized for allocation of small objects (less than 512
bytes) with a short liftime, you can replace override pymalloc (replace
``PyObject_Malloc()``).

Dummy Example wasting 2 bytes per allocation::

    #include <stdlib.h>

    int padding = 2;

    void* my_malloc(void *ctx, size_t size)
    {
        int padding = *(int *)ctx;
        return malloc(size + padding);
    }

    void* my_realloc(void *ctx, void *ptr, size_t new_size)
    {
        int padding = *(int *)ctx;
        return realloc(ptr, new_size + padding);
    }

    void my_free(void *ctx, void *ptr)
    {
        free(ptr);
    }

    void setup_custom_allocators(void)
    {
        PyMemAllocators alloc;
        alloc.ctx = &padding;
        alloc.malloc = my_malloc;
        alloc.realloc = my_realloc;
        alloc.free = my_free;

        PyMem_SetRawAllocators(&alloc);
        PyMem_SetAllocators(&alloc);
        PyObject_SetAllocators(&alloc);

        PyMem_SetupDebugHooks();
    }

.. warning::
   Remove the call ``PyMem_SetRawAllocators(&alloc)`` if the new allocators
   are not thread-safe.



Use case 3: Setup Allocator Hooks
---------------------------------

Example to setup hooks on memory allocators::

    struct {
        PyMemAllocators pymem;
        PyMemAllocators pymem_raw;
        PyMemAllocators pyobj;
        /* ... */
    } hook;

    static void* hook_malloc(void *ctx, size_t size)
    {
        PyMemAllocators *alloc = (PyMemAllocators *)ctx;
        /* ... */
        ptr = alloc->malloc(alloc->ctx, size);
        /* ... */
        return ptr;
    }

    static void* hook_realloc(void *ctx, void *ptr, size_t new_size)
    {
        PyMemAllocators *alloc = (PyMemAllocators *)ctx;
        void *ptr2;
        /* ... */
        ptr2 = alloc->realloc(alloc->ctx, ptr, new_size);
        /* ... */
        return ptr2;
    }

    static void hook_free(void *ctx, void *ptr)
    {
        PyMemAllocators *alloc = (PyMemAllocators *)ctx;
        /* ... */
        alloc->free(alloc->ctx, ptr);
        /* ... */
    }

    void setup_hooks(void)
    {
        PyMemAllocators alloc;
        static int registered = 0;

        if (registered)
            return;
        registered = 1;

        alloc.malloc = hook_malloc;
        alloc.realloc = hook_realloc;
        alloc.free = hook_free;

        PyMem_GetRawAllocators(&hook.pymem_raw);
        alloc.ctx = &hook.pymem_raw;
        PyMem_SetRawAllocators(&alloc);

        PyMem_GetAllocators(&hook.pymem);
        alloc.ctx = &hook.pymem;
        PyMem_SetAllocators(&alloc);

        PyObject_GetAllocators(&hook.pyobj);
        alloc.ctx = &hook.pyobj;
        PyObject_SetAllocators(&alloc);
    }

.. warning::
   Remove the call ``PyMem_SetRawAllocators(&alloc)`` if hooks are not
   thread-safe.

.. note::
   ``PyMem_SetupDebugHooks()`` does not need to be called: Python debug hooks
   are installed automatically at startup.


Performances
============

The `Python benchmarks suite <http://hg.python.org/benchmarks>`_ (-b 2n3): some
tests are 1.04x faster, some tests are 1.04 slower, significant is between 115
and -191. I don't understand these output, but I guess that the overhead cannot
be seen with such test.

pybench: "+0.1%" (diff between -4.9% and +5.6%).

The full output is attached to the issue #3329.


Alternatives
============

Only one get and one set function
---------------------------------

Replace the 6 functions:

* ``PyMem_GetRawAllocators()``
* ``PyMem_GetAllocators()``
* ``PyObject_GetAllocators()``
* ``PyMem_SetRawAllocators(allocators)``
* ``PyMem_SetAllocators(allocators)``
* ``PyObject_SetAllocators(allocators)``

with 2 functions with an additional *domain* argument:

* ``Py_GetAllocators(domain)``
* ``Py_SetAllocators(domain, allocators)``

where domain is one of these values:

* ``PYALLOC_PYMEM``
* ``PYALLOC_PYMEM_RAW``
* ``PYALLOC_PYOBJECT``


Add a new PYDEBUGMALLOC environment variable
--------------------------------------------

To be able to use Python builtin debug hooks even when a custom memory
allocator is set, an environment variable ``PYDEBUGMALLOC`` can be added to
setup these debug function hooks, instead of adding the new function
``PyMem_SetupDebugHooks()``. If the environment variable is present,
``PyMem_SetRawAllocators()``, ``PyMem_SetAllocators()`` and
``PyObject_SetAllocators()`` will reinstall automatically the hook on top of
the new allocator.

An new environment variable would make the Python initialization even more
complex. The `PEP 432 <http://www.python.org/dev/peps/pep-0432/>`_ tries to
simply the CPython startup sequence.


Use macros to get customizable allocators
-----------------------------------------

To have no overhead in the default configuration, customizable allocators would
be an optional feature enabled by a configuration option or by macros.


Pass the C filename and line number
-----------------------------------

Use C macros using ``__FILE__`` and ``__LINE__`` to get the C filename
and line number of a memory allocation.

Passing a filename and a line number to each allocator makes the API more
complex: pass 3 new arguments instead of just a context argument, to each
allocator function. GC allocator functions should also be patched,
``_PyObject_GC_Malloc()`` is used in many C functions for example. Such changes
add too much complexity, for a little gain.


No context argument
-------------------

Simplify the signature of allocator functions, remove the context argument:

* ``void* malloc(size_t size)``
* ``void* realloc(void *ptr, size_t new_size)``
* ``void free(void *ptr)``

The context is a convenient way to reuse the same allocator for different APIs
(ex: PyMem and PyObject).


PyMem_Malloc() GIL-free
-----------------------

There is no real reason to require the GIL when calling ``PyMem_Malloc()``.

Allowing to call ``PyMem_Malloc()`` without holding the GIL might break
applications which setup their own allocator or their allocator hooks. Holding
the GIL is very convinient to develop a custom allocator or a hook (no need to
care of other threads, no need to handle mutexes, etc.).


Don't add PyMem_RawMalloc()
---------------------------

Replace ``malloc()`` with ``PyMem_Malloc()``, but if the GIL is not held: keep
``malloc()`` unchanged.

The ``PyMem_Malloc()`` is sometimes already misused. For example, the
``main()`` and ``Py_Main()`` functions of Python call ``PyMem_Malloc()``
whereas the GIL do not exist yet. In this case, ``PyMem_Malloc()`` should
be replaced with ``malloc()``.

If an hook is used to the track memory usage, the ``malloc()`` memory will not
be seen. Remaining ``malloc()`` may allocate a lot of memory and so would be
missed in reports.


CCP API
-------

XXX To be done (Kristján Valur Jónsson) XXX


External libraries
==================

* glib: `g_mem_set_vtable()
  <http://developer.gnome.org/glib/unstable/glib-Memory-Allocation.html#g-mem-set-vtable>`_

See also the `GNU libc: Memory Allocation Hooks
<http://www.gnu.org/software/libc/manual/html_node/Hooks-for-Malloc.html>`_.



Memory allocators
=================

The C standard library provides the well known ``malloc()`` function. Its
implementation depends on the platform and of the C library. The GNU C library
uses a modified ptmalloc2, based on "Doug Lea's Malloc" (dlmalloc). FreeBSD
uses `jemalloc <http://www.canonware.com/jemalloc/>`_. Google provides
tcmalloc which is part of `gperftools <http://code.google.com/p/gperftools/>`_.

``malloc()`` uses two kinds of memory: heap and memory mappings. Memory
mappings are usually used for large allocations (ex: larger than 256 KB),
whereas the heap is used for small allocations.

The heap is handled by ``brk()`` and ``sbrk()`` system calls on Linux, and is
contiguous.  Memory mappings are handled by ``mmap()`` on UNIX and
``VirtualAlloc()`` on Windows, they may be discontiguous.

Releasing a memory mapping gives back immediatly the memory to the system. For
the heap, memory is only given back to the system if it is at the end of the
heap. Otherwise, the memory will only be given back to the system when all the
memory located after the released memory are also released. To allocate memory
in the heap, the allocator tries to reuse free space. If there is no contiguous
space big enough, the heap must be increased, even if we have more free space
than required size.  This issue is called the "memory fragmentation": the
memory usage seen by the system may be much higher than real usage.

CPython has a pymalloc allocator using arenas of 256 KB for allocations smaller
than 512 bytes. This allocator is optimized for small objects with a short
lifetime.

Windows provides a `Low-fragmentation Heap
<http://msdn.microsoft.com/en-us/library/windows/desktop/aa366750%28v=vs.85%29.aspx>`_.

The Linux kernel uses `slab allocation
<http://en.wikipedia.org/wiki/Slab_allocation>`_.

The glib library has a `Memory Slice API
<https://developer.gnome.org/glib/unstable/glib-Memory-Slices.html>`_:
efficient way to allocate groups of equal-sized chunks of memory


Links
=====

CPython issues related to memory allocation:

* `Issue #3329: Add new APIs to customize memory allocators
  <http://bugs.python.org/issue3329>`_
* `Issue #13483: Use VirtualAlloc to allocate memory arenas
  <http://bugs.python.org/issue13483>`_
* `Issue #16742: PyOS_Readline drops GIL and calls PyOS_StdioReadline, which
  isn't thread safe <http://bugs.python.org/issue16742>`_
* `Issue #18203: Replace calls to malloc() with PyMem_Malloc() or PyMem_RawMalloc()
  <http://bugs.python.org/issue18203>`_
* `Issue #18227: Use Python memory allocators in external libraries like zlib
  or OpenSSL <http://bugs.python.org/issue18227>`_

Projects analyzing the memory usage of Python applications:

* `pytracemalloc
  <https://pypi.python.org/pypi/pytracemalloc>`_
* `Meliae: Python Memory Usage Analyzer
  <https://pypi.python.org/pypi/meliae>`_
* `Guppy-PE: umbrella package combining Heapy and GSL
  <http://guppy-pe.sourceforge.net/>`_
* `PySizer (developed for Python 2.4)
  <http://pysizer.8325.org/>`_