Memory Management (2023)

Overview¶

Memory management in Python involves a private heap containing all Pythonobjects and data structures. The management of this private heap is ensuredinternally by the Python memory manager. The Python memory manager hasdifferent components which deal with various dynamic storage management aspects,like sharing, segmentation, preallocation or caching.

At the lowest level, a raw memory allocator ensures that there is enough room inthe private heap for storing all Python-related data by interacting with thememory manager of the operating system. On top of the raw memory allocator,several object-specific allocators operate on the same heap and implementdistinct memory management policies adapted to the peculiarities of every objecttype. For example, integer objects are managed differently within the heap thanstrings, tuples or dictionaries because integers imply different storagerequirements and speed/space tradeoffs. The Python memory manager thus delegatessome of the work to the object-specific allocators, but ensures that the latteroperate within the bounds of the private heap.

It is important to understand that the management of the Python heap isperformed by the interpreter itself and that the user has no control over it,even if they regularly manipulate object pointers to memory blocks inside thatheap. The allocation of heap space for Python objects and other internalbuffers is performed on demand by the Python memory manager through the Python/CAPI functions listed in this document.

To avoid memory corruption, extension writers should never try to operate onPython objects with the functions exported by the C library: malloc(),calloc(), realloc() and free(). This will result in mixedcalls between the C allocator and the Python memory manager with fatalconsequences, because they implement different algorithms and operate ondifferent heaps. However, one may safely allocate and release memory blockswith the C library allocator for individual purposes, as shown in the followingexample:

PyObject *res;char *buf = (char *) malloc(BUFSIZ); /* for I/O */if (buf == NULL) return PyErr_NoMemory();...Do some I/O operation involving buf...res = PyBytes_FromString(buf);free(buf); /* malloc'ed */return res;

In this example, the memory request for the I/O buffer is handled by the Clibrary allocator. The Python memory manager is involved only in the allocationof the bytes object returned as a result.

In most situations, however, it is recommended to allocate memory from thePython heap specifically because the latter is under control of the Pythonmemory manager. For example, this is required when the interpreter is extendedwith new object types written in C. Another reason for using the Python heap isthe desire to inform the Python memory manager about the memory needs of theextension module. Even when the requested memory is used exclusively forinternal, highly specific purposes, delegating all memory requests to the Pythonmemory manager causes the interpreter to have a more accurate image of itsmemory footprint as a whole. Consequently, under certain circumstances, thePython memory manager may or may not trigger appropriate actions, like garbagecollection, memory compaction or other preventive procedures. Note that by usingthe C library allocator as shown in the previous example, the allocated memoryfor the I/O buffer escapes completely the Python memory manager.

Allocator Domains¶

All allocating functions belong to one of three different “domains” (see alsoPyMemAllocatorDomain). These domains represent different allocationstrategies and are optimized for different purposes. The specific details onhow every domain allocates memory or what internal functions each domain callsis considered an implementation detail, but for debugging purposes a simplifiedtable can be found at here. There is no hardrequirement to use the memory returned by the allocation functions belonging toa given domain for only the purposes hinted by that domain (although this is therecommended practice). For example, one could use the memory returned byPyMem_RawMalloc() for allocating Python objects or the memory returnedby PyObject_Malloc() for allocating memory for buffers.

The three allocation domains are:

• Raw domain: intended for allocating memory for general-purpose memorybuffers where the allocation must go to the system allocator or where theallocator can operate without the GIL. The memory is requested directlyto the system.

• “Mem” domain: intended for allocating memory for Python buffers andgeneral-purpose memory buffers where the allocation must be performed withthe GIL held. The memory is taken from the Python private heap.

• Object domain: intended for allocating memory belonging to Python objects. Thememory is taken from the Python private heap.

When freeing memory previously allocated by the allocating functions belonging to agiven domain,the matching specific deallocating functions must be used. For example,PyMem_Free() must be used to free memory allocated using PyMem_Malloc().

Raw Memory Interface¶

The following function sets are wrappers to the system allocator. Thesefunctions are thread-safe, the GIL does notneed to be held.

The default raw memory allocator usesthe following functions: malloc(), calloc(), realloc()and free(); call malloc(1) (or calloc(1, 1)) when requestingzero bytes.

New in version 3.4.

void *PyMem_RawMalloc(size_t n)¶

Allocates n bytes and returns a pointer of type void* to theallocated memory, or NULL if the request fails.

Requesting zero bytes returns a distinct non-NULL pointer if possible, asif PyMem_RawMalloc(1) had been called instead. The memory will not havebeen initialized in any way.

void *PyMem_RawCalloc(size_t nelem, size_t elsize)¶

Allocates nelem elements each whose size in bytes is elsize and returnsa pointer of type void* to the allocated memory, or NULL if therequest fails. The memory is initialized to zeros.

Requesting zero elements or elements of size zero bytes returns a distinctnon-NULL pointer if possible, as if PyMem_RawCalloc(1, 1) had beencalled instead.

New in version 3.5.

void *PyMem_RawRealloc(void *p, size_t n)¶

Resizes the memory block pointed to by p to n bytes. The contents willbe unchanged to the minimum of the old and the new sizes.

If p is NULL, the call is equivalent to PyMem_RawMalloc(n); else ifn is equal to zero, the memory block is resized but is not freed, and thereturned pointer is non-NULL.

Unless p is NULL, it must have been returned by a previous call toPyMem_RawMalloc(), PyMem_RawRealloc() orPyMem_RawCalloc().

If the request fails, PyMem_RawRealloc() returns NULL and premains a valid pointer to the previous memory area.

void PyMem_RawFree(void *p)¶

Frees the memory block pointed to by p, which must have been returned by aprevious call to PyMem_RawMalloc(), PyMem_RawRealloc() orPyMem_RawCalloc(). Otherwise, or if PyMem_RawFree(p) has beencalled before, undefined behavior occurs.

If p is NULL, no operation is performed.

Memory Interface¶

The following function sets, modeled after the ANSI C standard, but specifyingbehavior when requesting zero bytes, are available for allocating and releasingmemory from the Python heap.

The default memory allocator uses thepymalloc memory allocator.

Changed in version 3.6: The default allocator is now pymalloc instead of system malloc().

void *PyMem_Malloc(size_t n)¶

Part of the Stable ABI.

Allocates n bytes and returns a pointer of type void* to theallocated memory, or NULL if the request fails.

Requesting zero bytes returns a distinct non-NULL pointer if possible, asif PyMem_Malloc(1) had been called instead. The memory will not havebeen initialized in any way.

void *PyMem_Calloc(size_t nelem, size_t elsize)¶

Part of the Stable ABI since version 3.7.

Allocates nelem elements each whose size in bytes is elsize and returnsa pointer of type void* to the allocated memory, or NULL if therequest fails. The memory is initialized to zeros.

See Also

Comment résoudre l'erreur de mémoire Python | HackerMidiPython, Memory, and ObjectsHow to Handle the MemoryError in PythonPython Memory Management: The Essential Guide | Scout APM Blog

Requesting zero elements or elements of size zero bytes returns a distinctnon-NULL pointer if possible, as if PyMem_Calloc(1, 1) had been calledinstead.

New in version 3.5.

void *PyMem_Realloc(void *p, size_t n)¶

Part of the Stable ABI.

Resizes the memory block pointed to by p to n bytes. The contents will beunchanged to the minimum of the old and the new sizes.

If p is NULL, the call is equivalent to PyMem_Malloc(n); else if nis equal to zero, the memory block is resized but is not freed, and thereturned pointer is non-NULL.

Unless p is NULL, it must have been returned by a previous call toPyMem_Malloc(), PyMem_Realloc() or PyMem_Calloc().

If the request fails, PyMem_Realloc() returns NULL and p remainsa valid pointer to the previous memory area.

void PyMem_Free(void *p)¶

Part of the Stable ABI.

Frees the memory block pointed to by p, which must have been returned by aprevious call to PyMem_Malloc(), PyMem_Realloc() orPyMem_Calloc(). Otherwise, or if PyMem_Free(p) has been calledbefore, undefined behavior occurs.

If p is NULL, no operation is performed.

The following type-oriented macros are provided for convenience. Note thatTYPE refers to any C type.

PyMem_New(TYPE, n)¶

Same as PyMem_Malloc(), but allocates (n * sizeof(TYPE)) bytes ofmemory. Returns a pointer cast to TYPE*. The memory will not havebeen initialized in any way.

PyMem_Resize(p, TYPE, n)¶

Same as PyMem_Realloc(), but the memory block is resized to (n *sizeof(TYPE)) bytes. Returns a pointer cast to TYPE*. On return,p will be a pointer to the new memory area, or NULL in the event offailure.

This is a C preprocessor macro; p is always reassigned. Save the originalvalue of p to avoid losing memory when handling errors.

void PyMem_Del(void *p)¶

Same as PyMem_Free().

In addition, the following macro sets are provided for calling the Python memoryallocator directly, without involving the C API functions listed above. However,note that their use does not preserve binary compatibility across Pythonversions and is therefore deprecated in extension modules.

• PyMem_MALLOC(size)

• PyMem_NEW(type, size)

• PyMem_REALLOC(ptr, size)

• PyMem_RESIZE(ptr, type, size)

• PyMem_FREE(ptr)

• PyMem_DEL(ptr)

Object allocators¶

The following function sets, modeled after the ANSI C standard, but specifyingbehavior when requesting zero bytes, are available for allocating and releasingmemory from the Python heap.

The default object allocator uses thepymalloc memory allocator.

void *PyObject_Malloc(size_t n)¶

Part of the Stable ABI.

Allocates n bytes and returns a pointer of type void* to theallocated memory, or NULL if the request fails.

Requesting zero bytes returns a distinct non-NULL pointer if possible, asif PyObject_Malloc(1) had been called instead. The memory will not havebeen initialized in any way.

void *PyObject_Calloc(size_t nelem, size_t elsize)¶

Part of the Stable ABI since version 3.7.

Allocates nelem elements each whose size in bytes is elsize and returnsa pointer of type void* to the allocated memory, or NULL if therequest fails. The memory is initialized to zeros.

Requesting zero elements or elements of size zero bytes returns a distinctnon-NULL pointer if possible, as if PyObject_Calloc(1, 1) had been calledinstead.

New in version 3.5.

void *PyObject_Realloc(void *p, size_t n)¶

Part of the Stable ABI.

Resizes the memory block pointed to by p to n bytes. The contents will beunchanged to the minimum of the old and the new sizes.

If p is NULL, the call is equivalent to PyObject_Malloc(n); else if nis equal to zero, the memory block is resized but is not freed, and thereturned pointer is non-NULL.

Unless p is NULL, it must have been returned by a previous call toPyObject_Malloc(), PyObject_Realloc() or PyObject_Calloc().

If the request fails, PyObject_Realloc() returns NULL and p remainsa valid pointer to the previous memory area.

void PyObject_Free(void *p)¶

Part of the Stable ABI.

Frees the memory block pointed to by p, which must have been returned by aprevious call to PyObject_Malloc(), PyObject_Realloc() orPyObject_Calloc(). Otherwise, or if PyObject_Free(p) has been calledbefore, undefined behavior occurs.

If p is NULL, no operation is performed.

Default Memory Allocators¶

Default memory allocators:

Legend:

• Name: value for PYTHONMALLOC environment variable.

• malloc: system allocators from the standard C library, C functions:malloc(), calloc(), realloc() and free().

• pymalloc: pymalloc memory allocator.

• “+ debug”: with debug hooks on the Python memory allocators.

• “Debug build”: Python build in debug mode.

Customize Memory Allocators¶

New in version 3.4.

type PyMemAllocatorEx¶

Structure used to describe a memory block allocator. The structure hasthe following fields:

Field	Meaning
void *ctx	user context passed as first argument
void* malloc(void *ctx, size_t size)	allocate a memory block
void* calloc(void *ctx, size_t nelem, size_t elsize)	allocate a memory block initializedwith zeros
void* realloc(void *ctx, void *ptr, size_t new_size)	allocate or resize a memory block
void free(void *ctx, void *ptr)	free a memory block

Changed in version 3.5: The PyMemAllocator structure was renamed toPyMemAllocatorEx and a new calloc field was added.

type PyMemAllocatorDomain¶

Enum used to identify an allocator domain. Domains:

PYMEM_DOMAIN_RAW¶

Functions:

• PyMem_RawMalloc()

• PyMem_RawRealloc()

• PyMem_RawCalloc()

• PyMem_RawFree()

PYMEM_DOMAIN_MEM¶

Functions:

• PyMem_Malloc(),

• PyMem_Realloc()

• PyMem_Calloc()

• PyMem_Free()

PYMEM_DOMAIN_OBJ¶

Functions:

• PyObject_Malloc()

• PyObject_Realloc()

• PyObject_Calloc()

• PyObject_Free()

void PyMem_GetAllocator(PyMemAllocatorDomain domain, PyMemAllocatorEx *allocator)¶

Get the memory block allocator of the specified domain.

void PyMem_SetAllocator(PyMemAllocatorDomain domain, PyMemAllocatorEx *allocator)¶

Set the memory block allocator of the specified domain.

The new allocator must return a distinct non-NULL pointer when requestingzero bytes.

For the PYMEM_DOMAIN_RAW domain, the allocator must bethread-safe: the GIL is not held when theallocator is called.

If the new allocator is not a hook (does not call the previous allocator),the PyMem_SetupDebugHooks() function must be called to reinstall thedebug hooks on top on the new allocator.

See also PyPreConfig.allocator and Preinitialize Pythonwith PyPreConfig.

Warning

PyMem_SetAllocator() does have the following contract:

• It can be called after Py_PreInitialize() and beforePy_InitializeFromConfig() to install a custom memoryallocator. There are no restrictions over the installed allocatorother than the ones imposed by the domain (for instance, the RawDomain allows the allocator to be called without the GIL held). Seethe section on allocator domains for moreinformation.

• If called after Python has finish initializing (afterPy_InitializeFromConfig() has been called) the allocatormust wrap the existing allocator. Substituting the currentallocator for some other arbitrary one is not supported.

void PyMem_SetupDebugHooks(void)¶

Setup debug hooks in the Python memory allocatorsto detect memory errors.

Debug hooks on the Python memory allocators¶

When Python is built in debug mode, thePyMem_SetupDebugHooks() function is called at the Pythonpreinitialization to setup debug hooks on Python memory allocatorsto detect memory errors.

The PYTHONMALLOC environment variable can be used to install debughooks on a Python compiled in release mode (ex: PYTHONMALLOC=debug).

The PyMem_SetupDebugHooks() function can be used to set debug hooksafter calling PyMem_SetAllocator().

These debug hooks fill dynamically allocated memory blocks with special,recognizable bit patterns. Newly allocated memory is filled with the byte0xCD (PYMEM_CLEANBYTE), freed memory is filled with the byte 0xDD(PYMEM_DEADBYTE). Memory blocks are surrounded by “forbidden bytes”filled with the byte 0xFD (PYMEM_FORBIDDENBYTE). Strings of these bytesare unlikely to be valid addresses, floats, or ASCII strings.

Runtime checks:

• Detect API violations. For example, detect if PyObject_Free() iscalled on a memory block allocated by PyMem_Malloc().

• Detect write before the start of the buffer (buffer underflow).

• Detect write after the end of the buffer (buffer overflow).

• Check that the GIL is held whenallocator functions of PYMEM_DOMAIN_OBJ (ex:PyObject_Malloc()) and PYMEM_DOMAIN_MEM (ex:PyMem_Malloc()) domains are called.

On error, the debug hooks use the tracemalloc module to get thetraceback where a memory block was allocated. The traceback is only displayedif tracemalloc is tracing Python memory allocations and the memory blockwas traced.

Let S = sizeof(size_t). 2*S bytes are added at each end of each blockof N bytes requested. The memory layout is like so, where p represents theaddress returned by a malloc-like or realloc-like function (p[i:j] meansthe slice of bytes from *(p+i) inclusive up to *(p+j) exclusive; notethat the treatment of negative indices differs from a Python slice):

p[-2*S:-S]

Number of bytes originally asked for. This is a size_t, big-endian (easierto read in a memory dump).

p[-S]

API identifier (ASCII character):

• 'r' for PYMEM_DOMAIN_RAW.

• 'm' for PYMEM_DOMAIN_MEM.

• 'o' for PYMEM_DOMAIN_OBJ.

p[-S+1:0]

Copies of PYMEM_FORBIDDENBYTE. Used to catch under- writes and reads.

p[0:N]

The requested memory, filled with copies of PYMEM_CLEANBYTE, used to catchreference to uninitialized memory. When a realloc-like function is calledrequesting a larger memory block, the new excess bytes are also filled withPYMEM_CLEANBYTE. When a free-like function is called, these areoverwritten with PYMEM_DEADBYTE, to catch reference to freed memory. Whena realloc- like function is called requesting a smaller memory block, theexcess old bytes are also filled with PYMEM_DEADBYTE.

p[N:N+S]

Copies of PYMEM_FORBIDDENBYTE. Used to catch over- writes and reads.

p[N+S:N+2*S]

Only used if the PYMEM_DEBUG_SERIALNO macro is defined (not defined bydefault).

A serial number, incremented by 1 on each call to a malloc-like orrealloc-like function. Big-endian size_t. If “bad memory” is detectedlater, the serial number gives an excellent way to set a breakpoint on thenext run, to capture the instant at which this block was passed out. Thestatic function bumpserialno() in obmalloc.c is the only place the serialnumber is incremented, and exists so you can set such a breakpoint easily.

A realloc-like or free-like function first checks that the PYMEM_FORBIDDENBYTEbytes at each end are intact. If they’ve been altered, diagnostic output iswritten to stderr, and the program is aborted via Py_FatalError(). The othermain failure mode is provoking a memory error when a program reads up one ofthe special bit patterns and tries to use it as an address. If you get in adebugger then and look at the object, you’re likely to see that it’s entirelyfilled with PYMEM_DEADBYTE (meaning freed memory is getting used) orPYMEM_CLEANBYTE (meaning uninitialized memory is getting used).

Changed in version 3.6: The PyMem_SetupDebugHooks() function now also works on Pythoncompiled in release mode. On error, the debug hooks now usetracemalloc to get the traceback where a memory block was allocated.The debug hooks now also check if the GIL is held when functions ofPYMEM_DOMAIN_OBJ and PYMEM_DOMAIN_MEM domains arecalled.

Changed in version 3.8: Byte patterns 0xCB (PYMEM_CLEANBYTE), 0xDB (PYMEM_DEADBYTE)and 0xFB (PYMEM_FORBIDDENBYTE) have been replaced with 0xCD,0xDD and 0xFD to use the same values than Windows CRT debugmalloc() and free().

The pymalloc allocator¶

Python has a pymalloc allocator optimized for small objects (smaller or equalto 512 bytes) with a short lifetime. It uses memory mappings called “arenas”with a fixed size of 256 KiB. It falls back to PyMem_RawMalloc() andPyMem_RawRealloc() for allocations larger than 512 bytes.

pymalloc is the default allocator of thePYMEM_DOMAIN_MEM (ex: PyMem_Malloc()) andPYMEM_DOMAIN_OBJ (ex: PyObject_Malloc()) domains.

The arena allocator uses the following functions:

• VirtualAlloc() and VirtualFree() on Windows,

• mmap() and munmap() if available,

• malloc() and free() otherwise.

This allocator is disabled if Python is configured with the--without-pymalloc option. It can also be disabled at runtime usingthe PYTHONMALLOC environment variable (ex: PYTHONMALLOC=malloc).

tracemalloc C API¶

New in version 3.7.

int PyTraceMalloc_Track(unsigned int domain, uintptr_t ptr, size_t size)¶

Track an allocated memory block in the tracemalloc module.

Return 0 on success, return -1 on error (failed to allocate memory tostore the trace). Return -2 if tracemalloc is disabled.

If memory block is already tracked, update the existing trace.

int PyTraceMalloc_Untrack(unsigned int domain, uintptr_t ptr)¶

Untrack an allocated memory block in the tracemalloc module.Do nothing if the block was not tracked.

Return -2 if tracemalloc is disabled, otherwise return 0.

Examples¶

Here is the example from section Overview, rewritten so that theI/O buffer is allocated from the Python heap by using the first function set:

PyObject *res;char *buf = (char *) PyMem_Malloc(BUFSIZ); /* for I/O */if (buf == NULL) return PyErr_NoMemory();/* ...Do some I/O operation involving buf... */res = PyBytes_FromString(buf);PyMem_Free(buf); /* allocated with PyMem_Malloc */return res;

The same code using the type-oriented function set:

PyObject *res;char *buf = PyMem_New(char, BUFSIZ); /* for I/O */if (buf == NULL) return PyErr_NoMemory();/* ...Do some I/O operation involving buf... */res = PyBytes_FromString(buf);PyMem_Del(buf); /* allocated with PyMem_New */return res;

Note that in the two examples above, the buffer is always manipulated viafunctions belonging to the same set. Indeed, it is required to use the samememory API family for a given memory block, so that the risk of mixing differentallocators is reduced to a minimum. The following code sequence contains twoerrors, one of which is labeled as fatal because it mixes two differentallocators operating on different heaps.

char *buf1 = PyMem_New(char, BUFSIZ);char *buf2 = (char *) malloc(BUFSIZ);char *buf3 = (char *) PyMem_Malloc(BUFSIZ);...PyMem_Del(buf3); /* Wrong -- should be PyMem_Free() */free(buf2); /* Right -- allocated via malloc() */free(buf1); /* Fatal -- should be PyMem_Del() */

In addition to the functions aimed at handling raw memory blocks from the Pythonheap, objects in Python are allocated and released with PyObject_New,PyObject_NewVar and PyObject_Del().

These will be explained in the next chapter on defining and implementing newobject types in C.