Sync pre-release CPython 3.15 branch from GitHub (2026-05-31)

Summary:
Imported python/cpython `3.15.0b1+` from upstream rev [`9a39343`](https://www.github.com/python/cpython/commit/9a393438a7e30f4909c288a90f7637f4ce68e21a) (committed 2026-05-31 04:22:11+00:00).

# Commit Info

- Base: (`3.15.0b1+`) - [`2f91315`](https://www.github.com/python/cpython/commit/2f9131575b611dfc749242e8bbc6805bbc14683b) (commit date: 2026-05-29 21:48:10+00:00)
- Imported: (`3.15.0b1+`) - [`9a39343`](https://www.github.com/python/cpython/commit/9a393438a7e30f4909c288a90f7637f4ce68e21a) (commit date: 2026-05-31 04:22:11+00:00)

# Noteworthy file changes

- Low-signal files (1 added) (NEWS.d, docs, .github)

Complete list of added/removed files: https://www.internalfb.com/intern/everpaste/?color=0&handle=GAKHUCNn0F9DXbAGAKlYOuSQtZMpbr0LAAAz

Differential Revision: D106969527

fbshipit-source-id: 20bafe4b6cea8e52a349efa4b8446fa4771356d9

Author: generatedunixname1734921407115435

Doc/howto/free-threading-python.rst

@@ -165,3 +165,132 @@ to false.  If the flag is true then the :class:`warnings.catch_warnings`
 context manager uses a context variable for warning filters.  If the flag is
 false then :class:`~warnings.catch_warnings` modifies the global filters list,
 which is not thread-safe.  See the :mod:`warnings` module for more details.
+
+
+Increased memory usage
+----------------------
+
+The free-threaded build will typically use more memory compared to the default
+build.  There are multiple reasons for this, mostly due to design decisions.
+
+
+All interned strings are immortal
+^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
+
+For modern Python versions (since version 2.3), interning a string (e.g. with
+:func:`sys.intern`) does not cause it to become immortal.  Instead, if the last
+reference to that string disappears, it will be removed from the interned
+string table.  This is not the case for the free-threaded build and any interned
+string will become immortal, surviving until interpreter shutdown.
+
+
+Non-GC objects have a larger object header
+^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
+
+The free-threaded build uses a different :c:type:`PyObject` structure.  Instead
+of having the GC related information allocated before the :c:type:`PyObject`
+structure, like in the default build, the GC related info is part of the normal
+object header.  For example, on the AMD64 platform, ``None`` uses 32 bytes on
+the free-threaded build vs 16 bytes for the default build.  GC objects (such as
+dicts and lists) are the same size for both builds since the free-threaded
+build does not use additional space for the GC info.
+
+
+QSBR can delay freeing of memory
+^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
+
+In order to safely implement lock-free data structures, a safe memory
+reclamation (SMR) scheme is used, known as quiescent state-based reclamation
+(QSBR).  This means that the memory backing data structures allowing lock-free
+access will use QSBR, which defers the free operation, rather than immediately
+freeing the memory.  Two examples of these data structures are the list object
+and the dictionary keys object.  See ``InternalDocs/qsbr.md`` in the CPython
+source tree for more details on how QSBR is implemented.  Running
+:func:`gc.collect` should cause all memory being held by QSBR to be actually
+freed.  Note that even when QSBR frees the memory, the underlying memory
+allocator may not immediately return that memory to the OS and so the resident
+set size (RSS) of the process might not decrease.
+
+
+mimalloc allocator vs pymalloc
+^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
+
+The default build will normally use the "pymalloc" memory allocator for small
+allocations (512 bytes or smaller).  The free-threaded build does not use
+pymalloc and allocates all Python objects using the "mimalloc" allocator.  The
+pymalloc allocator has the following properties that help keep memory usage
+low: small per-allocated-block overhead, effective memory fragmentation
+prevention, and quick return of free memory to the operating system.  The
+mimalloc allocator does quite well in these respects as well but can have some
+more overhead.
+
+In the free-threaded build, mimalloc manages memory in a number of separate
+heaps (currently four).  For example, all GC supporting objects are allocated
+from their own heap.  Using separate heaps means that free memory in one heap
+cannot be used for an allocation that uses another heap.  Also, some heaps are
+configured to use QSBR (quiescent-state based reclamation) when freeing the
+memory that backs up the heap (known as "pages" in mimalloc terminology).  The
+use of QSBR creates a delay between all memory blocks for a page being freed
+and the memory page being released, either for new allocations or back to the
+OS.
+
+The mimalloc allocator also defers returning freed memory back to the OS.  You
+can reduce that delay by setting the environment variable
+:envvar:`!MIMALLOC_PURGE_DELAY` to ``0``.  Note that this will likely reduce
+the performance of the allocator.
+
+
+Free-threaded reference counting can cause objects to live longer
+^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
+
+In the default build, when an object's reference count reaches zero, it is
+normally deallocated.  The free-threaded build uses "biased reference
+counting", with a fast-path for objects "owned" by the current thread and a
+slow path for other objects.  See :pep:`703` for additional details.  Any time
+an object's reference count ends up in a "queued" state, deallocation can be
+deferred.  The queued state is cleared from the "eval breaker" section of the
+bytecode evaluator.
+
+The free-threaded build also allows a different mode of reference counting,
+known as "deferred reference counting".  This mode is enabled by setting a flag
+on a per-object basis.  Deferred reference counting is enabled for the
+following types:
+
+* module objects
+* module top-level functions
+* class methods defined in the class scope
+* descriptor objects
+* thread-local objects, created by :class:`threading.local`
+
+When deferred reference counting is enabled, references from Python function
+stacks are not added to the reference count.  This scheme reduces the overhead
+of reference counting, especially for objects used from multiple threads.
+Because the stack references are not counted, objects with deferred reference
+counting are not immediately freed when their internal reference count goes to
+zero.  Instead, they are examined by the next GC run and, if no stack
+references to them are found, they are freed.  This means these objects are
+freed by the GC and not when their reference count goes to zero, as is typical.
+
+
+Per-thread reference counting can delay freeing objects
+^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
+
+To avoid contention on the reference count fields of frequently shared
+objects, the free-threaded build also uses "per-thread reference counting"
+for a few selected object types.  Rather than updating a single shared
+reference count, each thread maintains its own local reference count array,
+indexed by a unique id assigned to the object.  The true reference count is
+only computed by summing the per-thread counts when the object's local
+count drops to zero.  Per-thread reference counting is currently used for:
+
+* heap type objects (classes created in Python)
+* code objects
+* the ``__dict__`` of module objects
+
+Because the per-thread counts must be merged back to the object before it
+can be deallocated, objects using per-thread reference counting are
+typically freed later than they would be in the default build.  In
+particular, such an object is usually not freed until the thread that
+referenced it reaches a safe point (for example, in the "eval breaker"
+section of the bytecode evaluator) or exits.  Running :func:`gc.collect`
+will merge the per-thread counts and allow these objects to be freed.

Doc/library/inspect.rst

@@ -1614,10 +1614,11 @@ properties, will be invoked and :meth:`~object.__getattr__` and
 may be called.
 
 For cases where you want passive introspection, like documentation tools, this
-can be inconvenient. :func:`getattr_static` has the same signature as :func:`getattr`
+can be inconvenient. :func:`getattr_static` has a similar signature as :func:`getattr`
 but avoids executing code when it fetches attributes.
 
-.. function:: getattr_static(obj, attr, default=None)
+.. function:: getattr_static(obj, attr)
+              getattr_static(obj, attr, default)
 
    Retrieve attributes without triggering dynamic lookup via the
    descriptor protocol, :meth:`~object.__getattr__`

Doc/library/multiprocessing.rst

@@ -100,10 +100,10 @@ To show the individual process IDs involved, here is an expanded example::
 For an explanation of why the ``if __name__ == '__main__'`` part is
 necessary, see :ref:`multiprocessing-programming`.
 
-The arguments to :class:`Process` usually need to be unpickleable from within
-the child process. If you tried typing the above example directly into a REPL it
-could lead to an :exc:`AttributeError` in the child process trying to locate the
-*f* function in the ``__main__`` module.
+The arguments to :class:`Process` usually need to be picklable so they can be
+passed to the child process. If you tried typing the above example directly
+into a REPL it could lead to an :exc:`AttributeError` in the child process
+trying to locate the *f* function in the ``__main__`` module.
 
 
 .. _multiprocessing-start-methods:

Doc/library/urllib.request.rst

@@ -1051,7 +1051,7 @@ AbstractBasicAuthHandler Objects
    *headers* should be the error headers.
 
    *host* is either an authority (e.g. ``"python.org"``) or a URL containing an
-   authority component (e.g. ``"http://python.org/"``). In either case, the
+   authority component (e.g. ``"https://python.org/"``). In either case, the
    authority must not contain a userinfo component (so, ``"python.org"`` and
    ``"python.org:80"`` are fine, ``"joe:password@python.org"`` is not).
 
@@ -1247,10 +1247,14 @@ This example gets the python.org main page and displays the first 300 bytes of
 it::
 
    >>> import urllib.request
-   >>> with urllib.request.urlopen('http://www.python.org/') as f:
-   ...     print(f.read(300))
-   ...
-   b'<!doctype html>\n<!--[if lt IE 7]>   <html class="no-js ie6 lt-ie7 lt-ie8 lt-ie9">   <![endif]-->\n<!--[if IE 7]>      <html class="no-js ie7 lt-ie8 lt-ie9">          <![endif]-->\n<!--[if IE 8]>      <html class="no-js ie8 lt-ie9">
+   >>> with urllib.request.urlopen('https://www.python.org/') as f:
+   ...     # The response may be compressed (for example, 'gzip').
+   ...     print(f.headers.get('Content-Encoding'))
+   ...     data = f.read()
+   ...     if f.headers.get('Content-Encoding') == 'gzip':
+   ...         import gzip
+   ...         data = gzip.decompress(data)
+   ...     print(data[:300].decode('utf-8', errors='replace'))
 
 Note that urlopen returns a bytes object.  This is because there is no way
 for urlopen to automatically determine the encoding of the byte stream
@@ -1267,26 +1271,30 @@ For additional information, see the W3C document: https://www.w3.org/Internation
 As the python.org website uses *utf-8* encoding as specified in its meta tag, we
 will use the same for decoding the bytes object::
 
-   >>> with urllib.request.urlopen('http://www.python.org/') as f:
-   ...     print(f.read(100).decode('utf-8'))
+   >>> with urllib.request.urlopen('https://www.python.org/') as f:
+   ...     # Check for compression and decode appropriately.
+   ...     enc = f.headers.get('Content-Encoding')
+   ...     data = f.read()
+   ...     if enc == 'gzip':
+   ...         import gzip
+   ...         data = gzip.decompress(data)
+   ...     print(data[:100].decode('utf-8', errors='replace'))
    ...
-   <!doctype html>
-   <!--[if lt IE 7]>   <html class="no-js ie6 lt-ie7 lt-ie8 lt-ie9">   <![endif]-->
-   <!-
 
 It is also possible to achieve the same result without using the
 :term:`context manager` approach::
 
    >>> import urllib.request
-   >>> f = urllib.request.urlopen('http://www.python.org/')
+   >>> f = urllib.request.urlopen('https://www.python.org/')
    >>> try:
-   ...     print(f.read(100).decode('utf-8'))
+   ...     enc = f.headers.get('Content-Encoding')
+   ...     data = f.read()
+   ...     if enc == 'gzip':
+   ...         import gzip
+   ...         data = gzip.decompress(data)
+   ...     print(data[:100].decode('utf-8', errors='replace'))
    ... finally:
    ...     f.close()
-   ...
-   <!doctype html>
-   <!--[if lt IE 7]>   <html class="no-js ie6 lt-ie7 lt-ie8 lt-ie9">   <![endif]-->
-   <!--
 
 In the following example, we are sending a data-stream to the stdin of a CGI
 and reading the data it returns to us. Note that this example will only work
@@ -1357,7 +1365,7 @@ Use the *headers* argument to the :class:`Request` constructor, or::
 
    import urllib.request
    req = urllib.request.Request('http://www.example.com/')
-   req.add_header('Referer', 'http://www.python.org/')
+   req.add_header('Referer', 'https://www.python.org/')
    # Customize the default User-Agent header value:
    req.add_header('User-Agent', 'urllib-example/0.1 (Contact: . . .)')
    with urllib.request.urlopen(req) as f:
@@ -1386,7 +1394,7 @@ containing parameters::
    >>> import urllib.request
    >>> import urllib.parse
    >>> params = urllib.parse.urlencode({'spam': 1, 'eggs': 2, 'bacon': 0})
-   >>> url = "http://www.musi-cal.com/cgi-bin/query?%s" % params
+   >>> url = "https://www.python.org/?%s" % params
    >>> with urllib.request.urlopen(url) as f:
    ...     print(f.read().decode('utf-8'))
    ...
@@ -1398,7 +1406,7 @@ from urlencode is encoded to bytes before it is sent to urlopen as data::
    >>> import urllib.parse
    >>> data = urllib.parse.urlencode({'spam': 1, 'eggs': 2, 'bacon': 0})
    >>> data = data.encode('ascii')
-   >>> with urllib.request.urlopen("http://requestb.in/xrbl82xr", data) as f:
+   >>> with urllib.request.urlopen("https://httpbin.org/post", data) as f:
    ...     print(f.read().decode('utf-8'))
    ...
 
@@ -1408,15 +1416,15 @@ environment settings::
    >>> import urllib.request
    >>> proxies = {'http': 'http://proxy.example.com:8080/'}
    >>> opener = urllib.request.build_opener(urllib.request.ProxyHandler(proxies))
-   >>> with opener.open("http://www.python.org") as f:
+   >>> with opener.open("https://www.python.org") as f:
    ...     f.read().decode('utf-8')
    ...
 
 The following example uses no proxies at all, overriding environment settings::
 
    >>> import urllib.request
-   >>> opener = urllib.request.build_opener(urllib.request.ProxyHandler({}}))
-   >>> with opener.open("http://www.python.org/") as f:
+   >>> opener = urllib.request.build_opener(urllib.request.ProxyHandler({}))
+   >>> with opener.open("https://www.python.org/") as f:
    ...     f.read().decode('utf-8')
    ...
 
@@ -1449,7 +1457,7 @@ some point in the future.
    The following example illustrates the most common usage scenario::
 
       >>> import urllib.request
-      >>> local_filename, headers = urllib.request.urlretrieve('http://python.org/')
+      >>> local_filename, headers = urllib.request.urlretrieve('https://python.org/')
       >>> html = open(local_filename)
       >>> html.close()
 

Doc/using/cmdline.rst

@@ -50,8 +50,8 @@ additional methods of invocation:
 * When called with ``-c command``, it executes the Python statement(s) given as
   *command*.  Here *command* may contain multiple statements separated by
   newlines.
-* When called with ``-m module-name``, the given module is located on the
-  Python module path and executed as a script.
+* When called with ``-m module-name``, the given module is located using the standard
+  import mechanism and executed as a script.
 
 In non-interactive mode, the entire input is parsed before it is executed.
 
@@ -78,8 +78,8 @@ source.
 
 .. option:: -m <module-name>
 
-   Search :data:`sys.path` for the named module and execute its contents as
-   the :mod:`__main__` module.
+   Locate the module using the standard import mechanism and execute its contents
+   as the :mod:`__main__` module.
 
    Since the argument is a *module* name, you must not give a file extension
    (``.py``).  The module name should be a valid absolute Python module name, but

Misc/NEWS.d/next/Library/2026-05-25-17-00-00.gh-issue-150406.jF3g63.rst

@@ -0,0 +1,3 @@
+Fix a possible crash occurring during :mod:`socket` module initialization
+when the system is out of memory on platforms without a reentrant
+``gethostbyname``.

Modules/socketmodule.c

@@ -9285,6 +9285,9 @@ socket_exec(PyObject *m)
     /* Initialize gethostbyname lock */
 #if defined(USE_GETHOSTBYNAME_LOCK)
     netdb_lock = PyThread_allocate_lock();
+    if (netdb_lock == NULL) {
+        goto error;
+    }
 #endif
 
 #ifdef MS_WINDOWS