Merge branch 'main' into chain_next

Author: eendebakpt

.github/workflows/build.yml

@@ -527,6 +527,14 @@ jobs:
       config_hash: ${{ needs.build-context.outputs.config-hash }}
       free-threading: ${{ matrix.free-threading }}
 
+  build-ubsan:
+    name: Undefined behavior sanitizer
+    needs: build-context
+    if: needs.build-context.outputs.run-tests == 'true'
+    uses: ./.github/workflows/reusable-ubsan.yml
+    with:
+      config_hash: ${{ needs.build-context.outputs.config-hash }}
+
   cross-build-linux:
     name: Cross build Linux
     runs-on: ubuntu-latest

.github/workflows/reusable-ubsan.yml

@@ -0,0 +1,74 @@
+name: Reusable Undefined Behavior Sanitizer
+
+on:
+  workflow_call:
+    inputs:
+      config_hash:
+        required: true
+        type: string
+
+env:
+  FORCE_COLOR: 1
+
+jobs:
+  build-ubsan-reusable:
+    name: 'Undefined behavior sanitizer'
+    runs-on: ubuntu-24.04
+    timeout-minutes: 60
+    steps:
+    - uses: actions/checkout@v4
+      with:
+        persist-credentials: false
+    - name: Runner image version
+      run: echo "IMAGE_OS_VERSION=${ImageOS}-${ImageVersion}" >> "$GITHUB_ENV"
+    - name: Restore config.cache
+      uses: actions/cache@v4
+      with:
+        path: config.cache
+        key: ${{ github.job }}-${{ env.IMAGE_OS_VERSION }}-${{ inputs.config_hash }}
+    - name: Install dependencies
+      run: |
+        sudo ./.github/workflows/posix-deps-apt.sh
+        # Install clang-20
+        wget https://apt.llvm.org/llvm.sh
+        chmod +x llvm.sh
+        sudo ./llvm.sh 20
+    - name: UBSAN option setup
+      run: |
+        echo "UBSAN_OPTIONS=halt_on_error=1:abort_on_error=1:print_summary=1:print_stacktrace=1" >> "$GITHUB_ENV"
+        echo "CC=clang" >> "$GITHUB_ENV"
+        echo "CXX=clang++" >> "$GITHUB_ENV"
+    - name: Add ccache to PATH
+      run: |
+        echo "PATH=/usr/lib/ccache:$PATH" >> "$GITHUB_ENV"
+    - name: Configure ccache action
+      uses: hendrikmuhs/ccache-action@v1.2
+      with:
+        save: ${{ github.event_name == 'push' }}
+        max-size: "200M"
+    - name: Configure CPython
+      run: >-
+        ./configure
+        --config-cache
+        --with-undefined-behavior-sanitizer
+        --with-pydebug
+    - name: Set up UBSAN log after configuration
+      run: |
+        echo "UBSAN_OPTIONS=${UBSAN_OPTIONS}:log_path=${GITHUB_WORKSPACE}/ubsan_log" >> "$GITHUB_ENV"
+    - name: Build CPython
+      run: make -j4
+    - name: Display build info
+      run: make pythoninfo
+    - name: Tests
+      run: ./python -m test -j4
+    - name: Display UBSAN logs
+      if: always()
+      run: find "${GITHUB_WORKSPACE}" -name 'ubsan_log.*' | xargs head -n 1000
+    - name: Archive UBSAN logs
+      if: always()
+      uses: actions/upload-artifact@v4
+      with:
+        name: >-
+          ubsan-logs
+        path: ubsan_log.*
+        if-no-files-found: ignore

Doc/c-api/type.rst

@@ -282,6 +282,10 @@ Type Objects
    and other places where a method's defining class cannot be passed using the
    :c:type:`PyCMethod` calling convention.
 
+   The returned reference is :term:`borrowed <borrowed reference>` from *type*,
+   and will be valid as long as you hold a reference to *type*.
+   Do not release it with :c:func:`Py_DECREF` or similar.
+
    .. versionadded:: 3.11
 
 .. c:function:: int PyType_GetBaseByToken(PyTypeObject *type, void *token, PyTypeObject **result)

Doc/data/refcounts.dat

@@ -2385,6 +2385,10 @@ PyType_GetFlags:PyTypeObject*:type:0:
 PyType_GetName:PyObject*::+1:
 PyType_GetName:PyTypeObject*:type:0:
 
+PyType_GetModuleByDef:PyObject*::0:
+PyType_GetModuleByDef:PyTypeObject*:type:0:
+PyType_GetModuleByDef:PyModuleDef*:def::
+
 PyType_GetQualName:PyObject*::+1:
 PyType_GetQualName:PyTypeObject*:type:0:
 

Doc/howto/regex.rst

@@ -1016,7 +1016,9 @@ extension.  This regular expression matches ``foo.bar`` and
 Now, consider complicating the problem a bit; what if you want to match
 filenames where the extension is not ``bat``? Some incorrect attempts:
 
-``.*[.][^b].*$``  The first attempt above tries to exclude ``bat`` by requiring
+``.*[.][^b].*$``
+
+The first attempt above tries to exclude ``bat`` by requiring
 that the first character of the extension is not a ``b``.  This is wrong,
 because the pattern also doesn't match ``foo.bar``.
 
@@ -1043,7 +1045,9 @@ confusing.
 
 A negative lookahead cuts through all this confusion:
 
-``.*[.](?!bat$)[^.]*$``  The negative lookahead means: if the expression ``bat``
+``.*[.](?!bat$)[^.]*$``
+
+The negative lookahead means: if the expression ``bat``
 doesn't match at this point, try the rest of the pattern; if ``bat$`` does
 match, the whole pattern will fail.  The trailing ``$`` is required to ensure
 that something like ``sample.batch``, where the extension only starts with

Doc/library/concurrent.futures.rst

@@ -265,7 +265,7 @@ Each worker's interpreter is isolated from all the other interpreters.
 "Isolated" means each interpreter has its own runtime state and
 operates completely independently.  For example, if you redirect
 :data:`sys.stdout` in one interpreter, it will not be automatically
-redirected any other interpreter.  If you import a module in one
+redirected to any other interpreter.  If you import a module in one
 interpreter, it is not automatically imported in any other.  You
 would need to import the module separately in interpreter where
 you need it.  In fact, each module imported in an interpreter is
@@ -287,7 +287,7 @@ efficient alternative is to serialize with :mod:`pickle` and then send
 the bytes over a shared :mod:`socket <socket>` or
 :func:`pipe <os.pipe>`.
 
-.. class:: InterpreterPoolExecutor(max_workers=None, thread_name_prefix='', initializer=None, initargs=(), shared=None)
+.. class:: InterpreterPoolExecutor(max_workers=None, thread_name_prefix='', initializer=None, initargs=())
 
    A :class:`ThreadPoolExecutor` subclass that executes calls asynchronously
    using a pool of at most *max_workers* threads.  Each thread runs
@@ -304,32 +304,17 @@ the bytes over a shared :mod:`socket <socket>` or
    and *initargs* using :mod:`pickle` when sending them to the worker's
    interpreter.
 
-   .. note::
-      Functions defined in the ``__main__`` module cannot be pickled
-      and thus cannot be used.
-
    .. note::
       The executor may replace uncaught exceptions from *initializer*
       with :class:`~concurrent.futures.interpreter.ExecutionFailed`.
 
-   The optional *shared* argument is a :class:`dict` of objects that all
-   interpreters in the pool share.  The *shared* items are added to each
-   interpreter's ``__main__`` module.  Not all objects are shareable.
-   Shareable objects include the builtin singletons, :class:`str`
-   and :class:`bytes`, and :class:`memoryview`.  See :pep:`734`
-   for more info.
-
    Other caveats from parent :class:`ThreadPoolExecutor` apply here.
 
 :meth:`~Executor.submit` and :meth:`~Executor.map` work like normal,
 except the worker serializes the callable and arguments using
 :mod:`pickle` when sending them to its interpreter.  The worker
 likewise serializes the return value when sending it back.
 
-.. note::
-   Functions defined in the ``__main__`` module cannot be pickled
-   and thus cannot be used.
-
 When a worker's current task raises an uncaught exception, the worker
 always tries to preserve the exception as-is.  If that is successful
 then it also sets the ``__cause__`` to a corresponding

Doc/library/pkgutil.rst

@@ -69,8 +69,8 @@ support.
 
    Yield :term:`finder` objects for the given module name.
 
-   If fullname contains a ``'.'``, the finders will be for the package
-   containing fullname, otherwise they will be all registered top level
+   If *fullname* contains a ``'.'``, the finders will be for the package
+   containing *fullname*, otherwise they will be all registered top level
    finders (i.e. those on both :data:`sys.meta_path` and :data:`sys.path_hooks`).
 
    If the named module is in a package, that package is imported as a side

Doc/library/shutil.rst

@@ -327,6 +327,10 @@ Directory and files operations
    The deprecated *onerror* is similar to *onexc*, except that the third
    parameter it receives is the tuple returned from :func:`sys.exc_info`.
 
+   .. seealso::
+      :ref:`shutil-rmtree-example` for an example of handling the removal
+      of a directory tree that contains read-only files.
+
    .. audit-event:: shutil.rmtree path,dir_fd shutil.rmtree
 
    .. versionchanged:: 3.3

Doc/library/zoneinfo.rst

@@ -195,7 +195,7 @@ The ``ZoneInfo`` class
 
 The ``ZoneInfo`` class has two alternate constructors:
 
-.. classmethod:: ZoneInfo.from_file(fobj, /, key=None)
+.. classmethod:: ZoneInfo.from_file(file_obj, /, key=None)
 
     Constructs a ``ZoneInfo`` object from a file-like object returning bytes
     (e.g. a file opened in binary mode or an :class:`io.BytesIO` object).
@@ -325,7 +325,7 @@ The behavior of a ``ZoneInfo`` file depends on how it was constructed:
        >>> a is b
        False
 
-3. ``ZoneInfo.from_file(fobj, /, key=None)``: When constructed from a file, the
+3. ``ZoneInfo.from_file(file_obj, /, key=None)``: When constructed from a file, the
    ``ZoneInfo`` object raises an exception on pickling. If an end user wants to
    pickle a ``ZoneInfo`` constructed from a file, it is recommended that they
    use a wrapper type or a custom serialization function: either serializing by

Include/internal/pycore_long.h

@@ -112,9 +112,9 @@ PyAPI_DATA(PyObject*) _PyLong_Rshift(PyObject *, int64_t);
 // Export for 'math' shared extension
 PyAPI_DATA(PyObject*) _PyLong_Lshift(PyObject *, int64_t);
 
-PyAPI_FUNC(PyObject*) _PyCompactLong_Add(PyLongObject *left, PyLongObject *right);
-PyAPI_FUNC(PyObject*) _PyCompactLong_Multiply(PyLongObject *left, PyLongObject *right);
-PyAPI_FUNC(PyObject*) _PyCompactLong_Subtract(PyLongObject *left, PyLongObject *right);
+PyAPI_FUNC(_PyStackRef) _PyCompactLong_Add(PyLongObject *left, PyLongObject *right);
+PyAPI_FUNC(_PyStackRef) _PyCompactLong_Multiply(PyLongObject *left, PyLongObject *right);
+PyAPI_FUNC(_PyStackRef) _PyCompactLong_Subtract(PyLongObject *left, PyLongObject *right);
 
 // Export for 'binascii' shared extension.
 PyAPI_DATA(unsigned char) _PyLong_DigitValue[256];
@@ -213,7 +213,6 @@ _PyLong_BothAreCompact(const PyLongObject* a, const PyLongObject* b) {
     assert(PyLong_Check(b));
     return (a->long_value.lv_tag | b->long_value.lv_tag) < (2 << NON_SIZE_BITS);
 }
-
 static inline bool
 _PyLong_IsZero(const PyLongObject *op)
 {
@@ -313,6 +312,12 @@ _PyLong_FlipSign(PyLongObject *op) {
 #define _PyLong_FALSE_TAG TAG_FROM_SIGN_AND_SIZE(0, 0)
 #define _PyLong_TRUE_TAG TAG_FROM_SIGN_AND_SIZE(1, 1)
 
+static inline int
+_PyLong_CheckExactAndCompact(PyObject *op)
+{
+    return PyLong_CheckExact(op) && _PyLong_IsCompact((const PyLongObject *)op);
+}
+
 #ifdef __cplusplus
 }
 #endif