diff --git a/.github/workflows/build.yml b/.github/workflows/build.yml index cc5ecc09fbc592..949c4ae95da07f 100644 --- a/.github/workflows/build.yml +++ b/.github/workflows/build.yml @@ -132,7 +132,7 @@ jobs: with: python-version: '3.x' - name: Restore config.cache - uses: actions/cache@v3 + uses: actions/cache@v4 with: path: config.cache key: ${{ github.job }}-${{ runner.os }}-${{ needs.check_source.outputs.config_hash }}-${{ env.pythonLocation }} @@ -259,7 +259,7 @@ jobs: steps: - uses: actions/checkout@v4 - name: Restore config.cache - uses: actions/cache@v3 + uses: actions/cache@v4 with: path: config.cache key: ${{ github.job }}-${{ runner.os }}-${{ needs.check_source.outputs.config_hash }} @@ -274,7 +274,7 @@ jobs: echo "LD_LIBRARY_PATH=${GITHUB_WORKSPACE}/multissl/openssl/${OPENSSL_VER}/lib" >> $GITHUB_ENV - name: 'Restore OpenSSL build' id: cache-openssl - uses: actions/cache@v3 + uses: actions/cache@v4 with: path: ./multissl/openssl/${{ env.OPENSSL_VER }} key: ${{ runner.os }}-multissl-openssl-${{ env.OPENSSL_VER }} @@ -319,7 +319,7 @@ jobs: echo "LD_LIBRARY_PATH=${GITHUB_WORKSPACE}/multissl/openssl/${OPENSSL_VER}/lib" >> $GITHUB_ENV - name: 'Restore OpenSSL build' id: cache-openssl - uses: actions/cache@v3 + uses: actions/cache@v4 with: path: ./multissl/openssl/${{ env.OPENSSL_VER }} key: ${{ runner.os }}-multissl-openssl-${{ env.OPENSSL_VER }} @@ -342,7 +342,7 @@ jobs: - name: Bind mount sources read-only run: sudo mount --bind -o ro $GITHUB_WORKSPACE $CPYTHON_RO_SRCDIR - name: Restore config.cache - uses: actions/cache@v3 + uses: actions/cache@v4 with: path: ${{ env.CPYTHON_BUILDDIR }}/config.cache key: ${{ github.job }}-${{ runner.os }}-${{ needs.check_source.outputs.config_hash }} @@ -375,7 +375,7 @@ jobs: ./python -m venv $VENV_LOC && $VENV_PYTHON -m pip install -r ${GITHUB_WORKSPACE}/Tools/requirements-hypothesis.txt - name: 'Restore Hypothesis database' id: cache-hypothesis-database - uses: actions/cache@v3 + uses: actions/cache@v4 with: path: ./hypothesis key: hypothesis-database-${{ github.head_ref || github.run_id }} @@ -421,7 +421,7 @@ jobs: steps: - uses: actions/checkout@v4 - name: Restore config.cache - uses: actions/cache@v3 + uses: actions/cache@v4 with: path: config.cache key: ${{ github.job }}-${{ runner.os }}-${{ needs.check_source.outputs.config_hash }} @@ -440,7 +440,7 @@ jobs: echo "LD_LIBRARY_PATH=${GITHUB_WORKSPACE}/multissl/openssl/${OPENSSL_VER}/lib" >> $GITHUB_ENV - name: 'Restore OpenSSL build' id: cache-openssl - uses: actions/cache@v3 + uses: actions/cache@v4 with: path: ./multissl/openssl/${{ env.OPENSSL_VER }} key: ${{ runner.os }}-multissl-openssl-${{ env.OPENSSL_VER }} diff --git a/.github/workflows/jit.yml b/.github/workflows/jit.yml index 3da72919181255..69648d87947ad6 100644 --- a/.github/workflows/jit.yml +++ b/.github/workflows/jit.yml @@ -1,14 +1,24 @@ name: JIT on: pull_request: - paths: '**jit**' + paths: + - '**jit**' + - 'Python/bytecodes.c' push: - paths: '**jit**' + paths: + - '**jit**' + - 'Python/bytecodes.c' workflow_dispatch: + +concurrency: + group: ${{ github.workflow }}-${{ github.head_ref || github.run_id }} + cancel-in-progress: true + jobs: jit: name: ${{ matrix.target }} (${{ matrix.debug && 'Debug' || 'Release' }}) runs-on: ${{ matrix.runner }} + timeout-minutes: 60 strategy: fail-fast: false matrix: diff --git a/.github/workflows/reusable-docs.yml b/.github/workflows/reusable-docs.yml index e534751ee1011d..cea8f93d67b29c 100644 --- a/.github/workflows/reusable-docs.yml +++ b/.github/workflows/reusable-docs.yml @@ -89,7 +89,7 @@ jobs: timeout-minutes: 60 steps: - uses: actions/checkout@v4 - - uses: actions/cache@v3 + - uses: actions/cache@v4 with: path: ~/.cache/pip key: ubuntu-doc-${{ hashFiles('Doc/requirements.txt') }} diff --git a/.github/workflows/reusable-macos.yml b/.github/workflows/reusable-macos.yml index c24b6e963ddfd6..cad619b78ce5f2 100644 --- a/.github/workflows/reusable-macos.yml +++ b/.github/workflows/reusable-macos.yml @@ -12,20 +12,27 @@ on: jobs: build_macos: name: 'build and test' - runs-on: macos-latest timeout-minutes: 60 env: HOMEBREW_NO_ANALYTICS: 1 HOMEBREW_NO_AUTO_UPDATE: 1 HOMEBREW_NO_INSTALL_CLEANUP: 1 PYTHONSTRICTEXTENSIONBUILD: 1 + strategy: + fail-fast: false + matrix: + os: [ + "macos-14", # M1 + "macos-13", # Intel + ] + runs-on: ${{ matrix.os }} steps: - uses: actions/checkout@v4 - name: Restore config.cache - uses: actions/cache@v3 + uses: actions/cache@v4 with: path: config.cache - key: ${{ github.job }}-${{ runner.os }}-${{ inputs.config_hash }} + key: ${{ github.job }}-${{ matrix.os }}-${{ inputs.config_hash }} - name: Install Homebrew dependencies run: brew install pkg-config openssl@3.0 xz gdbm tcl-tk - name: Configure CPython diff --git a/.github/workflows/reusable-ubuntu.yml b/.github/workflows/reusable-ubuntu.yml index c2194280c0a50f..ef52d99c15191b 100644 --- a/.github/workflows/reusable-ubuntu.yml +++ b/.github/workflows/reusable-ubuntu.yml @@ -29,7 +29,7 @@ jobs: echo "LD_LIBRARY_PATH=${GITHUB_WORKSPACE}/multissl/openssl/${OPENSSL_VER}/lib" >> $GITHUB_ENV - name: 'Restore OpenSSL build' id: cache-openssl - uses: actions/cache@v3 + uses: actions/cache@v4 with: path: ./multissl/openssl/${{ env.OPENSSL_VER }} key: ${{ runner.os }}-multissl-openssl-${{ env.OPENSSL_VER }} @@ -53,7 +53,7 @@ jobs: - name: Bind mount sources read-only run: sudo mount --bind -o ro $GITHUB_WORKSPACE $CPYTHON_RO_SRCDIR - name: Restore config.cache - uses: actions/cache@v3 + uses: actions/cache@v4 with: path: ${{ env.CPYTHON_BUILDDIR }}/config.cache key: ${{ github.job }}-${{ runner.os }}-${{ inputs.config_hash }} diff --git a/.pre-commit-config.yaml b/.pre-commit-config.yaml index 19033ce243d9d3..69d85238985150 100644 --- a/.pre-commit-config.yaml +++ b/.pre-commit-config.yaml @@ -1,6 +1,6 @@ repos: - repo: https://github.com/astral-sh/ruff-pre-commit - rev: v0.1.7 + rev: v0.2.0 hooks: - id: ruff name: Run Ruff on Lib/test/ diff --git a/Doc/c-api/import.rst b/Doc/c-api/import.rst index 51c20b202f091c..7c74e9e88678dc 100644 --- a/Doc/c-api/import.rst +++ b/Doc/c-api/import.rst @@ -13,20 +13,8 @@ Importing Modules single: __all__ (package variable) single: modules (in module sys) - This is a simplified interface to :c:func:`PyImport_ImportModuleEx` below, - leaving the *globals* and *locals* arguments set to ``NULL`` and *level* set - to 0. When the *name* - argument contains a dot (when it specifies a submodule of a package), the - *fromlist* argument is set to the list ``['*']`` so that the return value is the - named module rather than the top-level package containing it as would otherwise - be the case. (Unfortunately, this has an additional side effect when *name* in - fact specifies a subpackage instead of a submodule: the submodules specified in - the package's ``__all__`` variable are loaded.) Return a new reference to the - imported module, or ``NULL`` with an exception set on failure. A failing - import of a module doesn't leave the module in :data:`sys.modules`. - - This function always uses absolute imports. - + This is a wrapper around :c:func:`PyImport_Import()` which takes a + :c:expr:`const char *` as an argument instead of a :c:expr:`PyObject *`. .. c:function:: PyObject* PyImport_ImportModuleNoBlock(const char *name) diff --git a/Doc/c-api/list.rst b/Doc/c-api/list.rst index c8b64bad702f50..53eb54d3e1021a 100644 --- a/Doc/c-api/list.rst +++ b/Doc/c-api/list.rst @@ -56,13 +56,21 @@ List Objects Similar to :c:func:`PyList_Size`, but without error checking. -.. c:function:: PyObject* PyList_GetItem(PyObject *list, Py_ssize_t index) +.. c:function:: PyObject* PyList_GetItemRef(PyObject *list, Py_ssize_t index) Return the object at position *index* in the list pointed to by *list*. The position must be non-negative; indexing from the end of the list is not - supported. If *index* is out of bounds (<0 or >=len(list)), + supported. If *index* is out of bounds (:code:`<0 or >=len(list)`), return ``NULL`` and set an :exc:`IndexError` exception. + .. versionadded:: 3.13 + + +.. c:function:: PyObject* PyList_GetItem(PyObject *list, Py_ssize_t index) + + Like :c:func:`PyList_GetItemRef`, but returns a + :term:`borrowed reference` instead of a :term:`strong reference`. + .. c:function:: PyObject* PyList_GET_ITEM(PyObject *list, Py_ssize_t i) diff --git a/Doc/data/refcounts.dat b/Doc/data/refcounts.dat index f719ce153b239a..62a96146d605ff 100644 --- a/Doc/data/refcounts.dat +++ b/Doc/data/refcounts.dat @@ -1133,6 +1133,10 @@ PyList_GetItem:PyObject*::0: PyList_GetItem:PyObject*:list:0: PyList_GetItem:Py_ssize_t:index:: +PyList_GetItemRef:PyObject*::+1: +PyList_GetItemRef:PyObject*:list:0: +PyList_GetItemRef:Py_ssize_t:index:: + PyList_GetSlice:PyObject*::+1: PyList_GetSlice:PyObject*:list:0: PyList_GetSlice:Py_ssize_t:low:: diff --git a/Doc/data/stable_abi.dat b/Doc/data/stable_abi.dat index da28a2be60bc1b..def1903204add7 100644 --- a/Doc/data/stable_abi.dat +++ b/Doc/data/stable_abi.dat @@ -336,6 +336,7 @@ var,PyListRevIter_Type,3.2,, function,PyList_Append,3.2,, function,PyList_AsTuple,3.2,, function,PyList_GetItem,3.2,, +function,PyList_GetItemRef,3.13,, function,PyList_GetSlice,3.2,, function,PyList_Insert,3.2,, function,PyList_New,3.2,, diff --git a/Doc/glossary.rst b/Doc/glossary.rst index 098bfffb104ef6..f656e32514c717 100644 --- a/Doc/glossary.rst +++ b/Doc/glossary.rst @@ -341,7 +341,7 @@ Glossary docstring A string literal which appears as the first expression in a class, function or module. While ignored when the suite is executed, it is - recognized by the compiler and put into the :attr:`__doc__` attribute + recognized by the compiler and put into the :attr:`!__doc__` attribute of the enclosing class, function or module. Since it is available via introspection, it is the canonical place for documentation of the object. @@ -1104,10 +1104,12 @@ Glossary The :class:`collections.abc.Sequence` abstract base class defines a much richer interface that goes beyond just :meth:`~object.__getitem__` and :meth:`~object.__len__`, adding - :meth:`count`, :meth:`index`, :meth:`~object.__contains__`, and + :meth:`!count`, :meth:`!index`, :meth:`~object.__contains__`, and :meth:`~object.__reversed__`. Types that implement this expanded interface can be registered explicitly using - :func:`~abc.ABCMeta.register`. + :func:`~abc.ABCMeta.register`. For more documentation on sequence + methods generally, see + :ref:`Common Sequence Operations `. set comprehension A compact way to process all or part of the elements in an iterable and diff --git a/Doc/howto/enum.rst b/Doc/howto/enum.rst index 1e9ac9b6761b64..30be15230fc088 100644 --- a/Doc/howto/enum.rst +++ b/Doc/howto/enum.rst @@ -497,13 +497,30 @@ the :meth:`~Enum.__repr__` omits the inherited class' name. For example:: >>> Creature.DOG -Use the :func:`!dataclass` argument ``repr=False`` +Use the :func:`~dataclasses.dataclass` argument ``repr=False`` to use the standard :func:`repr`. .. versionchanged:: 3.12 Only the dataclass fields are shown in the value area, not the dataclass' name. +.. note:: + + Adding :func:`~dataclasses.dataclass` decorator to :class:`Enum` + and its subclasses is not supported. It will not raise any errors, + but it will produce very strange results at runtime, such as members + being equal to each other:: + + >>> @dataclass # don't do this: it does not make any sense + ... class Color(Enum): + ... RED = 1 + ... BLUE = 2 + ... + >>> Color.RED is Color.BLUE + False + >>> Color.RED == Color.BLUE # problem is here: they should not be equal + True + Pickling -------- diff --git a/Doc/howto/logging-cookbook.rst b/Doc/howto/logging-cookbook.rst index ea494f2fdbbce4..80147e31fcbae1 100644 --- a/Doc/howto/logging-cookbook.rst +++ b/Doc/howto/logging-cookbook.rst @@ -1933,30 +1933,28 @@ This dictionary is passed to :func:`~config.dictConfig` to put the configuration LOGGING = { 'version': 1, - 'disable_existing_loggers': True, + 'disable_existing_loggers': False, 'formatters': { 'verbose': { - 'format': '%(levelname)s %(asctime)s %(module)s %(process)d %(thread)d %(message)s' + 'format': '{levelname} {asctime} {module} {process:d} {thread:d} {message}', + 'style': '{', }, 'simple': { - 'format': '%(levelname)s %(message)s' + 'format': '{levelname} {message}', + 'style': '{', }, }, 'filters': { 'special': { '()': 'project.logging.SpecialFilter', 'foo': 'bar', - } + }, }, 'handlers': { - 'null': { - 'level':'DEBUG', - 'class':'django.utils.log.NullHandler', - }, - 'console':{ - 'level':'DEBUG', - 'class':'logging.StreamHandler', - 'formatter': 'simple' + 'console': { + 'level': 'INFO', + 'class': 'logging.StreamHandler', + 'formatter': 'simple', }, 'mail_admins': { 'level': 'ERROR', @@ -1966,9 +1964,8 @@ This dictionary is passed to :func:`~config.dictConfig` to put the configuration }, 'loggers': { 'django': { - 'handlers':['null'], + 'handlers': ['console'], 'propagate': True, - 'level':'INFO', }, 'django.request': { 'handlers': ['mail_admins'], diff --git a/Doc/howto/logging.rst b/Doc/howto/logging.rst index f164b461c93b9c..347330e98dd00c 100644 --- a/Doc/howto/logging.rst +++ b/Doc/howto/logging.rst @@ -520,7 +520,7 @@ custom handlers) are the following configuration methods: * The :meth:`~Handler.setLevel` method, just as in logger objects, specifies the lowest severity that will be dispatched to the appropriate destination. Why - are there two :func:`setLevel` methods? The level set in the logger + are there two :meth:`~Handler.setLevel` methods? The level set in the logger determines which severity of messages it will pass to its handlers. The level set in each handler determines which messages that handler will send on. @@ -774,29 +774,29 @@ What happens if no configuration is provided If no logging configuration is provided, it is possible to have a situation where a logging event needs to be output, but no handlers can be found to -output the event. The behaviour of the logging package in these -circumstances is dependent on the Python version. +output the event. -For versions of Python prior to 3.2, the behaviour is as follows: +The event is output using a 'handler of last resort', stored in +:data:`lastResort`. This internal handler is not associated with any +logger, and acts like a :class:`~logging.StreamHandler` which writes the +event description message to the current value of ``sys.stderr`` (therefore +respecting any redirections which may be in effect). No formatting is +done on the message - just the bare event description message is printed. +The handler's level is set to ``WARNING``, so all events at this and +greater severities will be output. -* If *logging.raiseExceptions* is ``False`` (production mode), the event is - silently dropped. +.. versionchanged:: 3.2 -* If *logging.raiseExceptions* is ``True`` (development mode), a message - 'No handlers could be found for logger X.Y.Z' is printed once. + For versions of Python prior to 3.2, the behaviour is as follows: -In Python 3.2 and later, the behaviour is as follows: + * If :data:`raiseExceptions` is ``False`` (production mode), the event is + silently dropped. -* The event is output using a 'handler of last resort', stored in - ``logging.lastResort``. This internal handler is not associated with any - logger, and acts like a :class:`~logging.StreamHandler` which writes the - event description message to the current value of ``sys.stderr`` (therefore - respecting any redirections which may be in effect). No formatting is - done on the message - just the bare event description message is printed. - The handler's level is set to ``WARNING``, so all events at this and - greater severities will be output. + * If :data:`raiseExceptions` is ``True`` (development mode), a message + 'No handlers could be found for logger X.Y.Z' is printed once. -To obtain the pre-3.2 behaviour, ``logging.lastResort`` can be set to ``None``. + To obtain the pre-3.2 behaviour, + :data:`lastResort` can be set to ``None``. .. _library-config: @@ -998,7 +998,7 @@ Logged messages are formatted for presentation through instances of the use with the % operator and a dictionary. For formatting multiple messages in a batch, instances of -:class:`~handlers.BufferingFormatter` can be used. In addition to the format +:class:`BufferingFormatter` can be used. In addition to the format string (which is applied to each message in the batch), there is provision for header and trailer format strings. @@ -1034,7 +1034,8 @@ checks to see if a module-level variable, :data:`raiseExceptions`, is set. If set, a traceback is printed to :data:`sys.stderr`. If not set, the exception is swallowed. -.. note:: The default value of :data:`raiseExceptions` is ``True``. This is +.. note:: + The default value of :data:`raiseExceptions` is ``True``. This is because during development, you typically want to be notified of any exceptions that occur. It's advised that you set :data:`raiseExceptions` to ``False`` for production usage. @@ -1072,7 +1073,7 @@ You can write code like this:: expensive_func2()) so that if the logger's threshold is set above ``DEBUG``, the calls to -:func:`expensive_func1` and :func:`expensive_func2` are never made. +``expensive_func1`` and ``expensive_func2`` are never made. .. note:: In some cases, :meth:`~Logger.isEnabledFor` can itself be more expensive than you'd like (e.g. for deeply nested loggers where an explicit diff --git a/Doc/library/asyncio-sync.rst b/Doc/library/asyncio-sync.rst index 05bdf5488af143..3cf8e2737e85dc 100644 --- a/Doc/library/asyncio-sync.rst +++ b/Doc/library/asyncio-sync.rst @@ -216,8 +216,8 @@ Condition .. method:: notify(n=1) - Wake up at most *n* tasks (1 by default) waiting on this - condition. The method is no-op if no tasks are waiting. + Wake up *n* tasks (1 by default) waiting on this + condition. If fewer than *n* tasks are waiting they are all awakened. The lock must be acquired before this method is called and released shortly after. If called with an *unlocked* lock @@ -257,12 +257,18 @@ Condition Once awakened, the Condition re-acquires its lock and this method returns ``True``. + Note that a task *may* return from this call spuriously, + which is why the caller should always re-check the state + and be prepared to :meth:`wait` again. For this reason, you may + prefer to use :meth:`wait_for` instead. + .. coroutinemethod:: wait_for(predicate) Wait until a predicate becomes *true*. The predicate must be a callable which result will be - interpreted as a boolean value. The final value is the + interpreted as a boolean value. The method will repeatedly + :meth:`wait` until the predicate evaluates to *true*. The final value is the return value. diff --git a/Doc/library/calendar.rst b/Doc/library/calendar.rst index c4dcf5641d6066..e699a7284ac802 100644 --- a/Doc/library/calendar.rst +++ b/Doc/library/calendar.rst @@ -512,7 +512,7 @@ to interactively print a calendar. python -m calendar [-h] [-L LOCALE] [-e ENCODING] [-t {text,html}] [-w WIDTH] [-l LINES] [-s SPACING] [-m MONTHS] [-c CSS] - [year] [month] + [-f FIRST_WEEKDAY] [year] [month] For example, to print a calendar for the year 2000: @@ -586,12 +586,13 @@ The following options are accepted: or as an HTML document. -.. option:: --first-weekday WEEKDAY, -f WEEKDAY +.. option:: --first-weekday FIRST_WEEKDAY, -f FIRST_WEEKDAY The weekday to start each week. Must be a number between 0 (Monday) and 6 (Sunday). Defaults to 0. + .. versionadded:: 3.13 .. option:: year diff --git a/Doc/library/collections.abc.rst b/Doc/library/collections.abc.rst index 582bb18f752bd5..7bcaba60c6ddbd 100644 --- a/Doc/library/collections.abc.rst +++ b/Doc/library/collections.abc.rst @@ -136,8 +136,8 @@ ABC Inherits from Abstract Methods Mi :class:`Collection` ``__len__`` ``index``, and ``count`` :class:`MutableSequence` :class:`Sequence` ``__getitem__``, Inherited :class:`Sequence` methods and - ``__setitem__``, ``append``, ``reverse``, ``extend``, ``pop``, - ``__delitem__``, ``remove``, and ``__iadd__`` + ``__setitem__``, ``append``, ``clear``, ``reverse``, ``extend``, + ``__delitem__``, ``pop``, ``remove``, and ``__iadd__`` ``__len__``, ``insert`` diff --git a/Doc/library/csv.rst b/Doc/library/csv.rst index 66888c22b7cc28..fd62b225fcebb8 100644 --- a/Doc/library/csv.rst +++ b/Doc/library/csv.rst @@ -351,6 +351,8 @@ The :mod:`csv` module defines the following constants: Instructs :class:`reader` objects to interpret an empty (unquoted) field as None and to otherwise behave as :data:`QUOTE_ALL`. + .. versionadded:: 3.12 + .. data:: QUOTE_STRINGS Instructs :class:`writer` objects to always place quotes around fields @@ -360,6 +362,8 @@ The :mod:`csv` module defines the following constants: Instructs :class:`reader` objects to interpret an empty (unquoted) string as ``None`` and to otherwise behave as :data:`QUOTE_NONNUMERIC`. + .. versionadded:: 3.12 + The :mod:`csv` module defines the following exception: diff --git a/Doc/library/datetime.rst b/Doc/library/datetime.rst index 4ff049c8709289..930af6cbbe9e8d 100644 --- a/Doc/library/datetime.rst +++ b/Doc/library/datetime.rst @@ -130,8 +130,8 @@ Available Types .. class:: timedelta :noindex: - A duration expressing the difference between two :class:`date`, :class:`.time`, - or :class:`.datetime` instances to microsecond resolution. + A duration expressing the difference between two :class:`.datetime` + or :class:`date` instances to microsecond resolution. .. class:: tzinfo @@ -203,7 +203,7 @@ objects. -------------------------- A :class:`timedelta` object represents a duration, the difference between two -dates or times. +:class:`.datetime` or :class:`date` instances. .. class:: timedelta(days=0, seconds=0, microseconds=0, milliseconds=0, minutes=0, hours=0, weeks=0) @@ -400,30 +400,7 @@ objects (see below). the :func:`divmod` function. True division and multiplication of a :class:`timedelta` object by a :class:`float` object are now supported. - -Comparisons of :class:`timedelta` objects are supported, with some caveats. - -The comparisons ``==`` or ``!=`` *always* return a :class:`bool`, no matter -the type of the compared object:: - - >>> from datetime import timedelta - >>> delta1 = timedelta(seconds=57) - >>> delta2 = timedelta(hours=25, seconds=2) - >>> delta2 != delta1 - True - >>> delta2 == 5 - False - -For all other comparisons (such as ``<`` and ``>``), when a :class:`timedelta` -object is compared to an object of a different type, :exc:`TypeError` -is raised:: - - >>> delta2 > delta1 - True - >>> delta2 > 5 - Traceback (most recent call last): - File "", line 1, in - TypeError: '>' not supported between instances of 'datetime.timedelta' and 'int' +:class:`timedelta` objects support equality and order comparisons. In Boolean contexts, a :class:`timedelta` object is considered to be true if and only if it isn't equal to ``timedelta(0)``. @@ -536,7 +513,15 @@ Other constructors, all class methods: .. classmethod:: date.fromisoformat(date_string) Return a :class:`date` corresponding to a *date_string* given in any valid - ISO 8601 format, except ordinal dates (e.g. ``YYYY-DDD``):: + ISO 8601 format, with the following exceptions: + + 1. Reduced precision dates are not currently supported (``YYYY-MM``, + ``YYYY``). + 2. Extended date representations are not currently supported + (``±YYYYYY-MM-DD``). + 3. Ordinal dates are not currently supported (``YYYY-OOO``). + + Examples:: >>> from datetime import date >>> date.fromisoformat('2019-12-04') @@ -606,8 +591,13 @@ Supported operations: +-------------------------------+----------------------------------------------+ | ``timedelta = date1 - date2`` | \(3) | +-------------------------------+----------------------------------------------+ -| ``date1 < date2`` | *date1* is considered less than *date2* when | -| | *date1* precedes *date2* in time. (4) | +| | ``date1 == date2`` | Equality comparison. (4) | +| | ``date1 != date2`` | | ++-------------------------------+----------------------------------------------+ +| | ``date1 < date2`` | Order comparison. (5) | +| | ``date1 > date2`` | | +| | ``date1 <= date2`` | | +| | ``date1 >= date2`` | | +-------------------------------+----------------------------------------------+ Notes: @@ -627,15 +617,12 @@ Notes: timedelta.microseconds are 0, and date2 + timedelta == date1 after. (4) + :class:`date` objects are equal if they represent the same date. + +(5) + *date1* is considered less than *date2* when *date1* precedes *date2* in time. In other words, ``date1 < date2`` if and only if ``date1.toordinal() < - date2.toordinal()``. Date comparison raises :exc:`TypeError` if - the other comparand isn't also a :class:`date` object. However, - ``NotImplemented`` is returned instead if the other comparand has a - :attr:`~date.timetuple` attribute. This hook gives other kinds of date objects a - chance at implementing mixed-type comparison. If not, when a :class:`date` - object is compared to an object of a different type, :exc:`TypeError` is raised - unless the comparison is ``==`` or ``!=``. The latter cases return - :const:`False` or :const:`True`, respectively. + date2.toordinal()``. In Boolean contexts, all :class:`date` objects are considered to be true. @@ -1017,8 +1004,12 @@ Other constructors, all class methods: 1. Time zone offsets may have fractional seconds. 2. The ``T`` separator may be replaced by any single unicode character. - 3. Ordinal dates are not currently supported. - 4. Fractional hours and minutes are not supported. + 3. Fractional hours and minutes are not supported. + 4. Reduced precision dates are not currently supported (``YYYY-MM``, + ``YYYY``). + 5. Extended date representations are not currently supported + (``±YYYYYY-MM-DD``). + 6. Ordinal dates are not currently supported (``YYYY-OOO``). Examples:: @@ -1158,8 +1149,13 @@ Supported operations: +---------------------------------------+--------------------------------+ | ``timedelta = datetime1 - datetime2`` | \(3) | +---------------------------------------+--------------------------------+ -| ``datetime1 < datetime2`` | Compares :class:`.datetime` to | -| | :class:`.datetime`. (4) | +| | ``datetime1 == datetime2`` | Equality comparison. (4) | +| | ``datetime1 != datetime2`` | | ++---------------------------------------+--------------------------------+ +| | ``datetime1 < datetime2`` | Order comparison. (5) | +| | ``datetime1 > datetime2`` | | +| | ``datetime1 <= datetime2`` | | +| | ``datetime1 >= datetime2`` | | +---------------------------------------+--------------------------------+ (1) @@ -1187,39 +1183,40 @@ Supported operations: are done in this case. If both are aware and have different :attr:`~.datetime.tzinfo` attributes, ``a-b`` acts - as if *a* and *b* were first converted to naive UTC datetimes first. The + as if *a* and *b* were first converted to naive UTC datetimes. The result is ``(a.replace(tzinfo=None) - a.utcoffset()) - (b.replace(tzinfo=None) - b.utcoffset())`` except that the implementation never overflows. (4) - *datetime1* is considered less than *datetime2* when *datetime1* precedes - *datetime2* in time. + :class:`.datetime` objects are equal if they represent the same date + and time, taking into account the time zone. - If one comparand is naive and the other is aware, :exc:`TypeError` - is raised if an order comparison is attempted. For equality - comparisons, naive instances are never equal to aware instances. + Naive and aware :class:`!datetime` objects are never equal. + :class:`!datetime` objects are never equal to :class:`date` objects + that are not also :class:`!datetime` instances, even if they represent + the same date. - If both comparands are aware, and have the same :attr:`~.datetime.tzinfo` attribute, the - common :attr:`~.datetime.tzinfo` attribute is ignored and the base datetimes are - compared. If both comparands are aware and have different :attr:`~.datetime.tzinfo` - attributes, the comparands are first adjusted by subtracting their UTC - offsets (obtained from ``self.utcoffset()``). + If both comparands are aware and have different :attr:`~.datetime.tzinfo` + attributes, the comparison acts as comparands were first converted to UTC + datetimes except that the implementation never overflows. + :class:`!datetime` instances in a repeated interval are never equal to + :class:`!datetime` instances in other time zone. - .. versionchanged:: 3.3 - Equality comparisons between aware and naive :class:`.datetime` - instances don't raise :exc:`TypeError`. +(5) + *datetime1* is considered less than *datetime2* when *datetime1* precedes + *datetime2* in time, taking into account the time zone. - .. note:: + Order comparison between naive and aware :class:`.datetime` objects, + as well as a :class:`!datetime` object and a :class:`!date` object + that is not also a :class:`!datetime` instance, raises :exc:`TypeError`. - In order to stop comparison from falling back to the default scheme of comparing - object addresses, datetime comparison normally raises :exc:`TypeError` if the - other comparand isn't also a :class:`.datetime` object. However, - ``NotImplemented`` is returned instead if the other comparand has a - :attr:`~.datetime.timetuple` attribute. This hook gives other kinds of date objects a - chance at implementing mixed-type comparison. If not, when a :class:`.datetime` - object is compared to an object of a different type, :exc:`TypeError` is raised - unless the comparison is ``==`` or ``!=``. The latter cases return - :const:`False` or :const:`True`, respectively. + If both comparands are aware and have different :attr:`~.datetime.tzinfo` + attributes, the comparison acts as comparands were first converted to UTC + datetimes except that the implementation never overflows. + +.. versionchanged:: 3.3 + Equality comparisons between aware and naive :class:`.datetime` + instances don't raise :exc:`TypeError`. Instance methods: @@ -1754,21 +1751,18 @@ Instance attributes (read-only): .. versionadded:: 3.6 -:class:`.time` objects support comparison of :class:`.time` to :class:`.time`, -where *a* is considered less -than *b* when *a* precedes *b* in time. If one comparand is naive and the other -is aware, :exc:`TypeError` is raised if an order comparison is attempted. For equality -comparisons, naive instances are never equal to aware instances. +:class:`.time` objects support equality and order comparisons, +where *a* is considered less than *b* when *a* precedes *b* in time. + +Naive and aware :class:`!time` objects are never equal. +Order comparison between naive and aware :class:`!time` objects raises +:exc:`TypeError`. If both comparands are aware, and have the same :attr:`~.time.tzinfo` attribute, the common :attr:`!tzinfo` attribute is ignored and the base times are compared. If both comparands are aware and have different :attr:`!tzinfo` attributes, the comparands are first adjusted by -subtracting their UTC offsets (obtained from ``self.utcoffset()``). In order -to stop mixed-type comparisons from falling back to the default comparison by -object address, when a :class:`.time` object is compared to an object of a -different type, :exc:`TypeError` is raised unless the comparison is ``==`` or -``!=``. The latter cases return :const:`False` or :const:`True`, respectively. +subtracting their UTC offsets (obtained from ``self.utcoffset()``). .. versionchanged:: 3.3 Equality comparisons between aware and naive :class:`.time` instances diff --git a/Doc/library/enum.rst b/Doc/library/enum.rst index 07b15e23b2c10a..f31e6ea848f3b2 100644 --- a/Doc/library/enum.rst +++ b/Doc/library/enum.rst @@ -337,6 +337,17 @@ Data Types >>> PowersOfThree.SECOND.value 9 + .. method:: Enum.__init__(self, \*args, \**kwds) + + By default, does nothing. If multiple values are given in the member + assignment, those values become separate arguments to ``__init__``; e.g. + + >>> from enum import Enum + >>> class Weekday(Enum): + ... MONDAY = 1, 'Mon' + + ``Weekday.__init__()`` would be called as ``Weekday.__init__(self, 1, 'Mon')`` + .. method:: Enum.__init_subclass__(cls, \**kwds) A *classmethod* that is used to further configure subsequent subclasses. @@ -364,6 +375,18 @@ Data Types >>> Build('deBUG') + .. method:: Enum.__new__(cls, \*args, \**kwds) + + By default, doesn't exist. If specified, either in the enum class + definition or in a mixin class (such as ``int``), all values given + in the member assignment will be passed; e.g. + + >>> from enum import Enum + >>> class MyIntEnum(Enum): + ... SEVENTEEN = '1a', 16 + + results in the call ``int('1a', 16)`` and a value of ``17`` for the member. + .. method:: Enum.__repr__(self) Returns the string used for *repr()* calls. By default, returns the @@ -477,9 +500,9 @@ Data Types .. class:: Flag - *Flag* members support the bitwise operators ``&`` (*AND*), ``|`` (*OR*), - ``^`` (*XOR*), and ``~`` (*INVERT*); the results of those operators are members - of the enumeration. + ``Flag`` is the same as :class:`Enum`, but its members support the bitwise + operators ``&`` (*AND*), ``|`` (*OR*), ``^`` (*XOR*), and ``~`` (*INVERT*); + the results of those operators are members of the enumeration. .. method:: __contains__(self, value) diff --git a/Doc/library/gzip.rst b/Doc/library/gzip.rst index 50cde09fa10a9d..79be215a766045 100644 --- a/Doc/library/gzip.rst +++ b/Doc/library/gzip.rst @@ -61,7 +61,7 @@ The module defines the following items: .. exception:: BadGzipFile - An exception raised for invalid gzip files. It inherits :exc:`OSError`. + An exception raised for invalid gzip files. It inherits from :exc:`OSError`. :exc:`EOFError` and :exc:`zlib.error` can also be raised for invalid gzip files. @@ -287,4 +287,3 @@ Command line options .. option:: -h, --help Show the help message. - diff --git a/Doc/library/io.rst b/Doc/library/io.rst index 6736aa9ee2b0ef..8eb531aa4ea248 100644 --- a/Doc/library/io.rst +++ b/Doc/library/io.rst @@ -466,7 +466,7 @@ I/O Base Classes .. class:: RawIOBase - Base class for raw binary streams. It inherits :class:`IOBase`. + Base class for raw binary streams. It inherits from :class:`IOBase`. Raw binary streams typically provide low-level access to an underlying OS device or API, and do not try to encapsulate it in high-level primitives @@ -519,7 +519,7 @@ I/O Base Classes .. class:: BufferedIOBase Base class for binary streams that support some kind of buffering. - It inherits :class:`IOBase`. + It inherits from :class:`IOBase`. The main difference with :class:`RawIOBase` is that methods :meth:`read`, :meth:`readinto` and :meth:`write` will try (respectively) to read as much @@ -633,7 +633,7 @@ Raw File I/O .. class:: FileIO(name, mode='r', closefd=True, opener=None) A raw binary stream representing an OS-level file containing bytes data. It - inherits :class:`RawIOBase`. + inherits from :class:`RawIOBase`. The *name* can be one of two things: @@ -696,7 +696,7 @@ than raw I/O does. .. class:: BytesIO(initial_bytes=b'') - A binary stream using an in-memory bytes buffer. It inherits + A binary stream using an in-memory bytes buffer. It inherits from :class:`BufferedIOBase`. The buffer is discarded when the :meth:`~IOBase.close` method is called. @@ -745,7 +745,7 @@ than raw I/O does. .. class:: BufferedReader(raw, buffer_size=DEFAULT_BUFFER_SIZE) A buffered binary stream providing higher-level access to a readable, non - seekable :class:`RawIOBase` raw binary stream. It inherits + seekable :class:`RawIOBase` raw binary stream. It inherits from :class:`BufferedIOBase`. When reading data from this object, a larger amount of data may be @@ -783,7 +783,7 @@ than raw I/O does. .. class:: BufferedWriter(raw, buffer_size=DEFAULT_BUFFER_SIZE) A buffered binary stream providing higher-level access to a writeable, non - seekable :class:`RawIOBase` raw binary stream. It inherits + seekable :class:`RawIOBase` raw binary stream. It inherits from :class:`BufferedIOBase`. When writing to this object, data is normally placed into an internal @@ -818,7 +818,7 @@ than raw I/O does. .. class:: BufferedRandom(raw, buffer_size=DEFAULT_BUFFER_SIZE) A buffered binary stream providing higher-level access to a seekable - :class:`RawIOBase` raw binary stream. It inherits :class:`BufferedReader` + :class:`RawIOBase` raw binary stream. It inherits from :class:`BufferedReader` and :class:`BufferedWriter`. The constructor creates a reader and writer for a seekable raw stream, given @@ -834,7 +834,7 @@ than raw I/O does. A buffered binary stream providing higher-level access to two non seekable :class:`RawIOBase` raw binary streams---one readable, the other writeable. - It inherits :class:`BufferedIOBase`. + It inherits from :class:`BufferedIOBase`. *reader* and *writer* are :class:`RawIOBase` objects that are readable and writeable respectively. If the *buffer_size* is omitted it defaults to @@ -857,7 +857,7 @@ Text I/O .. class:: TextIOBase Base class for text streams. This class provides a character and line based - interface to stream I/O. It inherits :class:`IOBase`. + interface to stream I/O. It inherits from :class:`IOBase`. :class:`TextIOBase` provides or overrides these data attributes and methods in addition to those from :class:`IOBase`: @@ -946,7 +946,7 @@ Text I/O line_buffering=False, write_through=False) A buffered text stream providing higher-level access to a - :class:`BufferedIOBase` buffered binary stream. It inherits + :class:`BufferedIOBase` buffered binary stream. It inherits from :class:`TextIOBase`. *encoding* gives the name of the encoding that the stream will be decoded or @@ -1073,7 +1073,7 @@ Text I/O .. class:: StringIO(initial_value='', newline='\n') - A text stream using an in-memory text buffer. It inherits + A text stream using an in-memory text buffer. It inherits from :class:`TextIOBase`. The text buffer is discarded when the :meth:`~IOBase.close` method is @@ -1124,7 +1124,7 @@ Text I/O .. class:: IncrementalNewlineDecoder A helper codec that decodes newlines for :term:`universal newlines` mode. - It inherits :class:`codecs.IncrementalDecoder`. + It inherits from :class:`codecs.IncrementalDecoder`. Performance diff --git a/Doc/library/logging.rst b/Doc/library/logging.rst index 4b756d10b4c586..39eb41ce1f1670 100644 --- a/Doc/library/logging.rst +++ b/Doc/library/logging.rst @@ -531,12 +531,12 @@ subclasses. However, the :meth:`!__init__` method in subclasses needs to call This method should be called from handlers when an exception is encountered during an :meth:`emit` call. If the module-level attribute - ``raiseExceptions`` is ``False``, exceptions get silently ignored. This is + :data:`raiseExceptions` is ``False``, exceptions get silently ignored. This is what is mostly wanted for a logging system - most users will not care about errors in the logging system, they are more interested in application errors. You could, however, replace this with a custom handler if you wish. The specified record is the one which was being processed when the exception - occurred. (The default value of ``raiseExceptions`` is ``True``, as that is + occurred. (The default value of :data:`raiseExceptions` is ``True``, as that is more useful during development). @@ -1494,6 +1494,18 @@ Module-Level Attributes .. versionadded:: 3.2 +.. attribute:: raiseExceptions + + Used to see if exceptions during handling should be propagated. + + Default: ``True``. + + If :data:`raiseExceptions` is ``False``, + exceptions get silently ignored. This is what is mostly wanted + for a logging system - most users will not care about errors in + the logging system, they are more interested in application errors. + + Integration with the warnings module ------------------------------------ diff --git a/Doc/library/multiprocessing.rst b/Doc/library/multiprocessing.rst index 789a84b02d59d2..b104a6483b70e6 100644 --- a/Doc/library/multiprocessing.rst +++ b/Doc/library/multiprocessing.rst @@ -56,7 +56,7 @@ will print to standard output :: The :class:`Process` class -~~~~~~~~~~~~~~~~~~~~~~~~~~ +^^^^^^^^^^^^^^^^^^^^^^^^^^ In :mod:`multiprocessing`, processes are spawned by creating a :class:`Process` object and then calling its :meth:`~Process.start` method. :class:`Process` @@ -102,7 +102,7 @@ necessary, see :ref:`multiprocessing-programming`. .. _multiprocessing-start-methods: Contexts and start methods -~~~~~~~~~~~~~~~~~~~~~~~~~~ +^^^^^^^^^^^^^^^^^^^^^^^^^^ Depending on the platform, :mod:`multiprocessing` supports three ways to start a process. These *start methods* are @@ -231,7 +231,7 @@ library user. Exchanging objects between processes -~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ +^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^ :mod:`multiprocessing` supports two types of communication channel between processes: @@ -283,7 +283,7 @@ processes: Synchronization between processes -~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ +^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^ :mod:`multiprocessing` contains equivalents of all the synchronization primitives from :mod:`threading`. For instance one can use a lock to ensure @@ -309,7 +309,7 @@ mixed up. Sharing state between processes -~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ +^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^ As mentioned above, when doing concurrent programming it is usually best to avoid using shared state as far as possible. This is particularly true when @@ -399,7 +399,7 @@ However, if you really do need to use some shared data then Using a pool of workers -~~~~~~~~~~~~~~~~~~~~~~~ +^^^^^^^^^^^^^^^^^^^^^^^ The :class:`~multiprocessing.pool.Pool` class represents a pool of worker processes. It has methods which allows tasks to be offloaded to the worker @@ -490,7 +490,7 @@ The :mod:`multiprocessing` package mostly replicates the API of the :class:`Process` and exceptions -~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ +^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^ .. class:: Process(group=None, target=None, name=None, args=(), kwargs={}, \ *, daemon=None) @@ -724,7 +724,7 @@ The :mod:`multiprocessing` package mostly replicates the API of the Raised by methods with a timeout when the timeout expires. Pipes and Queues -~~~~~~~~~~~~~~~~ +^^^^^^^^^^^^^^^^ When using multiple processes, one generally uses message passing for communication between processes and avoids having to use any synchronization @@ -981,7 +981,7 @@ For an example of the usage of queues for interprocess communication see Miscellaneous -~~~~~~~~~~~~~ +^^^^^^^^^^^^^ .. function:: active_children() @@ -1150,7 +1150,7 @@ Miscellaneous Connection Objects -~~~~~~~~~~~~~~~~~~ +^^^^^^^^^^^^^^^^^^ .. currentmodule:: multiprocessing.connection @@ -1292,7 +1292,7 @@ For example: Synchronization primitives -~~~~~~~~~~~~~~~~~~~~~~~~~~ +^^^^^^^^^^^^^^^^^^^^^^^^^^ .. currentmodule:: multiprocessing @@ -1481,7 +1481,7 @@ object -- see :ref:`multiprocessing-managers`. Shared :mod:`ctypes` Objects -~~~~~~~~~~~~~~~~~~~~~~~~~~~~ +^^^^^^^^^^^^^^^^^^^^^^^^^^^^ It is possible to create shared objects using shared memory which can be inherited by child processes. @@ -1543,7 +1543,7 @@ inherited by child processes. The :mod:`multiprocessing.sharedctypes` module ->>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>> +"""""""""""""""""""""""""""""""""""""""""""""" .. module:: multiprocessing.sharedctypes :synopsis: Allocate ctypes objects from shared memory. @@ -1709,7 +1709,7 @@ The results printed are :: .. _multiprocessing-managers: Managers -~~~~~~~~ +^^^^^^^^ Managers provide a way to create data which can be shared between different processes, including sharing over a network between processes running on @@ -1954,7 +1954,7 @@ their parent process exits. The manager classes are defined in the Customized managers ->>>>>>>>>>>>>>>>>>> +""""""""""""""""""" To create one's own manager, one creates a subclass of :class:`BaseManager` and uses the :meth:`~BaseManager.register` classmethod to register new types or @@ -1981,7 +1981,7 @@ callables with the manager class. For example:: Using a remote manager ->>>>>>>>>>>>>>>>>>>>>> +"""""""""""""""""""""" It is possible to run a manager server on one machine and have clients use it from other machines (assuming that the firewalls involved allow it). @@ -2044,7 +2044,7 @@ client to access it remotely:: .. _multiprocessing-proxy_objects: Proxy Objects -~~~~~~~~~~~~~ +^^^^^^^^^^^^^ A proxy is an object which *refers* to a shared object which lives (presumably) in a different process. The shared object is said to be the *referent* of the @@ -2196,7 +2196,7 @@ demonstrates a level of control over the synchronization. Cleanup ->>>>>>> +""""""" A proxy object uses a weakref callback so that when it gets garbage collected it deregisters itself from the manager which owns its referent. @@ -2206,7 +2206,7 @@ any proxies referring to it. Process Pools -~~~~~~~~~~~~~ +^^^^^^^^^^^^^ .. module:: multiprocessing.pool :synopsis: Create pools of processes. @@ -2442,7 +2442,7 @@ The following example demonstrates the use of a pool:: .. _multiprocessing-listeners-clients: Listeners and Clients -~~~~~~~~~~~~~~~~~~~~~ +^^^^^^^^^^^^^^^^^^^^^ .. module:: multiprocessing.connection :synopsis: API for dealing with sockets. @@ -2665,7 +2665,7 @@ wait for messages from multiple processes at once:: .. _multiprocessing-address-formats: Address Formats ->>>>>>>>>>>>>>> +""""""""""""""" * An ``'AF_INET'`` address is a tuple of the form ``(hostname, port)`` where *hostname* is a string and *port* is an integer. @@ -2685,7 +2685,7 @@ an ``'AF_PIPE'`` address rather than an ``'AF_UNIX'`` address. .. _multiprocessing-auth-keys: Authentication keys -~~~~~~~~~~~~~~~~~~~ +^^^^^^^^^^^^^^^^^^^ When one uses :meth:`Connection.recv `, the data received is automatically @@ -2711,7 +2711,7 @@ Suitable authentication keys can also be generated by using :func:`os.urandom`. Logging -~~~~~~~ +^^^^^^^ Some support for logging is available. Note, however, that the :mod:`logging` package does not use process shared locks so it is possible (depending on the @@ -2759,7 +2759,7 @@ For a full table of logging levels, see the :mod:`logging` module. The :mod:`multiprocessing.dummy` module -~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ +^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^ .. module:: multiprocessing.dummy :synopsis: Dumb wrapper around threading. @@ -2818,7 +2818,7 @@ There are certain guidelines and idioms which should be adhered to when using All start methods -~~~~~~~~~~~~~~~~~ +^^^^^^^^^^^^^^^^^ The following applies to all start methods. @@ -2977,7 +2977,7 @@ Beware of replacing :data:`sys.stdin` with a "file like object" For more information, see :issue:`5155`, :issue:`5313` and :issue:`5331` The *spawn* and *forkserver* start methods -~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ +^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^ There are a few extra restriction which don't apply to the *fork* start method. diff --git a/Doc/library/multiprocessing.shared_memory.rst b/Doc/library/multiprocessing.shared_memory.rst index 10d7f061fb759b..933fd07d62418a 100644 --- a/Doc/library/multiprocessing.shared_memory.rst +++ b/Doc/library/multiprocessing.shared_memory.rst @@ -23,7 +23,7 @@ processes, a :class:`~multiprocessing.managers.BaseManager` subclass, :class:`~multiprocessing.managers.SharedMemoryManager`, is also provided in the :mod:`multiprocessing.managers` module. -In this module, shared memory refers to "System V style" shared memory blocks +In this module, shared memory refers to "POSIX style" shared memory blocks (though is not necessarily implemented explicitly as such) and does not refer to "distributed shared memory". This style of shared memory permits distinct processes to potentially read and write to a common (or shared) region of diff --git a/Doc/library/pickle.rst b/Doc/library/pickle.rst index cfb251fca5c7cd..acada092afb679 100644 --- a/Doc/library/pickle.rst +++ b/Doc/library/pickle.rst @@ -272,13 +272,13 @@ The :mod:`pickle` module defines three exceptions: .. exception:: PickleError - Common base class for the other pickling exceptions. It inherits + Common base class for the other pickling exceptions. It inherits from :exc:`Exception`. .. exception:: PicklingError Error raised when an unpicklable object is encountered by :class:`Pickler`. - It inherits :exc:`PickleError`. + It inherits from :exc:`PickleError`. Refer to :ref:`pickle-picklable` to learn what kinds of objects can be pickled. @@ -286,7 +286,7 @@ The :mod:`pickle` module defines three exceptions: .. exception:: UnpicklingError Error raised when there is a problem unpickling an object, such as a data - corruption or a security violation. It inherits :exc:`PickleError`. + corruption or a security violation. It inherits from :exc:`PickleError`. Note that other exceptions may also be raised during unpickling, including (but not necessarily limited to) AttributeError, EOFError, ImportError, and @@ -356,7 +356,7 @@ The :mod:`pickle` module exports three classes, :class:`Pickler`, :func:`copyreg.pickle`. It is a mapping whose keys are classes and whose values are reduction functions. A reduction function takes a single argument of the associated class and should - conform to the same interface as a :meth:`__reduce__` + conform to the same interface as a :meth:`~object.__reduce__` method. By default, a pickler object will not have a @@ -376,7 +376,7 @@ The :mod:`pickle` module exports three classes, :class:`Pickler`, Special reducer that can be defined in :class:`Pickler` subclasses. This method has priority over any reducer in the :attr:`dispatch_table`. It - should conform to the same interface as a :meth:`__reduce__` method, and + should conform to the same interface as a :meth:`~object.__reduce__` method, and can optionally return ``NotImplemented`` to fallback on :attr:`dispatch_table`-registered reducers to pickle ``obj``. @@ -516,7 +516,7 @@ The following types can be pickled: * classes accessible from the top level of a module; -* instances of such classes whose the result of calling :meth:`__getstate__` +* instances of such classes whose the result of calling :meth:`~object.__getstate__` is picklable (see section :ref:`pickle-inst` for details). Attempts to pickle unpicklable objects will raise the :exc:`PicklingError` @@ -552,7 +552,7 @@ purpose, so you can fix bugs in a class or add methods to the class and still load objects that were created with an earlier version of the class. If you plan to have long-lived objects that will see many versions of a class, it may be worthwhile to put a version number in the objects so that suitable -conversions can be made by the class's :meth:`__setstate__` method. +conversions can be made by the class's :meth:`~object.__setstate__` method. .. _pickle-inst: @@ -567,7 +567,7 @@ customize, and control how class instances are pickled and unpickled. In most cases, no additional code is needed to make instances picklable. By default, pickle will retrieve the class and the attributes of an instance via -introspection. When a class instance is unpickled, its :meth:`__init__` method +introspection. When a class instance is unpickled, its :meth:`~object.__init__` method is usually *not* invoked. The default behaviour first creates an uninitialized instance and then restores the saved attributes. The following code shows an implementation of this behaviour:: @@ -658,30 +658,30 @@ methods: Refer to the section :ref:`pickle-state` for more information about how to use -the methods :meth:`__getstate__` and :meth:`__setstate__`. +the methods :meth:`~object.__getstate__` and :meth:`~object.__setstate__`. .. note:: - At unpickling time, some methods like :meth:`__getattr__`, - :meth:`__getattribute__`, or :meth:`__setattr__` may be called upon the + At unpickling time, some methods like :meth:`~object.__getattr__`, + :meth:`~object.__getattribute__`, or :meth:`~object.__setattr__` may be called upon the instance. In case those methods rely on some internal invariant being - true, the type should implement :meth:`__new__` to establish such an - invariant, as :meth:`__init__` is not called when unpickling an + true, the type should implement :meth:`~object.__new__` to establish such an + invariant, as :meth:`~object.__init__` is not called when unpickling an instance. .. index:: pair: copy; protocol As we shall see, pickle does not use directly the methods described above. In fact, these methods are part of the copy protocol which implements the -:meth:`__reduce__` special method. The copy protocol provides a unified +:meth:`~object.__reduce__` special method. The copy protocol provides a unified interface for retrieving the data necessary for pickling and copying objects. [#]_ -Although powerful, implementing :meth:`__reduce__` directly in your classes is +Although powerful, implementing :meth:`~object.__reduce__` directly in your classes is error prone. For this reason, class designers should use the high-level -interface (i.e., :meth:`__getnewargs_ex__`, :meth:`__getstate__` and -:meth:`__setstate__`) whenever possible. We will show, however, cases where -using :meth:`__reduce__` is the only option or leads to more efficient pickling +interface (i.e., :meth:`~object.__getnewargs_ex__`, :meth:`~object.__getstate__` and +:meth:`~object.__setstate__`) whenever possible. We will show, however, cases where +using :meth:`!__reduce__` is the only option or leads to more efficient pickling or both. .. method:: object.__reduce__() @@ -716,8 +716,9 @@ or both. These items will be appended to the object either using ``obj.append(item)`` or, in batch, using ``obj.extend(list_of_items)``. This is primarily used for list subclasses, but may be used by other - classes as long as they have :meth:`append` and :meth:`extend` methods with - the appropriate signature. (Whether :meth:`append` or :meth:`extend` is + classes as long as they have + :ref:`append and extend methods ` with + the appropriate signature. (Whether :meth:`!append` or :meth:`!extend` is used depends on which pickle protocol version is used as well as the number of items to append, so both must be supported.) @@ -793,8 +794,8 @@ any other code which depends on pickling, then one can create a pickler with a private dispatch table. The global dispatch table managed by the :mod:`copyreg` module is -available as :data:`copyreg.dispatch_table`. Therefore, one may -choose to use a modified copy of :data:`copyreg.dispatch_table` as a +available as :data:`!copyreg.dispatch_table`. Therefore, one may +choose to use a modified copy of :data:`!copyreg.dispatch_table` as a private dispatch table. For example :: @@ -833,12 +834,12 @@ Handling Stateful Objects single: __setstate__() (copy protocol) Here's an example that shows how to modify pickling behavior for a class. -The :class:`TextReader` class opens a text file, and returns the line number and +The :class:`!TextReader` class below opens a text file, and returns the line number and line contents each time its :meth:`!readline` method is called. If a -:class:`TextReader` instance is pickled, all attributes *except* the file object +:class:`!TextReader` instance is pickled, all attributes *except* the file object member are saved. When the instance is unpickled, the file is reopened, and -reading resumes from the last location. The :meth:`__setstate__` and -:meth:`__getstate__` methods are used to implement this behavior. :: +reading resumes from the last location. The :meth:`!__setstate__` and +:meth:`!__getstate__` methods are used to implement this behavior. :: class TextReader: """Print and number lines in a text file.""" @@ -903,7 +904,7 @@ functions and classes. For those cases, it is possible to subclass from the :class:`Pickler` class and implement a :meth:`~Pickler.reducer_override` method. This method can return an -arbitrary reduction tuple (see :meth:`__reduce__`). It can alternatively return +arbitrary reduction tuple (see :meth:`~object.__reduce__`). It can alternatively return ``NotImplemented`` to fallback to the traditional behavior. If both the :attr:`~Pickler.dispatch_table` and @@ -971,7 +972,7 @@ provided by pickle protocol 5 and higher. Provider API ^^^^^^^^^^^^ -The large data objects to be pickled must implement a :meth:`__reduce_ex__` +The large data objects to be pickled must implement a :meth:`~object.__reduce_ex__` method specialized for protocol 5 and higher, which returns a :class:`PickleBuffer` instance (instead of e.g. a :class:`bytes` object) for any large data. diff --git a/Doc/library/shutil.rst b/Doc/library/shutil.rst index d9ec2cbc47e611..7a7dd23177e672 100644 --- a/Doc/library/shutil.rst +++ b/Doc/library/shutil.rst @@ -586,7 +586,9 @@ provided. They rely on the :mod:`zipfile` and :mod:`tarfile` modules. Create an archive file (such as zip or tar) and return its name. *base_name* is the name of the file to create, including the path, minus - any format-specific extension. *format* is the archive format: one of + any format-specific extension. + + *format* is the archive format: one of "zip" (if the :mod:`zlib` module is available), "tar", "gztar" (if the :mod:`zlib` module is available), "bztar" (if the :mod:`bz2` module is available), or "xztar" (if the :mod:`lzma` module is available). diff --git a/Doc/library/socketserver.rst b/Doc/library/socketserver.rst index 5fd213fa613c8d..864b1dadb78562 100644 --- a/Doc/library/socketserver.rst +++ b/Doc/library/socketserver.rst @@ -494,7 +494,7 @@ This is the server side:: def handle(self): # self.request is the TCP socket connected to the client self.data = self.request.recv(1024).strip() - print("{} wrote:".format(self.client_address[0])) + print("Received from {}:".format(self.client_address[0])) print(self.data) # just send back the same data, but upper-cased self.request.sendall(self.data.upper()) @@ -525,8 +525,9 @@ objects that simplify communication by providing the standard file interface):: The difference is that the ``readline()`` call in the second handler will call ``recv()`` multiple times until it encounters a newline character, while the -single ``recv()`` call in the first handler will just return what has been sent -from the client in one ``sendall()`` call. +single ``recv()`` call in the first handler will just return what has been +received so far from the client's ``sendall()`` call (typically all of it, but +this is not guaranteed by the TCP protocol). This is the client side:: diff --git a/Doc/library/stdtypes.rst b/Doc/library/stdtypes.rst index 9028ff5c134fa9..1a4c12590c1018 100644 --- a/Doc/library/stdtypes.rst +++ b/Doc/library/stdtypes.rst @@ -1528,7 +1528,7 @@ between them will be implicitly converted to a single string literal. That is, ``("spam " "eggs") == "spam eggs"``. See :ref:`strings` for more about the various forms of string literal, -including supported escape sequences, and the ``r`` ("raw") prefix that +including supported :ref:`escape sequences `, and the ``r`` ("raw") prefix that disables most escape sequence processing. Strings may also be created from other objects using the :class:`str` diff --git a/Doc/library/subprocess.rst b/Doc/library/subprocess.rst index c437ce770b37d0..f63ca73b3ec067 100644 --- a/Doc/library/subprocess.rst +++ b/Doc/library/subprocess.rst @@ -664,7 +664,8 @@ functions. If given, *startupinfo* will be a :class:`STARTUPINFO` object, which is passed to the underlying ``CreateProcess`` function. - *creationflags*, if given, can be one or more of the following flags: + + If given, *creationflags*, can be one or more of the following flags: * :data:`CREATE_NEW_CONSOLE` * :data:`CREATE_NEW_PROCESS_GROUP` diff --git a/Doc/library/symtable.rst b/Doc/library/symtable.rst index 46159dcef940e7..47568387f9a7ce 100644 --- a/Doc/library/symtable.rst +++ b/Doc/library/symtable.rst @@ -97,7 +97,7 @@ Examining Symbol Tables .. class:: Function - A namespace for a function or method. This class inherits + A namespace for a function or method. This class inherits from :class:`SymbolTable`. .. method:: get_parameters() @@ -123,7 +123,7 @@ Examining Symbol Tables .. class:: Class - A namespace of a class. This class inherits :class:`SymbolTable`. + A namespace of a class. This class inherits from :class:`SymbolTable`. .. method:: get_methods() diff --git a/Doc/library/sys.rst b/Doc/library/sys.rst index abf2c393a44928..a97a369b77b88a 100644 --- a/Doc/library/sys.rst +++ b/Doc/library/sys.rst @@ -1293,7 +1293,10 @@ always available. The list of the original command line arguments passed to the Python executable. - See also :data:`sys.argv`. + The elements of :data:`sys.orig_argv` are the arguments to the Python interpreter, + while the elements of :data:`sys.argv` are the arguments to the user's program. + Arguments consumed by the interpreter itself will be present in :data:`sys.orig_argv` + and missing from :data:`sys.argv`. .. versionadded:: 3.10 diff --git a/Doc/library/warnings.rst b/Doc/library/warnings.rst index a9c469707e8227..500398636e11ae 100644 --- a/Doc/library/warnings.rst +++ b/Doc/library/warnings.rst @@ -396,7 +396,7 @@ Available Functions ------------------- -.. function:: warn(message, category=None, stacklevel=1, source=None, \*, skip_file_prefixes=None) +.. function:: warn(message, category=None, stacklevel=1, source=None, *, skip_file_prefixes=None) Issue a warning, or maybe ignore it or raise an exception. The *category* argument, if given, must be a :ref:`warning category class `; it diff --git a/Doc/library/xml.etree.elementtree.rst b/Doc/library/xml.etree.elementtree.rst index fe92400fb08dfd..75a7915c15240d 100644 --- a/Doc/library/xml.etree.elementtree.rst +++ b/Doc/library/xml.etree.elementtree.rst @@ -625,6 +625,8 @@ Functions target. Returns an :term:`iterator` providing ``(event, elem)`` pairs; it has a ``root`` attribute that references the root element of the resulting XML tree once *source* is fully read. + The iterator has the :meth:`!close` method that closes the internal + file object if *source* is a filename. Note that while :func:`iterparse` builds the tree incrementally, it issues blocking reads on *source* (or the file it names). As such, it's unsuitable @@ -647,6 +649,9 @@ Functions .. versionchanged:: 3.8 The ``comment`` and ``pi`` events were added. + .. versionchanged:: 3.13 + Added the :meth:`!close` method. + .. function:: parse(source, parser=None) @@ -664,7 +669,7 @@ Functions given. Returns an element instance, representing a processing instruction. Note that :class:`XMLParser` skips over processing instructions - in the input instead of creating comment objects for them. An + in the input instead of creating PI objects for them. An :class:`ElementTree` will only contain processing instruction nodes if they have been inserted into to the tree using one of the :class:`Element` methods. @@ -1302,8 +1307,8 @@ TreeBuilder Objects .. method:: pi(target, text) - Creates a comment with the given *target* name and *text*. If - ``insert_pis`` is true, this will also add it to the tree. + Creates a process instruction with the given *target* name and *text*. + If ``insert_pis`` is true, this will also add it to the tree. .. versionadded:: 3.8 diff --git a/Doc/tools/.nitignore b/Doc/tools/.nitignore index 7eacb46d6299b3..f96478b45e44c0 100644 --- a/Doc/tools/.nitignore +++ b/Doc/tools/.nitignore @@ -18,7 +18,6 @@ Doc/extending/extending.rst Doc/glossary.rst Doc/howto/descriptor.rst Doc/howto/enum.rst -Doc/howto/logging.rst Doc/library/ast.rst Doc/library/asyncio-extending.rst Doc/library/asyncio-policy.rst @@ -47,7 +46,6 @@ Doc/library/mmap.rst Doc/library/multiprocessing.rst Doc/library/optparse.rst Doc/library/os.rst -Doc/library/pickle.rst Doc/library/pickletools.rst Doc/library/platform.rst Doc/library/plistlib.rst diff --git a/Doc/tutorial/datastructures.rst b/Doc/tutorial/datastructures.rst index 87614d082a1d4e..de2827461e2f24 100644 --- a/Doc/tutorial/datastructures.rst +++ b/Doc/tutorial/datastructures.rst @@ -48,10 +48,9 @@ objects: :noindex: Remove the item at the given position in the list, and return it. If no index - is specified, ``a.pop()`` removes and returns the last item in the list. (The - square brackets around the *i* in the method signature denote that the parameter - is optional, not that you should type square brackets at that position. You - will see this notation frequently in the Python Library Reference.) + is specified, ``a.pop()`` removes and returns the last item in the list. + It raises an :exc:`IndexError` if the list is empty or the index is + outside the list range. .. method:: list.clear() diff --git a/Doc/whatsnew/2.0.rst b/Doc/whatsnew/2.0.rst index f4a9d23699de53..af8171487fbcfa 100644 --- a/Doc/whatsnew/2.0.rst +++ b/Doc/whatsnew/2.0.rst @@ -1039,12 +1039,12 @@ is an implementation of the Secure Socket Layer, which encrypts the data being sent over a socket. When compiling Python, you can edit :file:`Modules/Setup` to include SSL support, which adds an additional function to the :mod:`socket` module: ``socket.ssl(socket, keyfile, certfile)``, which takes a socket -object and returns an SSL socket. The :mod:`httplib` and :mod:`urllib` modules +object and returns an SSL socket. The :mod:`httplib ` and :mod:`urllib` modules were also changed to support ``https://`` URLs, though no one has implemented FTP or SMTP over SSL. -The :mod:`httplib` module has been rewritten by Greg Stein to support HTTP/1.1. -Backward compatibility with the 1.5 version of :mod:`httplib` is provided, +The :mod:`httplib ` module has been rewritten by Greg Stein to support HTTP/1.1. +Backward compatibility with the 1.5 version of :mod:`!httplib` is provided, though using HTTP/1.1 features such as pipelining will require rewriting code to use a different set of interfaces. @@ -1108,7 +1108,7 @@ module. * :mod:`pyexpat`: An interface to the Expat XML parser. (Contributed by Paul Prescod.) -* :mod:`robotparser`: Parse a :file:`robots.txt` file, which is used for writing +* :mod:`robotparser `: Parse a :file:`robots.txt` file, which is used for writing web spiders that politely avoid certain areas of a web site. The parser accepts the contents of a :file:`robots.txt` file, builds a set of rules from it, and can then answer questions about the fetchability of a given URL. (Contributed @@ -1129,10 +1129,10 @@ module. :file:`Tools/idle/BrowserControl.py`, and adapted for the standard library by Fred.) -* :mod:`_winreg`: An interface to the Windows registry. :mod:`_winreg` is an +* :mod:`_winreg `: An interface to the Windows registry. :mod:`!_winreg` is an adaptation of functions that have been part of PythonWin since 1995, but has now been added to the core distribution, and enhanced to support Unicode. - :mod:`_winreg` was written by Bill Tutt and Mark Hammond. + :mod:`!_winreg` was written by Bill Tutt and Mark Hammond. * :mod:`zipfile`: A module for reading and writing ZIP-format archives. These are archives produced by :program:`PKZIP` on DOS/Windows or :program:`zip` on diff --git a/Doc/whatsnew/2.2.rst b/Doc/whatsnew/2.2.rst index 968bd7a126bdf0..e6c13f957b8d54 100644 --- a/Doc/whatsnew/2.2.rst +++ b/Doc/whatsnew/2.2.rst @@ -55,7 +55,7 @@ implemented in C. In particular, it's not possible to subclass built-in types, so you can't just subclass, say, lists in order to add a single useful method to them. The :mod:`!UserList` module provides a class that supports all of the methods of lists and that can be subclassed further, but there's lots of C code -that expects a regular Python list and won't accept a :class:`!UserList` +that expects a regular Python list and won't accept a :class:`~collections.UserList` instance. Python 2.2 fixes this, and in the process adds some exciting new capabilities. @@ -69,7 +69,7 @@ A brief summary: * It's also possible to automatically call methods on accessing or setting an instance attribute by using a new mechanism called :dfn:`properties`. Many uses - of :meth:`!__getattr__` can be rewritten to use properties instead, making the + of :meth:`~object.__getattr__` can be rewritten to use properties instead, making the resulting code simpler and faster. As a small side benefit, attributes can now have docstrings, too. @@ -933,7 +933,7 @@ anyway). New and Improved Modules ======================== -* The :mod:`!xmlrpclib` module was contributed to the standard library by Fredrik +* The :mod:`xmlrpclib ` module was contributed to the standard library by Fredrik Lundh, providing support for writing XML-RPC clients. XML-RPC is a simple remote procedure call protocol built on top of HTTP and XML. For example, the following snippet retrieves a list of RSS channels from the O'Reilly Network, @@ -956,7 +956,7 @@ New and Improved Modules # 'description': 'A utility which converts HTML to XSL FO.', # 'title': 'html2fo 0.3 (Default)'}, ... ] - The :mod:`!SimpleXMLRPCServer` module makes it easy to create straightforward + The :mod:`SimpleXMLRPCServer ` module makes it easy to create straightforward XML-RPC servers. See http://xmlrpc.scripting.com/ for more information about XML-RPC. * The new :mod:`hmac` module implements the HMAC algorithm described by @@ -964,7 +964,7 @@ New and Improved Modules * Several functions that originally returned lengthy tuples now return pseudo-sequences that still behave like tuples but also have mnemonic attributes such - as :attr:`!memberst_mtime` or :attr:`!tm_year`. The enhanced functions include + as :attr:`!memberst_mtime` or :attr:`~time.struct_time.tm_year`. The enhanced functions include :func:`~os.stat`, :func:`~os.fstat`, :func:`~os.statvfs`, and :func:`~os.fstatvfs` in the :mod:`os` module, and :func:`~time.localtime`, :func:`~time.gmtime`, and :func:`~time.strptime` in the :mod:`time` module. diff --git a/Doc/whatsnew/2.4.rst b/Doc/whatsnew/2.4.rst index 15d4003622c506..7e235d4370edaa 100644 --- a/Doc/whatsnew/2.4.rst +++ b/Doc/whatsnew/2.4.rst @@ -1081,7 +1081,7 @@ complete list of changes, or look through the CVS logs for all the details. :func:`nsmallest` that use heaps to find the N largest or smallest values in a dataset without the expense of a full sort. (Contributed by Raymond Hettinger.) -* The :mod:`httplib` module now contains constants for HTTP status codes defined +* The :mod:`httplib ` module now contains constants for HTTP status codes defined in various HTTP-related RFC documents. Constants have names such as :const:`OK`, :const:`CREATED`, :const:`CONTINUE`, and :const:`MOVED_PERMANENTLY`; use pydoc to get a full list. (Contributed by @@ -1218,10 +1218,10 @@ complete list of changes, or look through the CVS logs for all the details. now include the string ``'%default'``, which will be replaced by the option's default value. (Contributed by Greg Ward.) -* The long-term plan is to deprecate the :mod:`rfc822` module in some future +* The long-term plan is to deprecate the :mod:`!rfc822` module in some future Python release in favor of the :mod:`email` package. To this end, the - :func:`email.Utils.formatdate` function has been changed to make it usable as a - replacement for :func:`rfc822.formatdate`. You may want to write new e-mail + :func:`email.Utils.formatdate ` function has been changed to make it usable as a + replacement for :func:`!rfc822.formatdate`. You may want to write new e-mail processing code with this in mind. (Change implemented by Anthony Baxter.) * A new ``urandom(n)`` function was added to the :mod:`os` module, returning @@ -1308,7 +1308,7 @@ complete list of changes, or look through the CVS logs for all the details. sockets, and regular expression pattern objects. (Contributed by Raymond Hettinger.) -* The :mod:`xmlrpclib` module now supports a multi-call extension for +* The :mod:`xmlrpclib ` module now supports a multi-call extension for transmitting multiple XML-RPC calls in a single HTTP operation. (Contributed by Brian Quinlan.) @@ -1323,8 +1323,8 @@ complete list of changes, or look through the CVS logs for all the details. cookielib --------- -The :mod:`cookielib` library supports client-side handling for HTTP cookies, -mirroring the :mod:`Cookie` module's server-side cookie support. Cookies are +The :mod:`cookielib ` library supports client-side handling for HTTP cookies, +mirroring the :mod:`Cookie ` module's server-side cookie support. Cookies are stored in cookie jars; the library transparently stores cookies offered by the web server in the cookie jar, and fetches the cookie from the jar when connecting to the server. As in web browsers, policy objects control whether @@ -1335,7 +1335,7 @@ are provided: one that stores cookies in the Netscape format so applications can use the Mozilla or Lynx cookie files, and one that stores cookies in the same format as the Perl libwww library. -:mod:`urllib2` has been changed to interact with :mod:`cookielib`: +:mod:`urllib2 ` has been changed to interact with :mod:`cookielib `: :class:`HTTPCookieProcessor` manages a cookie jar that is used when accessing URLs. diff --git a/Doc/whatsnew/2.5.rst b/Doc/whatsnew/2.5.rst index f45d70ea5a19a0..2ae26e7a106a0b 100644 --- a/Doc/whatsnew/2.5.rst +++ b/Doc/whatsnew/2.5.rst @@ -1478,8 +1478,8 @@ complete list of changes, or look through the SVN logs for all the details. .. Patch 790710 -* The :mod:`pickle` and :mod:`cPickle` modules no longer accept a return value - of ``None`` from the :meth:`__reduce__` method; the method must return a tuple +* The :mod:`pickle` and :mod:`!cPickle` modules no longer accept a return value + of ``None`` from the :meth:`~object.__reduce__` method; the method must return a tuple of arguments instead. The ability to return ``None`` was deprecated in Python 2.4, so this completes the removal of the feature. @@ -1519,7 +1519,7 @@ complete list of changes, or look through the SVN logs for all the details. .. Patch #1472854 -* The :mod:`SimpleXMLRPCServer` and :mod:`DocXMLRPCServer` classes now have a +* The :mod:`SimpleXMLRPCServer ` and :mod:`DocXMLRPCServer ` classes now have a :attr:`rpc_paths` attribute that constrains XML-RPC operations to a limited set of URL paths; the default is to allow only ``'/'`` and ``'/RPC2'``. Setting :attr:`rpc_paths` to ``None`` or an empty tuple disables this path checking. @@ -1650,9 +1650,9 @@ complete list of changes, or look through the SVN logs for all the details. .. Patch #754022 -* The :mod:`xmlrpclib` module now supports returning :class:`~datetime.datetime` objects - for the XML-RPC date type. Supply ``use_datetime=True`` to the :func:`loads` - function or the :class:`Unmarshaller` class to enable this feature. (Contributed +* The :mod:`xmlrpclib ` module now supports returning :class:`~datetime.datetime` objects + for the XML-RPC date type. Supply ``use_datetime=True`` to the :func:`~xmlrpc.client.loads` + function or the :class:`!Unmarshaller` class to enable this feature. (Contributed by Skip Montanaro.) .. Patch 1120353 @@ -2253,12 +2253,12 @@ code: appeared. In Python 2.5, the argument must be exactly one %char specifier with no surrounding text. -* Library: The :mod:`pickle` and :mod:`cPickle` modules no longer accept a - return value of ``None`` from the :meth:`__reduce__` method; the method must +* Library: The :mod:`pickle` and :mod:`!cPickle` modules no longer accept a + return value of ``None`` from the :meth:`~object.__reduce__` method; the method must return a tuple of arguments instead. The modules also no longer accept the deprecated *bin* keyword parameter. -* Library: The :mod:`SimpleXMLRPCServer` and :mod:`DocXMLRPCServer` classes now +* Library: The :mod:`SimpleXMLRPCServer ` and :mod:`DocXMLRPCServer ` classes now have a :attr:`rpc_paths` attribute that constrains XML-RPC operations to a limited set of URL paths; the default is to allow only ``'/'`` and ``'/RPC2'``. Setting :attr:`rpc_paths` to ``None`` or an empty tuple disables this path diff --git a/Doc/whatsnew/2.6.rst b/Doc/whatsnew/2.6.rst index c6bab93b7efdda..7d3769a22286e2 100644 --- a/Doc/whatsnew/2.6.rst +++ b/Doc/whatsnew/2.6.rst @@ -1082,7 +1082,7 @@ the :mod:`io` module: (In Python 2.6, :class:`io.StringIO` is implemented in pure Python, so it's pretty slow. You should therefore stick with the - existing :mod:`StringIO` module or :mod:`cStringIO` for now. At some + existing :mod:`!StringIO` module or :mod:`!cStringIO` for now. At some point Python 3.0's :mod:`io` module will be rewritten into C for speed, and perhaps the C implementation will be backported to the 2.x releases.) @@ -1807,7 +1807,7 @@ changes, or look through the Subversion logs for all the details. Nubis; :issue:`1817`.) The :func:`parse_qs` and :func:`parse_qsl` functions have been - relocated from the :mod:`!cgi` module to the :mod:`urlparse` module. + relocated from the :mod:`!cgi` module to the :mod:`urlparse ` module. The versions still available in the :mod:`!cgi` module will trigger :exc:`PendingDeprecationWarning` messages in 2.6 (:issue:`600362`). @@ -1895,8 +1895,8 @@ changes, or look through the Subversion logs for all the details. (Contributed by Raymond Hettinger.) -* The :mod:`Cookie` module's :class:`Morsel` objects now support an - :attr:`httponly` attribute. In some browsers. cookies with this attribute +* The :mod:`Cookie ` module's :class:`~http.cookies.Morsel` objects now support an + :attr:`~http.cookies.Morsel.httponly` attribute. In some browsers. cookies with this attribute set cannot be accessed or manipulated by JavaScript code. (Contributed by Arvin Schnell; :issue:`1638033`.) @@ -1987,8 +1987,8 @@ changes, or look through the Subversion logs for all the details. (Contributed by Raymond Hettinger.) * An optional ``timeout`` parameter, specifying a timeout measured in - seconds, was added to the :class:`httplib.HTTPConnection` and - :class:`HTTPSConnection` class constructors. (Added by Facundo + seconds, was added to the :class:`httplib.HTTPConnection ` and + :class:`HTTPSConnection ` class constructors. (Added by Facundo Batista.) * Most of the :mod:`inspect` module's functions, such as @@ -2371,10 +2371,10 @@ changes, or look through the Subversion logs for all the details. ``socket(socket.AF_INET, ...)`` may be all that's required to make your code work with IPv6. -* The base classes in the :mod:`SocketServer` module now support - calling a :meth:`handle_timeout` method after a span of inactivity - specified by the server's :attr:`timeout` attribute. (Contributed - by Michael Pomraning.) The :meth:`serve_forever` method +* The base classes in the :mod:`SocketServer ` module now support + calling a :meth:`~socketserver.BaseServer.handle_timeout` method after a span of inactivity + specified by the server's :attr:`~socketserver.BaseServer.timeout` attribute. (Contributed + by Michael Pomraning.) The :meth:`~socketserver.BaseServer.serve_forever` method now takes an optional poll interval measured in seconds, controlling how often the server will check for a shutdown request. (Contributed by Pedro Werneck and Jeffrey Yasskin; @@ -2478,9 +2478,9 @@ changes, or look through the Subversion logs for all the details. ``with tempfile.NamedTemporaryFile() as tmp: ...``. (Contributed by Alexander Belopolsky; :issue:`2021`.) -* The :mod:`test.test_support` module gained a number +* The :mod:`test.test_support ` module gained a number of context managers useful for writing tests. - :func:`EnvironmentVarGuard` is a + :func:`~test.support.os_helper.EnvironmentVarGuard` is a context manager that temporarily changes environment variables and automatically restores them to their old values. @@ -2577,9 +2577,9 @@ changes, or look through the Subversion logs for all the details. (:issue:`1513695`) * An optional ``timeout`` parameter was added to the - :func:`urllib.urlopen` function and the + :func:`urllib.urlopen ` function and the :class:`urllib.ftpwrapper` class constructor, as well as the - :func:`urllib2.urlopen` function. The parameter specifies a timeout + :func:`urllib2.urlopen ` function. The parameter specifies a timeout measured in seconds. For example:: >>> u = urllib2.urlopen("http://slow.example.com", @@ -2604,7 +2604,7 @@ changes, or look through the Subversion logs for all the details. intended for testing purposes that lets you temporarily modify the warning filters and then restore their original values (:issue:`3781`). -* The XML-RPC :class:`SimpleXMLRPCServer` and :class:`DocXMLRPCServer` +* The XML-RPC :class:`SimpleXMLRPCServer ` and :class:`DocXMLRPCServer ` classes can now be prevented from immediately opening and binding to their socket by passing ``False`` as the *bind_and_activate* constructor parameter. This can be used to modify the instance's @@ -2621,11 +2621,11 @@ changes, or look through the Subversion logs for all the details. information. (Contributed by Alan McIntyre as part of his project for Google's Summer of Code 2007.) -* The :mod:`xmlrpclib` module no longer automatically converts +* The :mod:`xmlrpclib ` module no longer automatically converts :class:`datetime.date` and :class:`datetime.time` to the - :class:`xmlrpclib.DateTime` type; the conversion semantics were + :class:`xmlrpclib.DateTime ` type; the conversion semantics were not necessarily correct for all applications. Code using - :mod:`xmlrpclib` should convert :class:`date` and :class:`~datetime.time` + :mod:`!xmlrpclib` should convert :class:`date` and :class:`~datetime.time` instances. (:issue:`1330538`) The code can also handle dates before 1900 (contributed by Ralf Schmitt; :issue:`2014`) and 64-bit integers represented by using ```` in XML-RPC responses @@ -3274,11 +3274,11 @@ that may require changes to your code: :exc:`StandardError` but now it is, through :exc:`IOError`. (Implemented by Gregory P. Smith; :issue:`1706815`.) -* The :mod:`xmlrpclib` module no longer automatically converts +* The :mod:`xmlrpclib ` module no longer automatically converts :class:`datetime.date` and :class:`datetime.time` to the - :class:`xmlrpclib.DateTime` type; the conversion semantics were + :class:`xmlrpclib.DateTime ` type; the conversion semantics were not necessarily correct for all applications. Code using - :mod:`xmlrpclib` should convert :class:`date` and :class:`~datetime.time` + :mod:`!xmlrpclib` should convert :class:`date` and :class:`~datetime.time` instances. (:issue:`1330538`) * (3.0-warning mode) The :class:`Exception` class now warns diff --git a/Doc/whatsnew/2.7.rst b/Doc/whatsnew/2.7.rst index 524967b4524234..ada05aa22b46f6 100644 --- a/Doc/whatsnew/2.7.rst +++ b/Doc/whatsnew/2.7.rst @@ -915,7 +915,7 @@ used with the :option:`-W` switch, separated by commas. (Contributed by Brian Curtin; :issue:`7301`.) For example, the following setting will print warnings every time -they occur, but turn warnings from the :mod:`Cookie` module into an +they occur, but turn warnings from the :mod:`Cookie ` module into an error. (The exact syntax for setting an environment variable varies across operating systems and shells.) @@ -1012,12 +1012,12 @@ Several performance enhancements have been added: scan. This is sometimes faster by a factor of 10. (Added by Florent Xicluna; :issue:`7462` and :issue:`7622`.) -* The :mod:`pickle` and :mod:`cPickle` modules now automatically +* The :mod:`pickle` and :mod:`!cPickle` modules now automatically intern the strings used for attribute names, reducing memory usage of the objects resulting from unpickling. (Contributed by Jake McGuire; :issue:`5084`.) -* The :mod:`cPickle` module now special-cases dictionaries, +* The :mod:`!cPickle` module now special-cases dictionaries, nearly halving the time required to pickle them. (Contributed by Collin Winter; :issue:`5670`.) @@ -1163,7 +1163,7 @@ changes, or look through the Subversion logs for all the details. statement, has been deprecated, because the :keyword:`!with` statement now supports multiple context managers. -* The :mod:`cookielib` module now ignores cookies that have an invalid +* The :mod:`cookielib ` module now ignores cookies that have an invalid version field, one that doesn't contain an integer value. (Fixed by John J. Lee; :issue:`3924`.) @@ -1306,11 +1306,11 @@ changes, or look through the Subversion logs for all the details. ``('md5', 'sha1', 'sha224', 'sha256', 'sha384', 'sha512')``. (Contributed by Carl Chenet; :issue:`7418`.) -* The default :class:`~httplib.HTTPResponse` class used by the :mod:`httplib` module now +* The default :class:`~http.client.HTTPResponse` class used by the :mod:`httplib ` module now supports buffering, resulting in much faster reading of HTTP responses. (Contributed by Kristján Valur Jónsson; :issue:`4879`.) - The :class:`~httplib.HTTPConnection` and :class:`~httplib.HTTPSConnection` classes + The :class:`~http.client.HTTPConnection` and :class:`~http.client.HTTPSConnection` classes now support a *source_address* parameter, a ``(host, port)`` 2-tuple giving the source address that will be used for the connection. (Contributed by Eldon Ziegler; :issue:`3972`.) @@ -1518,16 +1518,16 @@ changes, or look through the Subversion logs for all the details. the :class:`bytearray` and :class:`memoryview` objects. (Implemented by Antoine Pitrou; :issue:`8104`.) -* The :mod:`SocketServer` module's :class:`~SocketServer.TCPServer` class now +* The :mod:`SocketServer ` module's :class:`~socketserver.TCPServer` class now supports socket timeouts and disabling the Nagle algorithm. - The :attr:`~SocketServer.TCPServer.disable_nagle_algorithm` class attribute + The :attr:`!disable_nagle_algorithm` class attribute defaults to ``False``; if overridden to be true, new request connections will have the TCP_NODELAY option set to prevent buffering many small sends into a single TCP packet. - The :attr:`~SocketServer.BaseServer.timeout` class attribute can hold + The :attr:`~socketserver.BaseServer.timeout` class attribute can hold a timeout in seconds that will be applied to the request socket; if - no request is received within that time, :meth:`~SocketServer.BaseServer.handle_timeout` - will be called and :meth:`~SocketServer.BaseServer.handle_request` will return. + no request is received within that time, :meth:`~socketserver.BaseServer.handle_timeout` + will be called and :meth:`~socketserver.BaseServer.handle_request` will return. (Contributed by Kristján Valur Jónsson; :issue:`6192` and :issue:`6267`.) * Updated module: the :mod:`sqlite3` module has been updated to @@ -1648,7 +1648,7 @@ changes, or look through the Subversion logs for all the details. and has been updated to version 5.2.0 (updated by Florent Xicluna; :issue:`8024`). -* The :mod:`urlparse` module's :func:`~urlparse.urlsplit` now handles +* The :mod:`urlparse ` module's :func:`~urllib.parse.urlsplit` now handles unknown URL schemes in a fashion compliant with :rfc:`3986`: if the URL is of the form ``"://..."``, the text before the ``://`` is treated as the scheme, even if it's a made-up scheme that @@ -1675,7 +1675,7 @@ changes, or look through the Subversion logs for all the details. (Python 2.7 actually produces slightly different output, since it returns a named tuple instead of a standard tuple.) - The :mod:`urlparse` module also supports IPv6 literal addresses as defined by + The :mod:`urlparse ` module also supports IPv6 literal addresses as defined by :rfc:`2732` (contributed by Senthil Kumaran; :issue:`2987`). .. doctest:: @@ -1697,8 +1697,8 @@ changes, or look through the Subversion logs for all the details. or comment (which looks like ````). (Patch by Neil Muller; :issue:`2746`.) -* The XML-RPC client and server, provided by the :mod:`xmlrpclib` and - :mod:`SimpleXMLRPCServer` modules, have improved performance by +* The XML-RPC client and server, provided by the :mod:`xmlrpclib ` and + :mod:`SimpleXMLRPCServer ` modules, have improved performance by supporting HTTP/1.1 keep-alive and by optionally using gzip encoding to compress the XML being exchanged. The gzip compression is controlled by the :attr:`encode_threshold` attribute of @@ -2334,11 +2334,11 @@ Port-Specific Changes: Windows and :data:`LIBRARIES_ASSEMBLY_NAME_PREFIX`. (Contributed by David Cournapeau; :issue:`4365`.) -* The :mod:`_winreg` module for accessing the registry now implements - the :func:`~_winreg.CreateKeyEx` and :func:`~_winreg.DeleteKeyEx` +* The :mod:`_winreg ` module for accessing the registry now implements + the :func:`~winreg.CreateKeyEx` and :func:`~winreg.DeleteKeyEx` functions, extended versions of previously supported functions that - take several extra arguments. The :func:`~_winreg.DisableReflectionKey`, - :func:`~_winreg.EnableReflectionKey`, and :func:`~_winreg.QueryReflectionKey` + take several extra arguments. The :func:`~winreg.DisableReflectionKey`, + :func:`~winreg.EnableReflectionKey`, and :func:`~winreg.QueryReflectionKey` were also tested and documented. (Implemented by Brian Curtin: :issue:`7347`.) @@ -2508,7 +2508,7 @@ In the standard library: which raises an exception if there's an error. (Changed by Lars Gustäbel; :issue:`7357`.) -* The :mod:`urlparse` module's :func:`~urlparse.urlsplit` now handles +* The :mod:`urlparse ` module's :func:`~urllib.parse.urlsplit` now handles unknown URL schemes in a fashion compliant with :rfc:`3986`: if the URL is of the form ``"://..."``, the text before the ``://`` is treated as the scheme, even if it's a made-up scheme that @@ -2711,8 +2711,8 @@ and :ref:`setuptools-index`. PEP 476: Enabling certificate verification by default for stdlib http clients ----------------------------------------------------------------------------- -:pep:`476` updated :mod:`httplib` and modules which use it, such as -:mod:`urllib2` and :mod:`xmlrpclib`, to now verify that the server +:pep:`476` updated :mod:`httplib ` and modules which use it, such as +:mod:`urllib2 ` and :mod:`xmlrpclib`, to now verify that the server presents a certificate which is signed by a Certificate Authority in the platform trust store and whose hostname matches the hostname being requested by default, significantly improving security for many applications. This diff --git a/Doc/whatsnew/3.0.rst b/Doc/whatsnew/3.0.rst index 1df5209f22c6a5..888e6279754fc2 100644 --- a/Doc/whatsnew/3.0.rst +++ b/Doc/whatsnew/3.0.rst @@ -337,7 +337,7 @@ changed. (However, the standard library remains ASCII-only with the exception of contributor names in comments.) -* The :mod:`StringIO` and :mod:`cStringIO` modules are gone. Instead, +* The :mod:`!StringIO` and :mod:`!cStringIO` modules are gone. Instead, import the :mod:`io` module and use :class:`io.StringIO` or :class:`io.BytesIO` for text and data respectively. @@ -563,7 +563,7 @@ review: removal in Python 3.0 due to lack of use or because a better replacement exists. See :pep:`3108` for an exhaustive list. -* The :mod:`bsddb3` package was removed because its presence in the +* The :mod:`!bsddb3` package was removed because its presence in the core standard library has proved over time to be a particular burden for the core developers due to testing instability and Berkeley DB's release schedule. However, the package is alive and well, @@ -588,40 +588,40 @@ review: * A common pattern in Python 2.x is to have one version of a module implemented in pure Python, with an optional accelerated version implemented as a C extension; for example, :mod:`pickle` and - :mod:`cPickle`. This places the burden of importing the accelerated + :mod:`!cPickle`. This places the burden of importing the accelerated version and falling back on the pure Python version on each user of these modules. In Python 3.0, the accelerated versions are considered implementation details of the pure Python versions. Users should always import the standard version, which attempts to import the accelerated version and falls back to the pure Python - version. The :mod:`pickle` / :mod:`cPickle` pair received this + version. The :mod:`pickle` / :mod:`!cPickle` pair received this treatment. The :mod:`profile` module is on the list for 3.1. The - :mod:`StringIO` module has been turned into a class in the :mod:`io` + :mod:`!StringIO` module has been turned into a class in the :mod:`io` module. * Some related modules have been grouped into packages, and usually the submodule names have been simplified. The resulting new packages are: - * :mod:`dbm` (:mod:`anydbm`, :mod:`dbhash`, :mod:`dbm`, - :mod:`dumbdbm`, :mod:`gdbm`, :mod:`whichdb`). + * :mod:`dbm` (:mod:`!anydbm`, :mod:`!dbhash`, :mod:`!dbm`, + :mod:`!dumbdbm`, :mod:`!gdbm`, :mod:`!whichdb`). - * :mod:`html` (:mod:`HTMLParser`, :mod:`htmlentitydefs`). + * :mod:`html` (:mod:`!HTMLParser`, :mod:`!htmlentitydefs`). - * :mod:`http` (:mod:`httplib`, :mod:`BaseHTTPServer`, - :mod:`CGIHTTPServer`, :mod:`SimpleHTTPServer`, :mod:`Cookie`, - :mod:`cookielib`). + * :mod:`http` (:mod:`!httplib`, :mod:`!BaseHTTPServer`, + :mod:`!CGIHTTPServer`, :mod:`!SimpleHTTPServer`, :mod:`!Cookie`, + :mod:`!cookielib`). * :mod:`tkinter` (all :mod:`Tkinter`-related modules except :mod:`turtle`). The target audience of :mod:`turtle` doesn't really care about :mod:`tkinter`. Also note that as of Python 2.6, the functionality of :mod:`turtle` has been greatly enhanced. - * :mod:`urllib` (:mod:`urllib`, :mod:`urllib2`, :mod:`urlparse`, - :mod:`robotparse`). + * :mod:`urllib` (:mod:`!urllib`, :mod:`!urllib2`, :mod:`!urlparse`, + :mod:`!robotparse`). - * :mod:`xmlrpc` (:mod:`xmlrpclib`, :mod:`DocXMLRPCServer`, - :mod:`SimpleXMLRPCServer`). + * :mod:`xmlrpc` (:mod:`!xmlrpclib`, :mod:`!DocXMLRPCServer`, + :mod:`!SimpleXMLRPCServer`). Some other changes to standard library modules, not covered by :pep:`3108`: @@ -642,9 +642,9 @@ Some other changes to standard library modules, not covered by * Cleanup of the :mod:`operator` module: removed :func:`sequenceIncludes` and :func:`isCallable`. -* Cleanup of the :mod:`thread` module: :func:`acquire_lock` and - :func:`release_lock` are gone; use :func:`acquire` and - :func:`release` instead. +* Cleanup of the :mod:`!thread` module: :func:`!acquire_lock` and + :func:`!release_lock` are gone; use :meth:`~threading.Lock.acquire` and + :meth:`~threading.Lock.release` instead. * Cleanup of the :mod:`random` module: removed the :func:`jumpahead` API. diff --git a/Doc/whatsnew/3.13.rst b/Doc/whatsnew/3.13.rst index b33203efbb05c0..77f4fce6c321fe 100644 --- a/Doc/whatsnew/3.13.rst +++ b/Doc/whatsnew/3.13.rst @@ -81,6 +81,13 @@ Important deprecations, removals or restrictions: * Python 3.13 and later have two years of full support, followed by three years of security fixes. +Interpreter improvements: + +* A basic :ref:`JIT compiler ` was added. + It is currently disabled by default (though we may turn it on later). + Performance improvements are modest -- we expect to be improving this + over the next few releases. + New Features ============ @@ -259,8 +266,8 @@ mmap ---- * The :class:`mmap.mmap` class now has an :meth:`~mmap.mmap.seekable` method - that can be used where it requires a file-like object with seekable and - the :meth:`~mmap.mmap.seek` method return the new absolute position. + that can be used when a seekable file-like object is required. + The :meth:`~mmap.mmap.seek` method now returns the new absolute position. (Contributed by Donghee Na and Sylvie Liberman in :gh:`111835`.) * :class:`mmap.mmap` now has a *trackfd* parameter on Unix; if it is ``False``, the file descriptor specified by *fileno* will not be duplicated. @@ -465,6 +472,14 @@ warnings warning may also be emitted when a decorated function or class is used at runtime. See :pep:`702`. (Contributed by Jelle Zijlstra in :gh:`104003`.) +xml.etree.ElementTree +--------------------- + +* Add the :meth:`!close` method for the iterator returned by + :func:`~xml.etree.ElementTree.iterparse` for explicit cleaning up. + (Contributed by Serhiy Storchaka in :gh:`69893`.) + + Optimizations ============= @@ -477,6 +492,46 @@ Optimizations FreeBSD and Solaris. See the ``subprocess`` section above for details. (Contributed by Jakub Kulik in :gh:`113117`.) +.. _whatsnew313-jit-compiler: + +Experimental JIT Compiler +========================= + +When CPython is configured using the ``--enable-experimental-jit`` option, +a just-in-time compiler is added which can speed up some Python programs. + +The internal architecture is roughly as follows. + +* We start with specialized *Tier 1 bytecode*. + See :ref:`What's new in 3.11 ` for details. + +* When the Tier 1 bytecode gets hot enough, it gets translated + to a new, purely internal *Tier 2 IR*, a.k.a. micro-ops ("uops"). + +* The Tier 2 IR uses the same stack-based VM as Tier 1, but the + instruction format is better suited to translation to machine code. + +* We have several optimization passes for Tier 2 IR, which are applied + before it is interpreted or translated to machine code. + +* There is a Tier 2 interpreter, but it is mostly intended for debugging + the earlier stages of the optimization pipeline. If the JIT is not + enabled, the Tier 2 interpreter can be invoked by passing Python the + ``-X uops`` option or by setting the ``PYTHON_UOPS`` environment + variable to ``1``. + +* When the ``--enable-experimental-jit`` option is used, the optimized + Tier 2 IR is translated to machine code, which is then executed. + This does not require additional runtime options. + +* The machine code translation process uses an architecture called + *copy-and-patch*. It has no runtime dependencies, but there is a new + build-time dependency on LLVM. + +(JIT by Brandt Bucher, inspired by a paper by Haoran Xu and Fredrik Kjolstad. +Tier 2 IR by Mark Shannon and Guido van Rossum. +Tier 2 optimizer by Ken Jin.) + Deprecated ========== @@ -1329,6 +1384,10 @@ New Features UTF-8 encoded bytes string, rather than a :c:expr:`PyObject*`. (Contributed by Victor Stinner in :gh:`108314`.) +* Added :c:func:`PyList_GetItemRef` function: similar to + :c:func:`PyList_GetItem` but returns a :term:`strong reference` instead of + a :term:`borrowed reference`. + * Add :c:func:`Py_IsFinalizing` function: check if the main Python interpreter is :term:`shutting down `. (Contributed by Victor Stinner in :gh:`108014`.) diff --git a/Doc/whatsnew/3.5.rst b/Doc/whatsnew/3.5.rst index 1c7a9270af0aab..5c2ec230441b42 100644 --- a/Doc/whatsnew/3.5.rst +++ b/Doc/whatsnew/3.5.rst @@ -2418,7 +2418,7 @@ Changes in the Python API (Contributed by Victor Stinner in :issue:`21205`.) * The deprecated "strict" mode and argument of :class:`~html.parser.HTMLParser`, - :meth:`HTMLParser.error`, and the :exc:`HTMLParserError` exception have been + :meth:`!HTMLParser.error`, and the :exc:`!HTMLParserError` exception have been removed. (Contributed by Ezio Melotti in :issue:`15114`.) The *convert_charrefs* argument of :class:`~html.parser.HTMLParser` is now ``True`` by default. (Contributed by Berker Peksag in :issue:`21047`.) diff --git a/Include/cpython/dictobject.h b/Include/cpython/dictobject.h index 944965fb9e5351..1720fe6f01ea37 100644 --- a/Include/cpython/dictobject.h +++ b/Include/cpython/dictobject.h @@ -17,6 +17,9 @@ typedef struct { /* Dictionary version: globally unique, value change each time the dictionary is modified */ #ifdef Py_BUILD_CORE + /* Bits 0-7 are for dict watchers. + * Bits 8-11 are for the watched mutation counter (used by tier2 optimization) + * The remaining bits (12-63) are the actual version tag. */ uint64_t ma_version_tag; #else Py_DEPRECATED(3.12) uint64_t ma_version_tag; diff --git a/Include/cpython/optimizer.h b/Include/cpython/optimizer.h index ecf3cae4cbc3f1..5a9ccaea3b2209 100644 --- a/Include/cpython/optimizer.h +++ b/Include/cpython/optimizer.h @@ -47,7 +47,10 @@ typedef struct _PyExecutorObject { typedef struct _PyOptimizerObject _PyOptimizerObject; /* Should return > 0 if a new executor is created. O if no executor is produced and < 0 if an error occurred. */ -typedef int (*optimize_func)(_PyOptimizerObject* self, PyCodeObject *code, _Py_CODEUNIT *instr, _PyExecutorObject **, int curr_stackentries); +typedef int (*optimize_func)( + _PyOptimizerObject* self, struct _PyInterpreterFrame *frame, + _Py_CODEUNIT *instr, _PyExecutorObject **exec_ptr, + int curr_stackentries); typedef struct _PyOptimizerObject { PyObject_HEAD @@ -94,6 +97,9 @@ PyAPI_FUNC(PyObject *)PyUnstable_Optimizer_NewUOpOptimizer(void); /* Minimum of 16 additional executions before retry */ #define MINIMUM_TIER2_BACKOFF 4 +#define _Py_MAX_ALLOWED_BUILTINS_MODIFICATIONS 3 +#define _Py_MAX_ALLOWED_GLOBALS_MODIFICATIONS 6 + #ifdef __cplusplus } #endif diff --git a/Include/cpython/pystate.h b/Include/cpython/pystate.h index 1dbf97660f382f..9bc8758e72bd8f 100644 --- a/Include/cpython/pystate.h +++ b/Include/cpython/pystate.h @@ -229,7 +229,7 @@ struct _ts { #elif defined(__s390x__) # define Py_C_RECURSION_LIMIT 800 #elif defined(_WIN32) -# define Py_C_RECURSION_LIMIT 4000 +# define Py_C_RECURSION_LIMIT 3000 #elif defined(_Py_ADDRESS_SANITIZER) # define Py_C_RECURSION_LIMIT 4000 #else diff --git a/Include/internal/pycore_condvar.h b/Include/internal/pycore_condvar.h index 34c21aaad43197..ee9533484e8048 100644 --- a/Include/internal/pycore_condvar.h +++ b/Include/internal/pycore_condvar.h @@ -35,14 +35,14 @@ #include // CRITICAL_SECTION /* options */ -/* non-emulated condition variables are provided for those that want - * to target Windows Vista. Modify this macro to enable them. +/* emulated condition variables are provided for those that want + * to target Windows XP or earlier. Modify this macro to enable them. */ #ifndef _PY_EMULATED_WIN_CV -#define _PY_EMULATED_WIN_CV 1 /* use emulated condition variables */ +#define _PY_EMULATED_WIN_CV 0 /* use non-emulated condition variables */ #endif -/* fall back to emulation if not targeting Vista */ +/* fall back to emulation if targeting earlier than Vista */ #if !defined NTDDI_VISTA || NTDDI_VERSION < NTDDI_VISTA #undef _PY_EMULATED_WIN_CV #define _PY_EMULATED_WIN_CV 1 @@ -77,7 +77,7 @@ typedef struct _PyCOND_T #else /* !_PY_EMULATED_WIN_CV */ -/* Use native Win7 primitives if build target is Win7 or higher */ +/* Use native Windows primitives if build target is Vista or higher */ /* SRWLOCK is faster and better than CriticalSection */ typedef SRWLOCK PyMUTEX_T; diff --git a/Include/internal/pycore_dict.h b/Include/internal/pycore_dict.h index b4e1f8cf1e320b..233da058f464d1 100644 --- a/Include/internal/pycore_dict.h +++ b/Include/internal/pycore_dict.h @@ -9,6 +9,7 @@ extern "C" { # error "this header requires Py_BUILD_CORE define" #endif +#include "pycore_freelist.h" // _PyFreeListState #include "pycore_identifier.h" // _Py_Identifier #include "pycore_object.h" // PyDictOrValues @@ -69,7 +70,7 @@ extern PyObject* _PyDictView_Intersect(PyObject* self, PyObject *other); /* runtime lifecycle */ -extern void _PyDict_Fini(PyInterpreterState *interp); +extern void _PyDict_Fini(PyInterpreterState *state); /* other API */ @@ -206,8 +207,8 @@ static inline PyDictUnicodeEntry* DK_UNICODE_ENTRIES(PyDictKeysObject *dk) { #define DK_IS_UNICODE(dk) ((dk)->dk_kind != DICT_KEYS_GENERAL) -#define DICT_VERSION_INCREMENT (1 << DICT_MAX_WATCHERS) -#define DICT_VERSION_MASK (DICT_VERSION_INCREMENT - 1) +#define DICT_VERSION_INCREMENT (1 << (DICT_MAX_WATCHERS + DICT_WATCHED_MUTATION_BITS)) +#define DICT_WATCHER_MASK ((1 << DICT_MAX_WATCHERS) - 1) #ifdef Py_GIL_DISABLED #define DICT_NEXT_VERSION(INTERP) \ @@ -233,7 +234,7 @@ _PyDict_NotifyEvent(PyInterpreterState *interp, PyObject *value) { assert(Py_REFCNT((PyObject*)mp) > 0); - int watcher_bits = mp->ma_version_tag & DICT_VERSION_MASK; + int watcher_bits = mp->ma_version_tag & DICT_WATCHER_MASK; if (watcher_bits) { _PyDict_SendEvent(watcher_bits, event, mp, key, value); return DICT_NEXT_VERSION(interp) | watcher_bits; diff --git a/Include/internal/pycore_dict_state.h b/Include/internal/pycore_dict_state.h index ece0f10ca25170..1a44755c7a01a3 100644 --- a/Include/internal/pycore_dict_state.h +++ b/Include/internal/pycore_dict_state.h @@ -8,17 +8,8 @@ extern "C" { # error "this header requires Py_BUILD_CORE define" #endif - -#ifndef WITH_FREELISTS -// without freelists -# define PyDict_MAXFREELIST 0 -#endif - -#ifndef PyDict_MAXFREELIST -# define PyDict_MAXFREELIST 80 -#endif - #define DICT_MAX_WATCHERS 8 +#define DICT_WATCHED_MUTATION_BITS 4 struct _Py_dict_state { /*Global counter used to set ma_version_tag field of dictionary. @@ -26,15 +17,6 @@ struct _Py_dict_state { * time that a dictionary is modified. */ uint64_t global_version; uint32_t next_keys_version; - -#if PyDict_MAXFREELIST > 0 - /* Dictionary reuse scheme to save calls to malloc and free */ - PyDictObject *free_list[PyDict_MAXFREELIST]; - PyDictKeysObject *keys_free_list[PyDict_MAXFREELIST]; - int numfree; - int keys_numfree; -#endif - PyDict_WatchCallback watchers[DICT_MAX_WATCHERS]; }; diff --git a/Include/internal/pycore_freelist.h b/Include/internal/pycore_freelist.h index b91d2bc066b783..82a42300991ecc 100644 --- a/Include/internal/pycore_freelist.h +++ b/Include/internal/pycore_freelist.h @@ -17,6 +17,7 @@ extern "C" { # define PyTuple_NFREELISTS PyTuple_MAXSAVESIZE # define PyTuple_MAXFREELIST 2000 # define PyList_MAXFREELIST 80 +# define PyDict_MAXFREELIST 80 # define PyFloat_MAXFREELIST 100 # define PyContext_MAXFREELIST 255 # define _PyAsyncGen_MAXFREELIST 80 @@ -25,6 +26,7 @@ extern "C" { # define PyTuple_NFREELISTS 0 # define PyTuple_MAXFREELIST 0 # define PyList_MAXFREELIST 0 +# define PyDict_MAXFREELIST 0 # define PyFloat_MAXFREELIST 0 # define PyContext_MAXFREELIST 0 # define _PyAsyncGen_MAXFREELIST 0 @@ -65,6 +67,16 @@ struct _Py_float_state { #endif }; +struct _Py_dict_freelist { +#ifdef WITH_FREELISTS + /* Dictionary reuse scheme to save calls to malloc and free */ + PyDictObject *free_list[PyDict_MAXFREELIST]; + PyDictKeysObject *keys_free_list[PyDict_MAXFREELIST]; + int numfree; + int keys_numfree; +#endif +}; + struct _Py_slice_state { #ifdef WITH_FREELISTS /* Using a cache is very effective since typically only a single slice is @@ -106,6 +118,7 @@ typedef struct _Py_freelist_state { struct _Py_float_state floats; struct _Py_tuple_state tuples; struct _Py_list_state lists; + struct _Py_dict_freelist dicts; struct _Py_slice_state slices; struct _Py_context_state contexts; struct _Py_async_gen_state async_gens; diff --git a/Include/internal/pycore_gc.h b/Include/internal/pycore_gc.h index b362a294a59042..ca1d9fdf5253b8 100644 --- a/Include/internal/pycore_gc.h +++ b/Include/internal/pycore_gc.h @@ -267,7 +267,7 @@ extern void _PyTuple_ClearFreeList(_PyFreeListState *state, int is_finalization) extern void _PyFloat_ClearFreeList(_PyFreeListState *state, int is_finalization); extern void _PyList_ClearFreeList(_PyFreeListState *state, int is_finalization); extern void _PySlice_ClearCache(_PyFreeListState *state); -extern void _PyDict_ClearFreeList(PyInterpreterState *interp); +extern void _PyDict_ClearFreeList(_PyFreeListState *state, int is_finalization); extern void _PyAsyncGen_ClearFreeLists(_PyFreeListState *state, int is_finalization); extern void _PyContext_ClearFreeList(_PyFreeListState *state, int is_finalization); extern void _Py_ScheduleGC(PyInterpreterState *interp); diff --git a/Include/internal/pycore_interp.h b/Include/internal/pycore_interp.h index 04e75940dcb573..f7c332ed747cfa 100644 --- a/Include/internal/pycore_interp.h +++ b/Include/internal/pycore_interp.h @@ -20,6 +20,7 @@ extern "C" { #include "pycore_dtoa.h" // struct _dtoa_state #include "pycore_exceptions.h" // struct _Py_exc_state #include "pycore_floatobject.h" // struct _Py_float_state +#include "pycore_freelist.h" // struct _Py_freelist_state #include "pycore_function.h" // FUNC_MAX_WATCHERS #include "pycore_gc.h" // struct _gc_runtime_state #include "pycore_genobject.h" // struct _Py_async_gen_state @@ -71,7 +72,6 @@ typedef struct _rare_events { uint8_t set_eval_frame_func; /* Modifying the builtins, __builtins__.__dict__[var] = ... */ uint8_t builtin_dict; - int builtins_dict_watcher_id; /* Modifying a function, e.g. func.__defaults__ = ..., etc. */ uint8_t func_modification; } _rare_events; @@ -230,7 +230,6 @@ struct _is { struct _dtoa_state dtoa; struct _py_func_state func_state; - struct _Py_tuple_state tuple; struct _Py_dict_state dict_state; struct _Py_exc_state exc_state; @@ -243,6 +242,7 @@ struct _is { uint16_t optimizer_backedge_threshold; uint32_t next_func_version; _rare_events rare_events; + PyDict_WatchCallback builtins_dict_watcher; _Py_GlobalMonitors monitors; bool sys_profile_initialized; diff --git a/Include/internal/pycore_list.h b/Include/internal/pycore_list.h index 6c29d882335512..4536f90e414493 100644 --- a/Include/internal/pycore_list.h +++ b/Include/internal/pycore_list.h @@ -24,12 +24,13 @@ extern void _PyList_Fini(_PyFreeListState *); extern int _PyList_AppendTakeRefListResize(PyListObject *self, PyObject *newitem); +// In free-threaded build: self should be locked by the caller, if it should be thread-safe. static inline int _PyList_AppendTakeRef(PyListObject *self, PyObject *newitem) { assert(self != NULL && newitem != NULL); assert(PyList_Check(self)); - Py_ssize_t len = PyList_GET_SIZE(self); + Py_ssize_t len = Py_SIZE(self); Py_ssize_t allocated = self->allocated; assert((size_t)len + 1 < PY_SSIZE_T_MAX); if (allocated > len) { diff --git a/Include/internal/pycore_object.h b/Include/internal/pycore_object.h index e32ea2f528940a..34a83ea228e8b1 100644 --- a/Include/internal/pycore_object.h +++ b/Include/internal/pycore_object.h @@ -315,16 +315,15 @@ static inline void _PyObject_GC_TRACK( _PyObject_ASSERT_FROM(op, !_PyObject_GC_IS_TRACKED(op), "object already tracked by the garbage collector", filename, lineno, __func__); - +#ifdef Py_GIL_DISABLED + op->ob_gc_bits |= _PyGC_BITS_TRACKED; +#else PyGC_Head *gc = _Py_AS_GC(op); _PyObject_ASSERT_FROM(op, (gc->_gc_prev & _PyGC_PREV_MASK_COLLECTING) == 0, "object is in generation which is garbage collected", filename, lineno, __func__); -#ifdef Py_GIL_DISABLED - op->ob_gc_bits |= _PyGC_BITS_TRACKED; -#else PyInterpreterState *interp = _PyInterpreterState_GET(); PyGC_Head *generation0 = interp->gc.generation0; PyGC_Head *last = (PyGC_Head*)(generation0->_gc_prev); @@ -594,8 +593,12 @@ _PyObject_IS_GC(PyObject *obj) static inline size_t _PyType_PreHeaderSize(PyTypeObject *tp) { - return _PyType_IS_GC(tp) * sizeof(PyGC_Head) + - _PyType_HasFeature(tp, Py_TPFLAGS_PREHEADER) * 2 * sizeof(PyObject *); + return ( +#ifndef Py_GIL_DISABLED + _PyType_IS_GC(tp) * sizeof(PyGC_Head) + +#endif + _PyType_HasFeature(tp, Py_TPFLAGS_PREHEADER) * 2 * sizeof(PyObject *) + ); } void _PyObject_GC_Link(PyObject *op); @@ -625,6 +628,14 @@ extern int _PyObject_StoreInstanceAttribute(PyObject *obj, PyDictValues *values, PyObject * _PyObject_GetInstanceAttribute(PyObject *obj, PyDictValues *values, PyObject *name); +#ifdef Py_GIL_DISABLED +# define MANAGED_DICT_OFFSET (((Py_ssize_t)sizeof(PyObject *))*-1) +# define MANAGED_WEAKREF_OFFSET (((Py_ssize_t)sizeof(PyObject *))*-2) +#else +# define MANAGED_DICT_OFFSET (((Py_ssize_t)sizeof(PyObject *))*-3) +# define MANAGED_WEAKREF_OFFSET (((Py_ssize_t)sizeof(PyObject *))*-4) +#endif + typedef union { PyObject *dict; /* Use a char* to generate a warning if directly assigning a PyDictValues */ @@ -635,7 +646,7 @@ static inline PyDictOrValues * _PyObject_DictOrValuesPointer(PyObject *obj) { assert(Py_TYPE(obj)->tp_flags & Py_TPFLAGS_MANAGED_DICT); - return ((PyDictOrValues *)obj)-3; + return (PyDictOrValues *)((char *)obj + MANAGED_DICT_OFFSET); } static inline int @@ -664,8 +675,6 @@ _PyDictOrValues_SetValues(PyDictOrValues *ptr, PyDictValues *values) ptr->values = ((char *)values) - 1; } -#define MANAGED_WEAKREF_OFFSET (((Py_ssize_t)sizeof(PyObject *))*-4) - extern PyObject ** _PyObject_ComputedDictPointer(PyObject *); extern void _PyObject_FreeInstanceAttributes(PyObject *obj); extern int _PyObject_IsInstanceDictEmpty(PyObject *); diff --git a/Include/internal/pycore_optimizer.h b/Include/internal/pycore_optimizer.h index 31f30c673f207a..e21412fc815540 100644 --- a/Include/internal/pycore_optimizer.h +++ b/Include/internal/pycore_optimizer.h @@ -8,8 +8,9 @@ extern "C" { # error "this header requires Py_BUILD_CORE define" #endif -int _Py_uop_analyze_and_optimize(PyCodeObject *code, - _PyUOpInstruction *trace, int trace_len, int curr_stackentries); +int _Py_uop_analyze_and_optimize(_PyInterpreterFrame *frame, + _PyUOpInstruction *trace, int trace_len, int curr_stackentries, + _PyBloomFilter *dependencies); extern PyTypeObject _PyCounterExecutor_Type; extern PyTypeObject _PyCounterOptimizer_Type; diff --git a/Include/internal/pycore_runtime.h b/Include/internal/pycore_runtime.h index 02ab22b967b38f..7c705d1224f915 100644 --- a/Include/internal/pycore_runtime.h +++ b/Include/internal/pycore_runtime.h @@ -268,7 +268,7 @@ typedef struct pyruntimestate { a pointer type. */ - /* PyInterpreterState.interpreters.main */ + /* _PyRuntimeState.interpreters.main */ PyInterpreterState _main_interpreter; #if defined(__EMSCRIPTEN__) && defined(PY_CALL_TRAMPOLINE) diff --git a/Include/internal/pycore_uop_ids.h b/Include/internal/pycore_uop_ids.h index a7056586ff04c0..b2476e1c6e5c4b 100644 --- a/Include/internal/pycore_uop_ids.h +++ b/Include/internal/pycore_uop_ids.h @@ -232,8 +232,12 @@ extern "C" { #define _CHECK_VALIDITY 379 #define _LOAD_CONST_INLINE 380 #define _LOAD_CONST_INLINE_BORROW 381 -#define _INTERNAL_INCREMENT_OPT_COUNTER 382 -#define MAX_UOP_ID 382 +#define _LOAD_CONST_INLINE_WITH_NULL 382 +#define _LOAD_CONST_INLINE_BORROW_WITH_NULL 383 +#define _CHECK_GLOBALS 384 +#define _CHECK_BUILTINS 385 +#define _INTERNAL_INCREMENT_OPT_COUNTER 386 +#define MAX_UOP_ID 386 #ifdef __cplusplus } diff --git a/Include/internal/pycore_uop_metadata.h b/Include/internal/pycore_uop_metadata.h index 14d3382e895cdf..2b5b37e6b8d6a4 100644 --- a/Include/internal/pycore_uop_metadata.h +++ b/Include/internal/pycore_uop_metadata.h @@ -204,6 +204,10 @@ const uint16_t _PyUop_Flags[MAX_UOP_ID+1] = { [_CHECK_VALIDITY] = HAS_DEOPT_FLAG, [_LOAD_CONST_INLINE] = 0, [_LOAD_CONST_INLINE_BORROW] = 0, + [_LOAD_CONST_INLINE_WITH_NULL] = 0, + [_LOAD_CONST_INLINE_BORROW_WITH_NULL] = 0, + [_CHECK_GLOBALS] = HAS_DEOPT_FLAG, + [_CHECK_BUILTINS] = HAS_DEOPT_FLAG, [_INTERNAL_INCREMENT_OPT_COUNTER] = 0, }; @@ -250,10 +254,12 @@ const char *const _PyOpcode_uop_name[MAX_UOP_ID+1] = { [_CHECK_ATTR_METHOD_LAZY_DICT] = "_CHECK_ATTR_METHOD_LAZY_DICT", [_CHECK_ATTR_MODULE] = "_CHECK_ATTR_MODULE", [_CHECK_ATTR_WITH_HINT] = "_CHECK_ATTR_WITH_HINT", + [_CHECK_BUILTINS] = "_CHECK_BUILTINS", [_CHECK_CALL_BOUND_METHOD_EXACT_ARGS] = "_CHECK_CALL_BOUND_METHOD_EXACT_ARGS", [_CHECK_EG_MATCH] = "_CHECK_EG_MATCH", [_CHECK_EXC_MATCH] = "_CHECK_EXC_MATCH", [_CHECK_FUNCTION_EXACT_ARGS] = "_CHECK_FUNCTION_EXACT_ARGS", + [_CHECK_GLOBALS] = "_CHECK_GLOBALS", [_CHECK_MANAGED_OBJECT_HAS_VALUES] = "_CHECK_MANAGED_OBJECT_HAS_VALUES", [_CHECK_PEP_523] = "_CHECK_PEP_523", [_CHECK_STACK_SPACE] = "_CHECK_STACK_SPACE", @@ -332,6 +338,8 @@ const char *const _PyOpcode_uop_name[MAX_UOP_ID+1] = { [_LOAD_CONST] = "_LOAD_CONST", [_LOAD_CONST_INLINE] = "_LOAD_CONST_INLINE", [_LOAD_CONST_INLINE_BORROW] = "_LOAD_CONST_INLINE_BORROW", + [_LOAD_CONST_INLINE_BORROW_WITH_NULL] = "_LOAD_CONST_INLINE_BORROW_WITH_NULL", + [_LOAD_CONST_INLINE_WITH_NULL] = "_LOAD_CONST_INLINE_WITH_NULL", [_LOAD_DEREF] = "_LOAD_DEREF", [_LOAD_FAST] = "_LOAD_FAST", [_LOAD_FAST_AND_CLEAR] = "_LOAD_FAST_AND_CLEAR", diff --git a/Include/listobject.h b/Include/listobject.h index 6b7041ba0b05d5..4e4084b43483a2 100644 --- a/Include/listobject.h +++ b/Include/listobject.h @@ -29,6 +29,7 @@ PyAPI_FUNC(PyObject *) PyList_New(Py_ssize_t size); PyAPI_FUNC(Py_ssize_t) PyList_Size(PyObject *); PyAPI_FUNC(PyObject *) PyList_GetItem(PyObject *, Py_ssize_t); +PyAPI_FUNC(PyObject *) PyList_GetItemRef(PyObject *, Py_ssize_t); PyAPI_FUNC(int) PyList_SetItem(PyObject *, Py_ssize_t, PyObject *); PyAPI_FUNC(int) PyList_Insert(PyObject *, Py_ssize_t, PyObject *); PyAPI_FUNC(int) PyList_Append(PyObject *, PyObject *); diff --git a/Include/object.h b/Include/object.h index ef3fb721c2b012..05187fe5dc4f20 100644 --- a/Include/object.h +++ b/Include/object.h @@ -212,8 +212,9 @@ struct _object { struct _PyMutex { uint8_t v; }; struct _object { - // ob_tid stores the thread id (or zero). It is also used by the GC to - // store linked lists and the computed "gc_refs" refcount. + // ob_tid stores the thread id (or zero). It is also used by the GC and the + // trashcan mechanism as a linked list pointer and by the GC to store the + // computed "gc_refs" refcount. uintptr_t ob_tid; uint16_t _padding; struct _PyMutex ob_mutex; // per-object lock @@ -428,7 +429,11 @@ static inline void Py_SET_TYPE(PyObject *ob, PyTypeObject *type) { static inline void Py_SET_SIZE(PyVarObject *ob, Py_ssize_t size) { assert(ob->ob_base.ob_type != &PyLong_Type); assert(ob->ob_base.ob_type != &PyBool_Type); +#ifdef Py_GIL_DISABLED + _Py_atomic_store_ssize_relaxed(&ob->ob_size, size); +#else ob->ob_size = size; +#endif } #if !defined(Py_LIMITED_API) || Py_LIMITED_API+0 < 0x030b0000 # define Py_SET_SIZE(ob, size) Py_SET_SIZE(_PyVarObject_CAST(ob), (size)) diff --git a/Lib/argparse.py b/Lib/argparse.py index c7c5f41fa48f66..243a73d9766013 100644 --- a/Lib/argparse.py +++ b/Lib/argparse.py @@ -564,22 +564,18 @@ def _format_action_invocation(self, action): return metavar else: - parts = [] # if the Optional doesn't take a value, format is: # -s, --long if action.nargs == 0: - parts.extend(action.option_strings) + return ', '.join(action.option_strings) # if the Optional takes a value, format is: - # -s ARGS, --long ARGS + # -s, --long ARGS else: default = self._get_default_metavar_for_optional(action) args_string = self._format_args(action, default) - for option_string in action.option_strings: - parts.append('%s %s' % (option_string, args_string)) - - return ', '.join(parts) + return ', '.join(action.option_strings) + ' ' + args_string def _metavar_formatter(self, action, default_metavar): if action.metavar is not None: diff --git a/Lib/asyncio/base_events.py b/Lib/asyncio/base_events.py index c60d7688ef8c77..aadc4f478f8b56 100644 --- a/Lib/asyncio/base_events.py +++ b/Lib/asyncio/base_events.py @@ -1340,9 +1340,9 @@ async def create_datagram_endpoint(self, protocol_factory, allow_broadcast=None, sock=None): """Create datagram connection.""" if sock is not None: - if sock.type != socket.SOCK_DGRAM: + if sock.type == socket.SOCK_STREAM: raise ValueError( - f'A UDP Socket was expected, got {sock!r}') + f'A datagram socket was expected, got {sock!r}') if (local_addr or remote_addr or family or proto or flags or reuse_port or allow_broadcast): diff --git a/Lib/asyncio/locks.py b/Lib/asyncio/locks.py index 04158e667a895f..aaee8ff0702923 100644 --- a/Lib/asyncio/locks.py +++ b/Lib/asyncio/locks.py @@ -24,25 +24,23 @@ class Lock(_ContextManagerMixin, mixins._LoopBoundMixin): """Primitive lock objects. A primitive lock is a synchronization primitive that is not owned - by a particular coroutine when locked. A primitive lock is in one + by a particular task when locked. A primitive lock is in one of two states, 'locked' or 'unlocked'. It is created in the unlocked state. It has two basic methods, acquire() and release(). When the state is unlocked, acquire() changes the state to locked and returns immediately. When the state is locked, acquire() blocks until a call to release() in - another coroutine changes it to unlocked, then the acquire() call + another task changes it to unlocked, then the acquire() call resets it to locked and returns. The release() method should only be called in the locked state; it changes the state to unlocked and returns immediately. If an attempt is made to release an unlocked lock, a RuntimeError will be raised. - When more than one coroutine is blocked in acquire() waiting for - the state to turn to unlocked, only one coroutine proceeds when a - release() call resets the state to unlocked; first coroutine which - is blocked in acquire() is being processed. - - acquire() is a coroutine and should be called with 'await'. + When more than one task is blocked in acquire() waiting for + the state to turn to unlocked, only one task proceeds when a + release() call resets the state to unlocked; successive release() + calls will unblock tasks in FIFO order. Locks also support the asynchronous context management protocol. 'async with lock' statement should be used. @@ -130,7 +128,7 @@ def release(self): """Release a lock. When the lock is locked, reset it to unlocked, and return. - If any other coroutines are blocked waiting for the lock to become + If any other tasks are blocked waiting for the lock to become unlocked, allow exactly one of them to proceed. When invoked on an unlocked lock, a RuntimeError is raised. @@ -182,8 +180,8 @@ def is_set(self): return self._value def set(self): - """Set the internal flag to true. All coroutines waiting for it to - become true are awakened. Coroutine that call wait() once the flag is + """Set the internal flag to true. All tasks waiting for it to + become true are awakened. Tasks that call wait() once the flag is true will not block at all. """ if not self._value: @@ -194,7 +192,7 @@ def set(self): fut.set_result(True) def clear(self): - """Reset the internal flag to false. Subsequently, coroutines calling + """Reset the internal flag to false. Subsequently, tasks calling wait() will block until set() is called to set the internal flag to true again.""" self._value = False @@ -203,7 +201,7 @@ async def wait(self): """Block until the internal flag is true. If the internal flag is true on entry, return True - immediately. Otherwise, block until another coroutine calls + immediately. Otherwise, block until another task calls set() to set the flag to true, then return True. """ if self._value: @@ -222,8 +220,8 @@ class Condition(_ContextManagerMixin, mixins._LoopBoundMixin): """Asynchronous equivalent to threading.Condition. This class implements condition variable objects. A condition variable - allows one or more coroutines to wait until they are notified by another - coroutine. + allows one or more tasks to wait until they are notified by another + task. A new Lock object is created and used as the underlying lock. """ @@ -250,50 +248,64 @@ def __repr__(self): async def wait(self): """Wait until notified. - If the calling coroutine has not acquired the lock when this + If the calling task has not acquired the lock when this method is called, a RuntimeError is raised. This method releases the underlying lock, and then blocks until it is awakened by a notify() or notify_all() call for - the same condition variable in another coroutine. Once + the same condition variable in another task. Once awakened, it re-acquires the lock and returns True. + + This method may return spuriously, + which is why the caller should always + re-check the state and be prepared to wait() again. """ if not self.locked(): raise RuntimeError('cannot wait on un-acquired lock') + fut = self._get_loop().create_future() self.release() try: - fut = self._get_loop().create_future() - self._waiters.append(fut) try: - await fut - return True - finally: - self._waiters.remove(fut) - - finally: - # Must re-acquire lock even if wait is cancelled. - # We only catch CancelledError here, since we don't want any - # other (fatal) errors with the future to cause us to spin. - err = None - while True: - try: - await self.acquire() - break - except exceptions.CancelledError as e: - err = e - - if err: + self._waiters.append(fut) try: - raise err # Re-raise most recent exception instance. + await fut + return True finally: - err = None # Break reference cycles. + self._waiters.remove(fut) + + finally: + # Must re-acquire lock even if wait is cancelled. + # We only catch CancelledError here, since we don't want any + # other (fatal) errors with the future to cause us to spin. + err = None + while True: + try: + await self.acquire() + break + except exceptions.CancelledError as e: + err = e + + if err is not None: + try: + raise err # Re-raise most recent exception instance. + finally: + err = None # Break reference cycles. + except BaseException: + # Any error raised out of here _may_ have occurred after this Task + # believed to have been successfully notified. + # Make sure to notify another Task instead. This may result + # in a "spurious wakeup", which is allowed as part of the + # Condition Variable protocol. + self._notify(1) + raise async def wait_for(self, predicate): """Wait until a predicate becomes true. - The predicate should be a callable which result will be - interpreted as a boolean value. The final predicate value is + The predicate should be a callable whose result will be + interpreted as a boolean value. The method will repeatedly + wait() until it evaluates to true. The final predicate value is the return value. """ result = predicate() @@ -303,20 +315,22 @@ async def wait_for(self, predicate): return result def notify(self, n=1): - """By default, wake up one coroutine waiting on this condition, if any. - If the calling coroutine has not acquired the lock when this method + """By default, wake up one task waiting on this condition, if any. + If the calling task has not acquired the lock when this method is called, a RuntimeError is raised. - This method wakes up at most n of the coroutines waiting for the - condition variable; it is a no-op if no coroutines are waiting. + This method wakes up n of the tasks waiting for the condition + variable; if fewer than n are waiting, they are all awoken. - Note: an awakened coroutine does not actually return from its + Note: an awakened task does not actually return from its wait() call until it can reacquire the lock. Since notify() does not release the lock, its caller should. """ if not self.locked(): raise RuntimeError('cannot notify on un-acquired lock') + self._notify(n) + def _notify(self, n): idx = 0 for fut in self._waiters: if idx >= n: @@ -374,7 +388,7 @@ async def acquire(self): If the internal counter is larger than zero on entry, decrement it by one and return True immediately. If it is - zero on entry, block, waiting until some other coroutine has + zero on entry, block, waiting until some other task has called release() to make it larger than 0, and then return True. """ @@ -414,8 +428,8 @@ async def acquire(self): def release(self): """Release a semaphore, incrementing the internal counter by one. - When it was zero on entry and another coroutine is waiting for it to - become larger than zero again, wake up that coroutine. + When it was zero on entry and another task is waiting for it to + become larger than zero again, wake up that task. """ self._value += 1 self._wake_up_next() diff --git a/Lib/asyncio/selector_events.py b/Lib/asyncio/selector_events.py index dcd5e0aa345029..10fbdd76e93f79 100644 --- a/Lib/asyncio/selector_events.py +++ b/Lib/asyncio/selector_events.py @@ -235,6 +235,10 @@ async def _accept_connection2( await waiter except BaseException: transport.close() + # gh-109534: When an exception is raised by the SSLProtocol object the + # exception set in this future can keep the protocol object alive and + # cause a reference cycle. + waiter = None raise # It's now up to the protocol to handle the connection. diff --git a/Lib/asyncio/sslproto.py b/Lib/asyncio/sslproto.py index 599e91ba0003d1..fa99d4533aa0a6 100644 --- a/Lib/asyncio/sslproto.py +++ b/Lib/asyncio/sslproto.py @@ -579,6 +579,7 @@ def _on_handshake_complete(self, handshake_exc): peercert = sslobj.getpeercert() except Exception as exc: + handshake_exc = None self._set_state(SSLProtocolState.UNWRAPPED) if isinstance(exc, ssl.CertificateError): msg = 'SSL handshake failed on verifying the certificate' diff --git a/Lib/dbm/dumb.py b/Lib/dbm/dumb.py index 754624ccc8f500..def120ffc3778b 100644 --- a/Lib/dbm/dumb.py +++ b/Lib/dbm/dumb.py @@ -98,7 +98,8 @@ def _update(self, flag): except OSError: if flag not in ('c', 'n'): raise - self._modified = True + with self._io.open(self._dirfile, 'w', encoding="Latin-1") as f: + self._chmod(self._dirfile) else: with f: for line in f: @@ -134,6 +135,7 @@ def _commit(self): # position; UTF-8, though, does care sometimes. entry = "%r, %r\n" % (key.decode('Latin-1'), pos_and_siz_pair) f.write(entry) + self._modified = False sync = _commit diff --git a/Lib/ensurepip/__init__.py b/Lib/ensurepip/__init__.py index 80ee125cfd4ed3..e8dd253bb55520 100644 --- a/Lib/ensurepip/__init__.py +++ b/Lib/ensurepip/__init__.py @@ -10,7 +10,7 @@ __all__ = ["version", "bootstrap"] -_PIP_VERSION = "23.3.2" +_PIP_VERSION = "24.0" # Directory of system wheel packages. Some Linux distribution packaging # policies recommend against bundling dependencies. For example, Fedora diff --git a/Lib/ensurepip/_bundled/pip-23.3.2-py3-none-any.whl b/Lib/ensurepip/_bundled/pip-24.0-py3-none-any.whl similarity index 83% rename from Lib/ensurepip/_bundled/pip-23.3.2-py3-none-any.whl rename to Lib/ensurepip/_bundled/pip-24.0-py3-none-any.whl index ae78b8a6ce0737..2e6aa9d2cb9923 100644 Binary files a/Lib/ensurepip/_bundled/pip-23.3.2-py3-none-any.whl and b/Lib/ensurepip/_bundled/pip-24.0-py3-none-any.whl differ diff --git a/Lib/enum.py b/Lib/enum.py index a8a50a58380375..98a8966f5eb159 100644 --- a/Lib/enum.py +++ b/Lib/enum.py @@ -409,10 +409,11 @@ def __setitem__(self, key, value): if isinstance(value, auto): single = True value = (value, ) - if type(value) is tuple and any(isinstance(v, auto) for v in value): + if isinstance(value, tuple) and any(isinstance(v, auto) for v in value): # insist on an actual tuple, no subclasses, in keeping with only supporting # top-level auto() usage (not contained in any other data structure) auto_valued = [] + t = type(value) for v in value: if isinstance(v, auto): non_auto_store = False @@ -427,7 +428,12 @@ def __setitem__(self, key, value): if single: value = auto_valued[0] else: - value = tuple(auto_valued) + try: + # accepts iterable as multiple arguments? + value = t(auto_valued) + except TypeError: + # then pass them in singlely + value = t(*auto_valued) self._member_names[key] = None if non_auto_store: self._last_values.append(value) diff --git a/Lib/idlelib/editor.py b/Lib/idlelib/editor.py index 6ad383f460c7ee..8ee8eba64367a5 100644 --- a/Lib/idlelib/editor.py +++ b/Lib/idlelib/editor.py @@ -166,8 +166,9 @@ def __init__(self, flist=None, filename=None, key=None, root=None): text.bind("<3>",self.right_menu_event) text.bind('', wheel_event) - text.bind('', wheel_event) - text.bind('', wheel_event) + if text._windowingsystem == 'x11': + text.bind('', wheel_event) + text.bind('', wheel_event) text.bind('', self.handle_winconfig) text.bind("<>", self.cut) text.bind("<>", self.copy) diff --git a/Lib/idlelib/idle_test/test_sidebar.py b/Lib/idlelib/idle_test/test_sidebar.py index fb52b3a0179553..605e7a892570d7 100644 --- a/Lib/idlelib/idle_test/test_sidebar.py +++ b/Lib/idlelib/idle_test/test_sidebar.py @@ -690,16 +690,22 @@ def test_mousewheel(self): last_lineno = get_end_linenumber(text) self.assertIsNotNone(text.dlineinfo(text.index(f'{last_lineno}.0'))) - # Scroll up using the event. - # The meaning of delta is platform-dependent. - delta = -1 if sys.platform == 'darwin' else 120 - sidebar.canvas.event_generate('', x=0, y=0, delta=delta) + # Delta for , whose meaning is platform-dependent. + delta = 1 if sidebar.canvas._windowingsystem == 'aqua' else 120 + + # Scroll up. + if sidebar.canvas._windowingsystem == 'x11': + sidebar.canvas.event_generate('', x=0, y=0) + else: + sidebar.canvas.event_generate('', x=0, y=0, delta=delta) yield - if sys.platform != 'darwin': # .update_idletasks() does not work. - self.assertIsNone(text.dlineinfo(text.index(f'{last_lineno}.0'))) + self.assertIsNone(text.dlineinfo(text.index(f'{last_lineno}.0'))) - # Scroll back down using the event. - sidebar.canvas.event_generate('', x=0, y=0) + # Scroll back down. + if sidebar.canvas._windowingsystem == 'x11': + sidebar.canvas.event_generate('', x=0, y=0) + else: + sidebar.canvas.event_generate('', x=0, y=0, delta=-delta) yield self.assertIsNotNone(text.dlineinfo(text.index(f'{last_lineno}.0'))) diff --git a/Lib/idlelib/tree.py b/Lib/idlelib/tree.py index 9c2eb47b24aec9..0726d7e23660f6 100644 --- a/Lib/idlelib/tree.py +++ b/Lib/idlelib/tree.py @@ -285,8 +285,9 @@ def drawtext(self): self.label.bind("<1>", self.select_or_edit) self.label.bind("", self.flip) self.label.bind("", lambda e: wheel_event(e, self.canvas)) - self.label.bind("", lambda e: wheel_event(e, self.canvas)) - self.label.bind("", lambda e: wheel_event(e, self.canvas)) + if self.label._windowingsystem == 'x11': + self.label.bind("", lambda e: wheel_event(e, self.canvas)) + self.label.bind("", lambda e: wheel_event(e, self.canvas)) self.text_id = id def select_or_edit(self, event=None): @@ -460,8 +461,9 @@ def __init__(self, master, **opts): self.canvas.bind("", self.unit_up) self.canvas.bind("", self.unit_down) self.canvas.bind("", wheel_event) - self.canvas.bind("", wheel_event) - self.canvas.bind("", wheel_event) + if self.canvas._windowingsystem == 'x11': + self.canvas.bind("", wheel_event) + self.canvas.bind("", wheel_event) #if isinstance(master, Toplevel) or isinstance(master, Tk): self.canvas.bind("", self.zoom_height) self.canvas.focus_set() diff --git a/Lib/ssl.py b/Lib/ssl.py index 74a9d2d8fd4fb0..03d0121891ff4c 100644 --- a/Lib/ssl.py +++ b/Lib/ssl.py @@ -994,71 +994,67 @@ def _create(cls, sock, server_side=False, do_handshake_on_connect=True, if context.check_hostname and not server_hostname: raise ValueError("check_hostname requires server_hostname") + sock_timeout = sock.gettimeout() kwargs = dict( family=sock.family, type=sock.type, proto=sock.proto, fileno=sock.fileno() ) self = cls.__new__(cls, **kwargs) super(SSLSocket, self).__init__(**kwargs) - sock_timeout = sock.gettimeout() sock.detach() - - self._context = context - self._session = session - self._closed = False - self._sslobj = None - self.server_side = server_side - self.server_hostname = context._encode_hostname(server_hostname) - self.do_handshake_on_connect = do_handshake_on_connect - self.suppress_ragged_eofs = suppress_ragged_eofs - - # See if we are connected + # Now SSLSocket is responsible for closing the file descriptor. try: - self.getpeername() - except OSError as e: - if e.errno != errno.ENOTCONN: - raise - connected = False - blocking = self.getblocking() - self.setblocking(False) + self._context = context + self._session = session + self._closed = False + self._sslobj = None + self.server_side = server_side + self.server_hostname = context._encode_hostname(server_hostname) + self.do_handshake_on_connect = do_handshake_on_connect + self.suppress_ragged_eofs = suppress_ragged_eofs + + # See if we are connected try: - # We are not connected so this is not supposed to block, but - # testing revealed otherwise on macOS and Windows so we do - # the non-blocking dance regardless. Our raise when any data - # is found means consuming the data is harmless. - notconn_pre_handshake_data = self.recv(1) + self.getpeername() except OSError as e: - # EINVAL occurs for recv(1) on non-connected on unix sockets. - if e.errno not in (errno.ENOTCONN, errno.EINVAL): + if e.errno != errno.ENOTCONN: raise - notconn_pre_handshake_data = b'' - self.setblocking(blocking) - if notconn_pre_handshake_data: - # This prevents pending data sent to the socket before it was - # closed from escaping to the caller who could otherwise - # presume it came through a successful TLS connection. - reason = "Closed before TLS handshake with data in recv buffer." - notconn_pre_handshake_data_error = SSLError(e.errno, reason) - # Add the SSLError attributes that _ssl.c always adds. - notconn_pre_handshake_data_error.reason = reason - notconn_pre_handshake_data_error.library = None - try: - self.close() - except OSError: - pass + connected = False + blocking = self.getblocking() + self.setblocking(False) try: - raise notconn_pre_handshake_data_error - finally: - # Explicitly break the reference cycle. - notconn_pre_handshake_data_error = None - else: - connected = True + # We are not connected so this is not supposed to block, but + # testing revealed otherwise on macOS and Windows so we do + # the non-blocking dance regardless. Our raise when any data + # is found means consuming the data is harmless. + notconn_pre_handshake_data = self.recv(1) + except OSError as e: + # EINVAL occurs for recv(1) on non-connected on unix sockets. + if e.errno not in (errno.ENOTCONN, errno.EINVAL): + raise + notconn_pre_handshake_data = b'' + self.setblocking(blocking) + if notconn_pre_handshake_data: + # This prevents pending data sent to the socket before it was + # closed from escaping to the caller who could otherwise + # presume it came through a successful TLS connection. + reason = "Closed before TLS handshake with data in recv buffer." + notconn_pre_handshake_data_error = SSLError(e.errno, reason) + # Add the SSLError attributes that _ssl.c always adds. + notconn_pre_handshake_data_error.reason = reason + notconn_pre_handshake_data_error.library = None + try: + raise notconn_pre_handshake_data_error + finally: + # Explicitly break the reference cycle. + notconn_pre_handshake_data_error = None + else: + connected = True - self.settimeout(sock_timeout) # Must come after setblocking() calls. - self._connected = connected - if connected: - # create the SSL object - try: + self.settimeout(sock_timeout) # Must come after setblocking() calls. + self._connected = connected + if connected: + # create the SSL object self._sslobj = self._context._wrap_socket( self, server_side, self.server_hostname, owner=self, session=self._session, @@ -1069,9 +1065,12 @@ def _create(cls, sock, server_side=False, do_handshake_on_connect=True, # non-blocking raise ValueError("do_handshake_on_connect should not be specified for non-blocking sockets") self.do_handshake() - except (OSError, ValueError): + except: + try: self.close() - raise + except OSError: + pass + raise return self @property diff --git a/Lib/tarfile.py b/Lib/tarfile.py index 20e0394507f5db..9775040cbe372c 100755 --- a/Lib/tarfile.py +++ b/Lib/tarfile.py @@ -2456,7 +2456,8 @@ def makedir(self, tarinfo, targetpath): # later in _extract_member(). os.mkdir(targetpath, 0o700) except FileExistsError: - pass + if not os.path.isdir(targetpath): + raise def makefile(self, tarinfo, targetpath): """Make a file called targetpath. diff --git a/Lib/test/.ruff.toml b/Lib/test/.ruff.toml index d6c1d8745036ec..1c9bac507209b1 100644 --- a/Lib/test/.ruff.toml +++ b/Lib/test/.ruff.toml @@ -1,7 +1,4 @@ fix = true -select = [ - "F811", # Redefinition of unused variable (useful for finding test methods with the same name) -] extend-exclude = [ # Excluded (run with the other AC files in its own separate ruff job in pre-commit) "test_clinic.py", @@ -21,3 +18,8 @@ extend-exclude = [ "test_pkg.py", "test_yield_from.py", ] + +[lint] +select = [ + "F811", # Redefinition of unused variable (useful for finding test methods with the same name) +] diff --git a/Lib/test/archiver_tests.py b/Lib/test/archiver_tests.py new file mode 100644 index 00000000000000..1a4bbb9e5706c5 --- /dev/null +++ b/Lib/test/archiver_tests.py @@ -0,0 +1,155 @@ +"""Tests common to tarfile and zipfile.""" + +import os +import sys + +from test.support import os_helper + +class OverwriteTests: + + def setUp(self): + os.makedirs(self.testdir) + self.addCleanup(os_helper.rmtree, self.testdir) + + def create_file(self, path, content=b''): + with open(path, 'wb') as f: + f.write(content) + + def open(self, path): + raise NotImplementedError + + def extractall(self, ar): + raise NotImplementedError + + + def test_overwrite_file_as_file(self): + target = os.path.join(self.testdir, 'test') + self.create_file(target, b'content') + with self.open(self.ar_with_file) as ar: + self.extractall(ar) + self.assertTrue(os.path.isfile(target)) + with open(target, 'rb') as f: + self.assertEqual(f.read(), b'newcontent') + + def test_overwrite_dir_as_dir(self): + target = os.path.join(self.testdir, 'test') + os.mkdir(target) + with self.open(self.ar_with_dir) as ar: + self.extractall(ar) + self.assertTrue(os.path.isdir(target)) + + def test_overwrite_dir_as_implicit_dir(self): + target = os.path.join(self.testdir, 'test') + os.mkdir(target) + with self.open(self.ar_with_implicit_dir) as ar: + self.extractall(ar) + self.assertTrue(os.path.isdir(target)) + self.assertTrue(os.path.isfile(os.path.join(target, 'file'))) + with open(os.path.join(target, 'file'), 'rb') as f: + self.assertEqual(f.read(), b'newcontent') + + def test_overwrite_dir_as_file(self): + target = os.path.join(self.testdir, 'test') + os.mkdir(target) + with self.open(self.ar_with_file) as ar: + with self.assertRaises(PermissionError if sys.platform == 'win32' + else IsADirectoryError): + self.extractall(ar) + self.assertTrue(os.path.isdir(target)) + + def test_overwrite_file_as_dir(self): + target = os.path.join(self.testdir, 'test') + self.create_file(target, b'content') + with self.open(self.ar_with_dir) as ar: + with self.assertRaises(FileExistsError): + self.extractall(ar) + self.assertTrue(os.path.isfile(target)) + with open(target, 'rb') as f: + self.assertEqual(f.read(), b'content') + + def test_overwrite_file_as_implicit_dir(self): + target = os.path.join(self.testdir, 'test') + self.create_file(target, b'content') + with self.open(self.ar_with_implicit_dir) as ar: + with self.assertRaises(FileNotFoundError if sys.platform == 'win32' + else NotADirectoryError): + self.extractall(ar) + self.assertTrue(os.path.isfile(target)) + with open(target, 'rb') as f: + self.assertEqual(f.read(), b'content') + + @os_helper.skip_unless_symlink + def test_overwrite_file_symlink_as_file(self): + # XXX: It is potential security vulnerability. + target = os.path.join(self.testdir, 'test') + target2 = os.path.join(self.testdir, 'test2') + self.create_file(target2, b'content') + os.symlink('test2', target) + with self.open(self.ar_with_file) as ar: + self.extractall(ar) + self.assertTrue(os.path.islink(target)) + self.assertTrue(os.path.isfile(target2)) + with open(target2, 'rb') as f: + self.assertEqual(f.read(), b'newcontent') + + @os_helper.skip_unless_symlink + def test_overwrite_broken_file_symlink_as_file(self): + # XXX: It is potential security vulnerability. + target = os.path.join(self.testdir, 'test') + target2 = os.path.join(self.testdir, 'test2') + os.symlink('test2', target) + with self.open(self.ar_with_file) as ar: + self.extractall(ar) + self.assertTrue(os.path.islink(target)) + self.assertTrue(os.path.isfile(target2)) + with open(target2, 'rb') as f: + self.assertEqual(f.read(), b'newcontent') + + @os_helper.skip_unless_symlink + def test_overwrite_dir_symlink_as_dir(self): + # XXX: It is potential security vulnerability. + target = os.path.join(self.testdir, 'test') + target2 = os.path.join(self.testdir, 'test2') + os.mkdir(target2) + os.symlink('test2', target, target_is_directory=True) + with self.open(self.ar_with_dir) as ar: + self.extractall(ar) + self.assertTrue(os.path.islink(target)) + self.assertTrue(os.path.isdir(target2)) + + @os_helper.skip_unless_symlink + def test_overwrite_dir_symlink_as_implicit_dir(self): + # XXX: It is potential security vulnerability. + target = os.path.join(self.testdir, 'test') + target2 = os.path.join(self.testdir, 'test2') + os.mkdir(target2) + os.symlink('test2', target, target_is_directory=True) + with self.open(self.ar_with_implicit_dir) as ar: + self.extractall(ar) + self.assertTrue(os.path.islink(target)) + self.assertTrue(os.path.isdir(target2)) + self.assertTrue(os.path.isfile(os.path.join(target2, 'file'))) + with open(os.path.join(target2, 'file'), 'rb') as f: + self.assertEqual(f.read(), b'newcontent') + + @os_helper.skip_unless_symlink + def test_overwrite_broken_dir_symlink_as_dir(self): + target = os.path.join(self.testdir, 'test') + target2 = os.path.join(self.testdir, 'test2') + os.symlink('test2', target, target_is_directory=True) + with self.open(self.ar_with_dir) as ar: + with self.assertRaises(FileExistsError): + self.extractall(ar) + self.assertTrue(os.path.islink(target)) + self.assertFalse(os.path.exists(target2)) + + @os_helper.skip_unless_symlink + def test_overwrite_broken_dir_symlink_as_implicit_dir(self): + target = os.path.join(self.testdir, 'test') + target2 = os.path.join(self.testdir, 'test2') + os.symlink('test2', target, target_is_directory=True) + with self.open(self.ar_with_implicit_dir) as ar: + with self.assertRaises(FileExistsError): + self.extractall(ar) + self.assertTrue(os.path.islink(target)) + self.assertFalse(os.path.exists(target2)) diff --git a/Lib/test/test_argparse.py b/Lib/test/test_argparse.py index 3221d564d198ed..b13f543aeb4e6f 100644 --- a/Lib/test/test_argparse.py +++ b/Lib/test/test_argparse.py @@ -3950,7 +3950,7 @@ class TestHelpUsageWithParentheses(HelpTestCase): options: -h, --help show this help message and exit - -p {1 (option A), 2 (option B)}, --optional {1 (option A), 2 (option B)} + -p, --optional {1 (option A), 2 (option B)} ''' version = '' @@ -4433,8 +4433,8 @@ class TestHelpAlternatePrefixChars(HelpTestCase): help = usage + '''\ options: - ^^foo foo help - ;b BAR, ;;bar BAR bar help + ^^foo foo help + ;b, ;;bar BAR bar help ''' version = '' diff --git a/Lib/test/test_asyncio/test_base_events.py b/Lib/test/test_asyncio/test_base_events.py index c2080977e9d587..82071edb252570 100644 --- a/Lib/test/test_asyncio/test_base_events.py +++ b/Lib/test/test_asyncio/test_base_events.py @@ -1232,7 +1232,7 @@ def test_create_datagram_endpoint_wrong_sock(self): with sock: coro = self.loop.create_datagram_endpoint(MyProto, sock=sock) with self.assertRaisesRegex(ValueError, - 'A UDP Socket was expected'): + 'A datagram socket was expected'): self.loop.run_until_complete(coro) def test_create_connection_no_host_port_sock(self): diff --git a/Lib/test/test_asyncio/test_locks.py b/Lib/test/test_asyncio/test_locks.py index 9029efd2355b46..a0884bffe6b0de 100644 --- a/Lib/test/test_asyncio/test_locks.py +++ b/Lib/test/test_asyncio/test_locks.py @@ -816,6 +816,98 @@ async def func(): # originally raised. self.assertIs(err.exception, raised) + async def test_cancelled_wakeup(self): + # Test that a task cancelled at the "same" time as it is woken + # up as part of a Condition.notify() does not result in a lost wakeup. + # This test simulates a cancel while the target task is awaiting initial + # wakeup on the wakeup queue. + condition = asyncio.Condition() + state = 0 + async def consumer(): + nonlocal state + async with condition: + while True: + await condition.wait_for(lambda: state != 0) + if state < 0: + return + state -= 1 + + # create two consumers + c = [asyncio.create_task(consumer()) for _ in range(2)] + # wait for them to settle + await asyncio.sleep(0) + async with condition: + # produce one item and wake up one + state += 1 + condition.notify(1) + + # Cancel it while it is awaiting to be run. + # This cancellation could come from the outside + c[0].cancel() + + # now wait for the item to be consumed + # if it doesn't means that our "notify" didn"t take hold. + # because it raced with a cancel() + try: + async with asyncio.timeout(0.01): + await condition.wait_for(lambda: state == 0) + except TimeoutError: + pass + self.assertEqual(state, 0) + + # clean up + state = -1 + condition.notify_all() + await c[1] + + async def test_cancelled_wakeup_relock(self): + # Test that a task cancelled at the "same" time as it is woken + # up as part of a Condition.notify() does not result in a lost wakeup. + # This test simulates a cancel while the target task is acquiring the lock + # again. + condition = asyncio.Condition() + state = 0 + async def consumer(): + nonlocal state + async with condition: + while True: + await condition.wait_for(lambda: state != 0) + if state < 0: + return + state -= 1 + + # create two consumers + c = [asyncio.create_task(consumer()) for _ in range(2)] + # wait for them to settle + await asyncio.sleep(0) + async with condition: + # produce one item and wake up one + state += 1 + condition.notify(1) + + # now we sleep for a bit. This allows the target task to wake up and + # settle on re-aquiring the lock + await asyncio.sleep(0) + + # Cancel it while awaiting the lock + # This cancel could come the outside. + c[0].cancel() + + # now wait for the item to be consumed + # if it doesn't means that our "notify" didn"t take hold. + # because it raced with a cancel() + try: + async with asyncio.timeout(0.01): + await condition.wait_for(lambda: state == 0) + except TimeoutError: + pass + self.assertEqual(state, 0) + + # clean up + state = -1 + condition.notify_all() + await c[1] + class SemaphoreTests(unittest.IsolatedAsyncioTestCase): def test_initial_value_zero(self): diff --git a/Lib/test/test_capi/test_list.py b/Lib/test/test_capi/test_list.py index eb03d51d3def37..dceb4fce3c077b 100644 --- a/Lib/test/test_capi/test_list.py +++ b/Lib/test/test_capi/test_list.py @@ -82,10 +82,8 @@ def test_list_get_size(self): # CRASHES size(UserList()) # CRASHES size(NULL) - - def test_list_getitem(self): - # Test PyList_GetItem() - getitem = _testcapi.list_getitem + def check_list_get_item(self, getitem, exctype): + # Common test cases for PyList_GetItem() and PyList_GetItemRef() lst = [1, 2, 3] self.assertEqual(getitem(lst, 0), 1) self.assertEqual(getitem(lst, 2), 3) @@ -93,12 +91,19 @@ def test_list_getitem(self): self.assertRaises(IndexError, getitem, lst, -1) self.assertRaises(IndexError, getitem, lst, PY_SSIZE_T_MIN) self.assertRaises(IndexError, getitem, lst, PY_SSIZE_T_MAX) - self.assertRaises(SystemError, getitem, 42, 1) - self.assertRaises(SystemError, getitem, (1, 2, 3), 1) - self.assertRaises(SystemError, getitem, {1: 2}, 1) - + self.assertRaises(exctype, getitem, 42, 1) + self.assertRaises(exctype, getitem, (1, 2, 3), 1) + self.assertRaises(exctype, getitem, {1: 2}, 1) # CRASHES getitem(NULL, 1) + def test_list_getitem(self): + # Test PyList_GetItem() + self.check_list_get_item(_testcapi.list_getitem, SystemError) + + def test_list_get_item_ref(self): + # Test PyList_GetItemRef() + self.check_list_get_item(_testcapi.list_get_item_ref, TypeError) + def test_list_get_item(self): # Test PyList_GET_ITEM() get_item = _testcapi.list_get_item @@ -112,7 +117,6 @@ def test_list_get_item(self): # CRASHES get_item(21, 2) # CRASHES get_item(NULL, 1) - def test_list_setitem(self): # Test PyList_SetItem() setitem = _testcapi.list_setitem diff --git a/Lib/test/test_capi/test_structmembers.py b/Lib/test/test_capi/test_structmembers.py index 2cf46b203478dc..415b8033bd16b3 100644 --- a/Lib/test/test_capi/test_structmembers.py +++ b/Lib/test/test_capi/test_structmembers.py @@ -14,6 +14,13 @@ PY_SSIZE_T_MAX, PY_SSIZE_T_MIN, ) + +class Index: + def __init__(self, value): + self.value = value + def __index__(self): + return self.value + # There are two classes: one using and another using # `Py_`-prefixed API. They should behave the same in Python @@ -72,6 +79,10 @@ def test_int(self): self.assertEqual(ts.T_INT, INT_MIN) ts.T_UINT = UINT_MAX self.assertEqual(ts.T_UINT, UINT_MAX) + ts.T_UINT = Index(0) + self.assertEqual(ts.T_UINT, 0) + ts.T_UINT = Index(INT_MAX) + self.assertEqual(ts.T_UINT, INT_MAX) def test_long(self): ts = self.ts @@ -81,6 +92,10 @@ def test_long(self): self.assertEqual(ts.T_LONG, LONG_MIN) ts.T_ULONG = ULONG_MAX self.assertEqual(ts.T_ULONG, ULONG_MAX) + ts.T_ULONG = Index(0) + self.assertEqual(ts.T_ULONG, 0) + ts.T_ULONG = Index(LONG_MAX) + self.assertEqual(ts.T_ULONG, LONG_MAX) def test_py_ssize_t(self): ts = self.ts @@ -173,6 +188,28 @@ def test_ushort_max(self): with warnings_helper.check_warnings(('', RuntimeWarning)): ts.T_USHORT = USHRT_MAX+1 + def test_int(self): + ts = self.ts + if LONG_MIN < INT_MIN: + with self.assertWarns(RuntimeWarning): + ts.T_INT = INT_MIN-1 + if LONG_MAX > INT_MAX: + with self.assertWarns(RuntimeWarning): + ts.T_INT = INT_MAX+1 + + def test_uint(self): + ts = self.ts + with self.assertWarns(RuntimeWarning): + ts.T_UINT = -1 + if ULONG_MAX > UINT_MAX: + with self.assertWarns(RuntimeWarning): + ts.T_UINT = UINT_MAX+1 + + def test_ulong(self): + ts = self.ts + with self.assertWarns(RuntimeWarning): + ts.T_ULONG = -1 + class TestWarnings_OldAPI(TestWarnings, unittest.TestCase): cls = _test_structmembersType_OldAPI diff --git a/Lib/test/test_capi/test_watchers.py b/Lib/test/test_capi/test_watchers.py index 5981712c80c3a9..ae062b1bda26b7 100644 --- a/Lib/test/test_capi/test_watchers.py +++ b/Lib/test/test_capi/test_watchers.py @@ -151,8 +151,8 @@ def test_watch_out_of_range_watcher_id(self): def test_watch_unassigned_watcher_id(self): d = {} - with self.assertRaisesRegex(ValueError, r"No dict watcher set for ID 1"): - self.watch(1, d) + with self.assertRaisesRegex(ValueError, r"No dict watcher set for ID 3"): + self.watch(3, d) def test_unwatch_non_dict(self): with self.watcher() as wid: @@ -168,8 +168,8 @@ def test_unwatch_out_of_range_watcher_id(self): def test_unwatch_unassigned_watcher_id(self): d = {} - with self.assertRaisesRegex(ValueError, r"No dict watcher set for ID 1"): - self.unwatch(1, d) + with self.assertRaisesRegex(ValueError, r"No dict watcher set for ID 3"): + self.unwatch(3, d) def test_clear_out_of_range_watcher_id(self): with self.assertRaisesRegex(ValueError, r"Invalid dict watcher ID -1"): @@ -178,8 +178,8 @@ def test_clear_out_of_range_watcher_id(self): self.clear_watcher(8) # DICT_MAX_WATCHERS = 8 def test_clear_unassigned_watcher_id(self): - with self.assertRaisesRegex(ValueError, r"No dict watcher set for ID 1"): - self.clear_watcher(1) + with self.assertRaisesRegex(ValueError, r"No dict watcher set for ID 3"): + self.clear_watcher(3) class TestTypeWatchers(unittest.TestCase): diff --git a/Lib/test/test_class.py b/Lib/test/test_class.py index 1531aad4f1f779..d59271435e9eb0 100644 --- a/Lib/test/test_class.py +++ b/Lib/test/test_class.py @@ -771,6 +771,22 @@ def add_one_level(): with self.assertRaises(RecursionError): add_one_level() + def testMetaclassCallOptimization(self): + calls = 0 + + class TypeMetaclass(type): + def __call__(cls, *args, **kwargs): + nonlocal calls + calls += 1 + return type.__call__(cls, *args, **kwargs) + + class Type(metaclass=TypeMetaclass): + def __init__(self, obj): + self._obj = obj + + for i in range(100): + Type(i) + self.assertEqual(calls, 100) if __name__ == '__main__': unittest.main() diff --git a/Lib/test/test_compile.py b/Lib/test/test_compile.py index 3b1ceceaa6305f..ebb479f2de7c63 100644 --- a/Lib/test/test_compile.py +++ b/Lib/test/test_compile.py @@ -1104,6 +1104,17 @@ async def test(aseq): code_lines = self.get_code_lines(test.__code__) self.assertEqual(expected_lines, code_lines) + def check_line_numbers(self, code, opnames=None): + # Check that all instructions whose op matches opnames + # have a line number. opnames can be a single name, or + # a sequence of names. If it is None, match all ops. + + if isinstance(opnames, str): + opnames = (opnames, ) + for inst in dis.Bytecode(code): + if opnames and inst.opname in opnames: + self.assertIsNotNone(inst.positions.lineno) + def test_line_number_synthetic_jump_multiple_predecessors(self): def f(): for x in it: @@ -1113,25 +1124,52 @@ def f(): except OSError: pass - # Ensure that all JUMP_BACKWARDs have line number - code = f.__code__ - for inst in dis.Bytecode(code): - if inst.opname == 'JUMP_BACKWARD': - self.assertIsNotNone(inst.positions.lineno) + self.check_line_numbers(f.__code__, 'JUMP_BACKWARD') - def test_lineno_of_backward_jump(self): + def test_line_number_synthetic_jump_multiple_predecessors_nested(self): + def f(): + for x in it: + try: + X = 3 + except OSError: + try: + if C3: + X = 4 + except OSError: + pass + return 42 + + self.check_line_numbers(f.__code__, 'JUMP_BACKWARD') + + def test_line_number_synthetic_jump_multiple_predecessors_more_nested(self): + def f(): + for x in it: + try: + X = 3 + except OSError: + try: + if C3: + if C4: + X = 4 + except OSError: + try: + if C3: + if C4: + X = 5 + except OSError: + pass + return 42 + + self.check_line_numbers(f.__code__, 'JUMP_BACKWARD') + + def test_lineno_of_backward_jump_conditional_in_loop(self): # Issue gh-107901 def f(): for i in x: if y: pass - linenos = list(inst.positions.lineno - for inst in dis.get_instructions(f.__code__) - if inst.opname == 'JUMP_BACKWARD') - - self.assertTrue(len(linenos) > 0) - self.assertTrue(all(l is not None for l in linenos)) + self.check_line_numbers(f.__code__, 'JUMP_BACKWARD') def test_big_dict_literal(self): # The compiler has a flushing point in "compiler_dict" that calls compiles diff --git a/Lib/test/test_ctypes/test_as_parameter.py b/Lib/test/test_ctypes/test_as_parameter.py index a1a8745e737fa2..ca75e748256083 100644 --- a/Lib/test/test_ctypes/test_as_parameter.py +++ b/Lib/test/test_ctypes/test_as_parameter.py @@ -221,5 +221,16 @@ class AsParamPropertyWrapperTestCase(BasicWrapTestCase): wrap = AsParamPropertyWrapper +class AsParamNestedWrapperTestCase(BasicWrapTestCase): + """Test that _as_parameter_ is evaluated recursively. + + The _as_parameter_ attribute can be another object which + defines its own _as_parameter_ attribute. + """ + + def wrap(self, param): + return AsParamWrapper(AsParamWrapper(AsParamWrapper(param))) + + if __name__ == '__main__': unittest.main() diff --git a/Lib/test/test_dbm_dumb.py b/Lib/test/test_dbm_dumb.py index a481175b3bfdbd..672f9092207cf6 100644 --- a/Lib/test/test_dbm_dumb.py +++ b/Lib/test/test_dbm_dumb.py @@ -246,9 +246,27 @@ def test_missing_data(self): _delete_files() with self.assertRaises(FileNotFoundError): dumbdbm.open(_fname, value) + self.assertFalse(os.path.exists(_fname + '.dat')) self.assertFalse(os.path.exists(_fname + '.dir')) self.assertFalse(os.path.exists(_fname + '.bak')) + for value in ('c', 'n'): + _delete_files() + with dumbdbm.open(_fname, value) as f: + self.assertTrue(os.path.exists(_fname + '.dat')) + self.assertTrue(os.path.exists(_fname + '.dir')) + self.assertFalse(os.path.exists(_fname + '.bak')) + self.assertFalse(os.path.exists(_fname + '.bak')) + + for value in ('c', 'n'): + _delete_files() + with dumbdbm.open(_fname, value) as f: + f['key'] = 'value' + self.assertTrue(os.path.exists(_fname + '.dat')) + self.assertTrue(os.path.exists(_fname + '.dir')) + self.assertFalse(os.path.exists(_fname + '.bak')) + self.assertTrue(os.path.exists(_fname + '.bak')) + def test_missing_index(self): with dumbdbm.open(_fname, 'n') as f: pass @@ -259,6 +277,60 @@ def test_missing_index(self): self.assertFalse(os.path.exists(_fname + '.dir')) self.assertFalse(os.path.exists(_fname + '.bak')) + for value in ('c', 'n'): + with dumbdbm.open(_fname, value) as f: + self.assertTrue(os.path.exists(_fname + '.dir')) + self.assertFalse(os.path.exists(_fname + '.bak')) + self.assertFalse(os.path.exists(_fname + '.bak')) + os.unlink(_fname + '.dir') + + for value in ('c', 'n'): + with dumbdbm.open(_fname, value) as f: + f['key'] = 'value' + self.assertTrue(os.path.exists(_fname + '.dir')) + self.assertFalse(os.path.exists(_fname + '.bak')) + self.assertTrue(os.path.exists(_fname + '.bak')) + os.unlink(_fname + '.dir') + os.unlink(_fname + '.bak') + + def test_sync_empty_unmodified(self): + with dumbdbm.open(_fname, 'n') as f: + pass + os.unlink(_fname + '.dir') + for value in ('c', 'n'): + with dumbdbm.open(_fname, value) as f: + self.assertTrue(os.path.exists(_fname + '.dir')) + self.assertFalse(os.path.exists(_fname + '.bak')) + f.sync() + self.assertTrue(os.path.exists(_fname + '.dir')) + self.assertFalse(os.path.exists(_fname + '.bak')) + os.unlink(_fname + '.dir') + f.sync() + self.assertFalse(os.path.exists(_fname + '.dir')) + self.assertFalse(os.path.exists(_fname + '.bak')) + self.assertFalse(os.path.exists(_fname + '.dir')) + self.assertFalse(os.path.exists(_fname + '.bak')) + + def test_sync_nonempty_unmodified(self): + with dumbdbm.open(_fname, 'n') as f: + pass + os.unlink(_fname + '.dir') + for value in ('c', 'n'): + with dumbdbm.open(_fname, value) as f: + f['key'] = 'value' + self.assertTrue(os.path.exists(_fname + '.dir')) + self.assertFalse(os.path.exists(_fname + '.bak')) + f.sync() + self.assertTrue(os.path.exists(_fname + '.dir')) + self.assertTrue(os.path.exists(_fname + '.bak')) + os.unlink(_fname + '.dir') + os.unlink(_fname + '.bak') + f.sync() + self.assertFalse(os.path.exists(_fname + '.dir')) + self.assertFalse(os.path.exists(_fname + '.bak')) + self.assertFalse(os.path.exists(_fname + '.dir')) + self.assertFalse(os.path.exists(_fname + '.bak')) + def test_invalid_flag(self): for flag in ('x', 'rf', None): with self.assertRaisesRegex(ValueError, diff --git a/Lib/test/test_enum.py b/Lib/test/test_enum.py index d045739efa46b8..39c1ae0ad5a078 100644 --- a/Lib/test/test_enum.py +++ b/Lib/test/test_enum.py @@ -2344,6 +2344,40 @@ class SomeTuple(tuple, Enum): globals()['SomeTuple'] = SomeTuple test_pickle_dump_load(self.assertIs, SomeTuple.first) + def test_tuple_subclass_with_auto_1(self): + from collections import namedtuple + T = namedtuple('T', 'index desc') + class SomeEnum(T, Enum): + __qualname__ = 'SomeEnum' # needed for pickle protocol 4 + first = auto(), 'for the money' + second = auto(), 'for the show' + third = auto(), 'for the music' + self.assertIs(type(SomeEnum.first), SomeEnum) + self.assertEqual(SomeEnum.third.value, (3, 'for the music')) + self.assertIsInstance(SomeEnum.third.value, T) + self.assertEqual(SomeEnum.first.index, 1) + self.assertEqual(SomeEnum.second.desc, 'for the show') + globals()['SomeEnum'] = SomeEnum + globals()['T'] = T + test_pickle_dump_load(self.assertIs, SomeEnum.first) + + def test_tuple_subclass_with_auto_2(self): + from collections import namedtuple + T = namedtuple('T', 'index desc') + class SomeEnum(Enum): + __qualname__ = 'SomeEnum' # needed for pickle protocol 4 + first = T(auto(), 'for the money') + second = T(auto(), 'for the show') + third = T(auto(), 'for the music') + self.assertIs(type(SomeEnum.first), SomeEnum) + self.assertEqual(SomeEnum.third.value, (3, 'for the music')) + self.assertIsInstance(SomeEnum.third.value, T) + self.assertEqual(SomeEnum.first.value.index, 1) + self.assertEqual(SomeEnum.second.value.desc, 'for the show') + globals()['SomeEnum'] = SomeEnum + globals()['T'] = T + test_pickle_dump_load(self.assertIs, SomeEnum.first) + def test_duplicate_values_give_unique_enum_items(self): class AutoNumber(Enum): first = () diff --git a/Lib/test/test_frame.py b/Lib/test/test_frame.py index 7f17666a8d9697..244ce8af7cdf08 100644 --- a/Lib/test/test_frame.py +++ b/Lib/test/test_frame.py @@ -55,6 +55,28 @@ class C: # The reference was released by .clear() self.assertIs(None, wr()) + def test_clear_locals_after_f_locals_access(self): + # see gh-113939 + class C: + pass + + wr = None + def inner(): + nonlocal wr + c = C() + wr = weakref.ref(c) + 1/0 + + try: + inner() + except ZeroDivisionError as exc: + support.gc_collect() + self.assertIsNotNone(wr()) + print(exc.__traceback__.tb_next.tb_frame.f_locals) + exc.__traceback__.tb_next.tb_frame.clear() + support.gc_collect() + self.assertIsNone(wr()) + def test_clear_does_not_clear_specials(self): class C: pass diff --git a/Lib/test/test_monitoring.py b/Lib/test/test_monitoring.py index a64d1ed79decd8..60b6326bfbad5e 100644 --- a/Lib/test/test_monitoring.py +++ b/Lib/test/test_monitoring.py @@ -1466,9 +1466,8 @@ def func(): ('branch', 'func', 4, 4), ('line', 'func', 5), ('line', 'meth', 1), - ('jump', 'func', 5, 5), - ('jump', 'func', 5, '[offset=114]'), - ('branch', 'func', '[offset=120]', '[offset=124]'), + ('jump', 'func', 5, '[offset=118]'), + ('branch', 'func', '[offset=122]', '[offset=126]'), ('line', 'get_events', 11)]) self.check_events(func, recorders = FLOW_AND_LINE_RECORDERS, expected = [ @@ -1482,9 +1481,8 @@ def func(): ('line', 'func', 5), ('line', 'meth', 1), ('return', 'meth', None), - ('jump', 'func', 5, 5), - ('jump', 'func', 5, '[offset=114]'), - ('branch', 'func', '[offset=120]', '[offset=124]'), + ('jump', 'func', 5, '[offset=118]'), + ('branch', 'func', '[offset=122]', '[offset=126]'), ('return', 'func', None), ('line', 'get_events', 11)]) diff --git a/Lib/test/test_ssl.py b/Lib/test/test_ssl.py index 3fdfa2960503b8..1b18230d83577d 100644 --- a/Lib/test/test_ssl.py +++ b/Lib/test/test_ssl.py @@ -2206,14 +2206,15 @@ def _test_get_server_certificate(test, host, port, cert=None): sys.stdout.write("\nVerified certificate for %s:%s is\n%s\n" % (host, port ,pem)) def _test_get_server_certificate_fail(test, host, port): - try: - pem = ssl.get_server_certificate((host, port), ca_certs=CERTFILE) - except ssl.SSLError as x: - #should fail - if support.verbose: - sys.stdout.write("%s\n" % x) - else: - test.fail("Got server certificate %s for %s:%s!" % (pem, host, port)) + with warnings_helper.check_no_resource_warning(test): + try: + pem = ssl.get_server_certificate((host, port), ca_certs=CERTFILE) + except ssl.SSLError as x: + #should fail + if support.verbose: + sys.stdout.write("%s\n" % x) + else: + test.fail("Got server certificate %s for %s:%s!" % (pem, host, port)) from test.ssl_servers import make_https_server @@ -3026,6 +3027,16 @@ def test_check_hostname_idn(self): server_hostname="python.example.org") as s: with self.assertRaises(ssl.CertificateError): s.connect((HOST, server.port)) + with ThreadedEchoServer(context=server_context, chatty=True) as server: + with warnings_helper.check_no_resource_warning(self): + with self.assertRaises(UnicodeError): + context.wrap_socket(socket.socket(), + server_hostname='.pythontest.net') + with ThreadedEchoServer(context=server_context, chatty=True) as server: + with warnings_helper.check_no_resource_warning(self): + with self.assertRaises(UnicodeDecodeError): + context.wrap_socket(socket.socket(), + server_hostname=b'k\xf6nig.idn.pythontest.net') def test_wrong_cert_tls12(self): """Connecting when the server rejects the client's certificate @@ -4983,7 +4994,8 @@ def call_after_accept(conn_to_client): self.assertIsNone(wrap_error.library, msg="attr must exist") finally: # gh-108342: Explicitly break the reference cycle - wrap_error = None + with warnings_helper.check_no_resource_warning(self): + wrap_error = None server = None def test_https_client_non_tls_response_ignored(self): @@ -5032,7 +5044,8 @@ def call_after_accept(conn_to_client): # socket; that fails if the connection is broken. It may seem pointless # to test this. It serves as an illustration of something that we never # want to happen... properly not happening. - with self.assertRaises(OSError): + with warnings_helper.check_no_resource_warning(self), \ + self.assertRaises(OSError): connection.request("HEAD", "/test", headers={"Host": "localhost"}) response = connection.getresponse() diff --git a/Lib/test/test_stable_abi_ctypes.py b/Lib/test/test_stable_abi_ctypes.py index 054e7f0feb1a19..8bd373976426ef 100644 --- a/Lib/test/test_stable_abi_ctypes.py +++ b/Lib/test/test_stable_abi_ctypes.py @@ -372,6 +372,7 @@ def test_windows_feature_macros(self): "PyList_Append", "PyList_AsTuple", "PyList_GetItem", + "PyList_GetItemRef", "PyList_GetSlice", "PyList_Insert", "PyList_New", diff --git a/Lib/test/test_sys.py b/Lib/test/test_sys.py index 6c87dfabad9f0f..71671a5a984256 100644 --- a/Lib/test/test_sys.py +++ b/Lib/test/test_sys.py @@ -1392,6 +1392,7 @@ def setUp(self): self.longdigit = sys.int_info.sizeof_digit import _testinternalcapi self.gc_headsize = _testinternalcapi.SIZEOF_PYGC_HEAD + self.managed_pre_header_size = _testinternalcapi.SIZEOF_MANAGED_PRE_HEADER check_sizeof = test.support.check_sizeof @@ -1427,7 +1428,7 @@ class OverflowSizeof(int): def __sizeof__(self): return int(self) self.assertEqual(sys.getsizeof(OverflowSizeof(sys.maxsize)), - sys.maxsize + self.gc_headsize*2) + sys.maxsize + self.gc_headsize + self.managed_pre_header_size) with self.assertRaises(OverflowError): sys.getsizeof(OverflowSizeof(sys.maxsize + 1)) with self.assertRaises(ValueError): @@ -1650,7 +1651,7 @@ def delx(self): del self.__x # type # static type: PyTypeObject fmt = 'P2nPI13Pl4Pn9Pn12PIPc' - s = vsize('2P' + fmt) + s = vsize(fmt) check(int, s) # class s = vsize(fmt + # PyTypeObject diff --git a/Lib/test/test_tarfile.py b/Lib/test/test_tarfile.py index da5009126b3815..51f070e96047a6 100644 --- a/Lib/test/test_tarfile.py +++ b/Lib/test/test_tarfile.py @@ -15,6 +15,7 @@ import unittest.mock import tarfile +from test import archiver_tests from test import support from test.support import os_helper from test.support import script_helper @@ -4135,6 +4136,38 @@ def valueerror_filter(tarinfo, path): self.expect_exception(TypeError) # errorlevel is not int +class OverwriteTests(archiver_tests.OverwriteTests, unittest.TestCase): + testdir = os.path.join(TEMPDIR, "testoverwrite") + + @classmethod + def setUpClass(cls): + p = cls.ar_with_file = os.path.join(TEMPDIR, 'tar-with-file.tar') + cls.addClassCleanup(os_helper.unlink, p) + with tarfile.open(p, 'w') as tar: + t = tarfile.TarInfo('test') + t.size = 10 + tar.addfile(t, io.BytesIO(b'newcontent')) + + p = cls.ar_with_dir = os.path.join(TEMPDIR, 'tar-with-dir.tar') + cls.addClassCleanup(os_helper.unlink, p) + with tarfile.open(p, 'w') as tar: + tar.addfile(tar.gettarinfo(os.curdir, 'test')) + + p = os.path.join(TEMPDIR, 'tar-with-implicit-dir.tar') + cls.ar_with_implicit_dir = p + cls.addClassCleanup(os_helper.unlink, p) + with tarfile.open(p, 'w') as tar: + t = tarfile.TarInfo('test/file') + t.size = 10 + tar.addfile(t, io.BytesIO(b'newcontent')) + + def open(self, path): + return tarfile.open(path, 'r') + + def extractall(self, ar): + ar.extractall(self.testdir, filter='fully_trusted') + + def setUpModule(): os_helper.unlink(TEMPDIR) os.makedirs(TEMPDIR) diff --git a/Lib/test/test_unittest/test_discovery.py b/Lib/test/test_unittest/test_discovery.py index dcb72d73efceab..004898ed431834 100644 --- a/Lib/test/test_unittest/test_discovery.py +++ b/Lib/test/test_unittest/test_discovery.py @@ -571,7 +571,7 @@ def _get_module_from_name(name): result = unittest.TestResult() suite.run(result) self.assertEqual(len(result.skipped), 1) - self.assertEqual(result.testsRun, 0) + self.assertEqual(result.testsRun, 1) self.assertEqual(import_calls, ['my_package']) # Check picklability diff --git a/Lib/test/test_unittest/test_skipping.py b/Lib/test/test_unittest/test_skipping.py index 1a6af06d32b433..f146dcac18ecc0 100644 --- a/Lib/test/test_unittest/test_skipping.py +++ b/Lib/test/test_unittest/test_skipping.py @@ -103,16 +103,16 @@ def test_dont_skip(self): pass result = LoggingResult(events) self.assertIs(suite.run(result), result) self.assertEqual(len(result.skipped), 1) - expected = ['addSkip', 'stopTest', 'startTest', - 'addSuccess', 'stopTest'] + expected = ['startTest', 'addSkip', 'stopTest', + 'startTest', 'addSuccess', 'stopTest'] self.assertEqual(events, expected) - self.assertEqual(result.testsRun, 1) + self.assertEqual(result.testsRun, 2) self.assertEqual(result.skipped, [(test_do_skip, "testing")]) self.assertTrue(result.wasSuccessful()) events = [] result = test_do_skip.run() - self.assertEqual(events, ['startTestRun', 'addSkip', + self.assertEqual(events, ['startTestRun', 'startTest', 'addSkip', 'stopTest', 'stopTestRun']) self.assertEqual(result.skipped, [(test_do_skip, "testing")]) @@ -135,13 +135,13 @@ def test_1(self): test = Foo("test_1") suite = unittest.TestSuite([test]) self.assertIs(suite.run(result), result) - self.assertEqual(events, ['addSkip', 'stopTest']) + self.assertEqual(events, ['startTest', 'addSkip', 'stopTest']) self.assertEqual(result.skipped, [(test, "testing")]) self.assertEqual(record, []) events = [] result = test.run() - self.assertEqual(events, ['startTestRun', 'addSkip', + self.assertEqual(events, ['startTestRun', 'startTest', 'addSkip', 'stopTest', 'stopTestRun']) self.assertEqual(result.skipped, [(test, "testing")]) self.assertEqual(record, []) diff --git a/Lib/test/test_xml_etree.py b/Lib/test/test_xml_etree.py index 221545b315fa44..a435ec7822ea0c 100644 --- a/Lib/test/test_xml_etree.py +++ b/Lib/test/test_xml_etree.py @@ -555,6 +555,17 @@ def test_iterparse(self): ('end', '{namespace}root'), ]) + with open(SIMPLE_XMLFILE, 'rb') as source: + context = iterparse(source) + action, elem = next(context) + self.assertEqual((action, elem.tag), ('end', 'element')) + self.assertEqual([(action, elem.tag) for action, elem in context], [ + ('end', 'element'), + ('end', 'empty-element'), + ('end', 'root'), + ]) + self.assertEqual(context.root.tag, 'root') + events = () context = iterparse(SIMPLE_XMLFILE, events) self.assertEqual([(action, elem.tag) for action, elem in context], []) @@ -646,12 +657,81 @@ def test_iterparse(self): # Not exhausting the iterator still closes the resource (bpo-43292) with warnings_helper.check_no_resource_warning(self): - it = iterparse(TESTFN) + it = iterparse(SIMPLE_XMLFILE) del it + with warnings_helper.check_no_resource_warning(self): + it = iterparse(SIMPLE_XMLFILE) + it.close() + del it + + with warnings_helper.check_no_resource_warning(self): + it = iterparse(SIMPLE_XMLFILE) + action, elem = next(it) + self.assertEqual((action, elem.tag), ('end', 'element')) + del it, elem + + with warnings_helper.check_no_resource_warning(self): + it = iterparse(SIMPLE_XMLFILE) + action, elem = next(it) + it.close() + self.assertEqual((action, elem.tag), ('end', 'element')) + del it, elem + with self.assertRaises(FileNotFoundError): iterparse("nonexistent") + def test_iterparse_close(self): + iterparse = ET.iterparse + + it = iterparse(SIMPLE_XMLFILE) + it.close() + with self.assertRaises(StopIteration): + next(it) + it.close() # idempotent + + with open(SIMPLE_XMLFILE, 'rb') as source: + it = iterparse(source) + it.close() + self.assertFalse(source.closed) + with self.assertRaises(StopIteration): + next(it) + it.close() # idempotent + + it = iterparse(SIMPLE_XMLFILE) + action, elem = next(it) + self.assertEqual((action, elem.tag), ('end', 'element')) + it.close() + with self.assertRaises(StopIteration): + next(it) + it.close() # idempotent + + with open(SIMPLE_XMLFILE, 'rb') as source: + it = iterparse(source) + action, elem = next(it) + self.assertEqual((action, elem.tag), ('end', 'element')) + it.close() + self.assertFalse(source.closed) + with self.assertRaises(StopIteration): + next(it) + it.close() # idempotent + + it = iterparse(SIMPLE_XMLFILE) + list(it) + it.close() + with self.assertRaises(StopIteration): + next(it) + it.close() # idempotent + + with open(SIMPLE_XMLFILE, 'rb') as source: + it = iterparse(source) + list(it) + it.close() + self.assertFalse(source.closed) + with self.assertRaises(StopIteration): + next(it) + it.close() # idempotent + def test_writefile(self): elem = ET.Element("tag") elem.text = "text" @@ -3044,8 +3124,7 @@ def test_basic(self): # With an explicit parser too (issue #9708) sourcefile = serialize(doc, to_string=False) parser = ET.XMLParser(target=ET.TreeBuilder()) - self.assertEqual(next(ET.iterparse(sourcefile, parser=parser))[0], - 'end') + self.assertEqual(next(ET.iterparse(sourcefile, parser=parser))[0], 'end') tree = ET.ElementTree(None) self.assertRaises(AttributeError, tree.iter) diff --git a/Lib/test/test_zipfile/test_core.py b/Lib/test/test_zipfile/test_core.py index 9bdb08aeabb781..087fa8d65cc336 100644 --- a/Lib/test/test_zipfile/test_core.py +++ b/Lib/test/test_zipfile/test_core.py @@ -18,6 +18,7 @@ from tempfile import TemporaryFile from random import randint, random, randbytes +from test import archiver_tests from test.support import script_helper from test.support import ( findfile, requires_zlib, requires_bz2, requires_lzma, @@ -1687,6 +1688,33 @@ def _test_extract_hackers_arcnames(self, hacknames): unlink(TESTFN2) +class OverwriteTests(archiver_tests.OverwriteTests, unittest.TestCase): + testdir = TESTFN + + @classmethod + def setUpClass(cls): + p = cls.ar_with_file = TESTFN + '-with-file.zip' + cls.addClassCleanup(unlink, p) + with zipfile.ZipFile(p, 'w') as zipfp: + zipfp.writestr('test', b'newcontent') + + p = cls.ar_with_dir = TESTFN + '-with-dir.zip' + cls.addClassCleanup(unlink, p) + with zipfile.ZipFile(p, 'w') as zipfp: + zipfp.mkdir('test') + + p = cls.ar_with_implicit_dir = TESTFN + '-with-implicit-dir.zip' + cls.addClassCleanup(unlink, p) + with zipfile.ZipFile(p, 'w') as zipfp: + zipfp.writestr('test/file', b'newcontent') + + def open(self, path): + return zipfile.ZipFile(path, 'r') + + def extractall(self, ar): + ar.extractall(self.testdir) + + class OtherTests(unittest.TestCase): def test_open_via_zip_info(self): # Create the ZIP archive @@ -2959,7 +2987,7 @@ def test_create_directory_with_write(self): directory = os.path.join(TESTFN2, "directory2") os.mkdir(directory) - mode = os.stat(directory).st_mode + mode = os.stat(directory).st_mode & 0xFFFF zf.write(directory, arcname="directory2/") zinfo = zf.filelist[1] self.assertEqual(zinfo.filename, "directory2/") diff --git a/Lib/unittest/case.py b/Lib/unittest/case.py index 811557498bb30e..001b640dc43ad6 100644 --- a/Lib/unittest/case.py +++ b/Lib/unittest/case.py @@ -606,6 +606,7 @@ def run(self, result=None): else: stopTestRun = None + result.startTest(self) try: testMethod = getattr(self, self._testMethodName) if (getattr(self.__class__, "__unittest_skip__", False) or @@ -616,9 +617,6 @@ def run(self, result=None): _addSkip(result, self, skip_why) return result - # Increase the number of tests only if it hasn't been skipped - result.startTest(self) - expecting_failure = ( getattr(self, "__unittest_expecting_failure__", False) or getattr(testMethod, "__unittest_expecting_failure__", False) diff --git a/Lib/unittest/result.py b/Lib/unittest/result.py index 9e56f658027f4d..3ace0a5b7bf2ef 100644 --- a/Lib/unittest/result.py +++ b/Lib/unittest/result.py @@ -97,12 +97,10 @@ def _restoreStdout(self): sys.stdout = self._original_stdout sys.stderr = self._original_stderr - if self._stdout_buffer is not None: - self._stdout_buffer.seek(0) - self._stdout_buffer.truncate() - if self._stderr_buffer is not None: - self._stderr_buffer.seek(0) - self._stderr_buffer.truncate() + self._stdout_buffer.seek(0) + self._stdout_buffer.truncate() + self._stderr_buffer.seek(0) + self._stderr_buffer.truncate() def stopTestRun(self): """Called once after all tests are executed. diff --git a/Lib/venv/scripts/nt/activate.bat b/Lib/venv/scripts/nt/activate.bat index 2c98122362a060..dd5ea8eb67b90a 100644 --- a/Lib/venv/scripts/nt/activate.bat +++ b/Lib/venv/scripts/nt/activate.bat @@ -15,8 +15,8 @@ if not defined PROMPT set PROMPT=$P$G if defined _OLD_VIRTUAL_PROMPT set PROMPT=%_OLD_VIRTUAL_PROMPT% if defined _OLD_VIRTUAL_PYTHONHOME set PYTHONHOME=%_OLD_VIRTUAL_PYTHONHOME% -set _OLD_VIRTUAL_PROMPT=%PROMPT% -set PROMPT=(__VENV_PROMPT__) %PROMPT% +set "_OLD_VIRTUAL_PROMPT=%PROMPT%" +set "PROMPT=(__VENV_PROMPT__) %PROMPT%" if defined PYTHONHOME set _OLD_VIRTUAL_PYTHONHOME=%PYTHONHOME% set PYTHONHOME= diff --git a/Lib/xml/etree/ElementTree.py b/Lib/xml/etree/ElementTree.py index bb7362d1634a72..a37fead41b750e 100644 --- a/Lib/xml/etree/ElementTree.py +++ b/Lib/xml/etree/ElementTree.py @@ -1248,10 +1248,17 @@ def iterator(source): if close_source: source.close() + gen = iterator(source) class IterParseIterator(collections.abc.Iterator): - __next__ = iterator(source).__next__ + __next__ = gen.__next__ + def close(self): + if close_source: + source.close() + gen.close() def __del__(self): + # TODO: Emit a ResourceWarning if it was not explicitly closed. + # (When the close() method will be supported in all maintained Python versions.) if close_source: source.close() diff --git a/Makefile.pre.in b/Makefile.pre.in index fff3d3c4914e7a..aad637876ead80 100644 --- a/Makefile.pre.in +++ b/Makefile.pre.in @@ -2643,7 +2643,18 @@ config.status: $(srcdir)/configure Python/asm_trampoline.o: $(srcdir)/Python/asm_trampoline.S $(CC) -c $(PY_CORE_CFLAGS) -o $@ $< -Python/jit.o: regen-jit + +JIT_DEPS = \ + $(srcdir)/Tools/jit/*.c \ + $(srcdir)/Tools/jit/*.py \ + $(srcdir)/Python/executor_cases.c.h \ + pyconfig.h + +jit_stencils.h: $(JIT_DEPS) + @REGEN_JIT_COMMAND@ + +Python/jit.o: $(srcdir)/Python/jit.c @JIT_STENCILS_H@ + $(CC) -c $(PY_CORE_CFLAGS) -o $@ $< .PHONY: regen-jit regen-jit: diff --git a/Misc/NEWS.d/3.13.0a2.rst b/Misc/NEWS.d/3.13.0a2.rst index d4be4fb8a3d3ab..e5841e14c02efb 100644 --- a/Misc/NEWS.d/3.13.0a2.rst +++ b/Misc/NEWS.d/3.13.0a2.rst @@ -565,9 +565,9 @@ part of a :exc:`BaseExceptionGroup`, in addition to the recent support for .. section: Library The :class:`mmap.mmap` class now has an :meth:`~mmap.mmap.seekable` method -that can be used where it requires a file-like object with seekable and the -:meth:`~mmap.mmap.seek` method return the new absolute position. Patch by -Donghee Na. +that can be used when a seekable file-like object is required. +The :meth:`~mmap.mmap.seek` method now returns the new absolute position. +Patch by Donghee Na. .. diff --git a/Misc/NEWS.d/next/Build/2024-02-01-20-08-11.gh-issue-114875.x_2iZ9.rst b/Misc/NEWS.d/next/Build/2024-02-01-20-08-11.gh-issue-114875.x_2iZ9.rst new file mode 100644 index 00000000000000..20e9d6376b973c --- /dev/null +++ b/Misc/NEWS.d/next/Build/2024-02-01-20-08-11.gh-issue-114875.x_2iZ9.rst @@ -0,0 +1 @@ +Add :c:func:`!getgrent` as a prerequisite for building the :mod:`grp` module. diff --git a/Misc/NEWS.d/next/C API/2024-01-23-21-45-02.gh-issue-114329.YRaBoe.rst b/Misc/NEWS.d/next/C API/2024-01-23-21-45-02.gh-issue-114329.YRaBoe.rst new file mode 100644 index 00000000000000..62d4ce0cfb8de5 --- /dev/null +++ b/Misc/NEWS.d/next/C API/2024-01-23-21-45-02.gh-issue-114329.YRaBoe.rst @@ -0,0 +1,3 @@ +Add :c:func:`PyList_GetItemRef`, which is similar to +:c:func:`PyList_GetItem` but returns a :term:`strong reference` instead of a +:term:`borrowed reference`. diff --git a/Misc/NEWS.d/next/Core and Builtins/2023-05-16-06-52-34.gh-issue-104530.mJnA0W.rst b/Misc/NEWS.d/next/Core and Builtins/2023-05-16-06-52-34.gh-issue-104530.mJnA0W.rst new file mode 100644 index 00000000000000..8643a25ae51b13 --- /dev/null +++ b/Misc/NEWS.d/next/Core and Builtins/2023-05-16-06-52-34.gh-issue-104530.mJnA0W.rst @@ -0,0 +1 @@ +Use native Win32 condition variables. diff --git a/Misc/NEWS.d/next/Core and Builtins/2024-01-12-16-40-07.gh-issue-113939.Yi3L-e.rst b/Misc/NEWS.d/next/Core and Builtins/2024-01-12-16-40-07.gh-issue-113939.Yi3L-e.rst new file mode 100644 index 00000000000000..28b8e4bdda6be4 --- /dev/null +++ b/Misc/NEWS.d/next/Core and Builtins/2024-01-12-16-40-07.gh-issue-113939.Yi3L-e.rst @@ -0,0 +1,4 @@ +frame.clear(): +Clear frame.f_locals as well, and not only the fast locals. +This is relevant once frame.f_locals was accessed, +which would contain also references to all the locals. diff --git a/Misc/NEWS.d/next/Core and Builtins/2024-01-21-17-29-32.gh-issue-114388.UVGO4K.rst b/Misc/NEWS.d/next/Core and Builtins/2024-01-21-17-29-32.gh-issue-114388.UVGO4K.rst new file mode 100644 index 00000000000000..52c2742001d9ca --- /dev/null +++ b/Misc/NEWS.d/next/Core and Builtins/2024-01-21-17-29-32.gh-issue-114388.UVGO4K.rst @@ -0,0 +1,5 @@ +Fix a :exc:`RuntimeWarning` emitted when assign an integer-like value that +is not an instance of :class:`int` to an attribute that corresponds to a C +struct member of :ref:`type ` T_UINT and T_ULONG. Fix a +double :exc:`RuntimeWarning` emitted when assign a negative integer value to +an attribute that corresponds to a C struct member of type T_UINT. diff --git a/Misc/NEWS.d/next/Core and Builtins/2024-01-25-18-50-49.gh-issue-112529.IbbApA.rst b/Misc/NEWS.d/next/Core and Builtins/2024-01-25-18-50-49.gh-issue-112529.IbbApA.rst new file mode 100644 index 00000000000000..2a6d74fb222702 --- /dev/null +++ b/Misc/NEWS.d/next/Core and Builtins/2024-01-25-18-50-49.gh-issue-112529.IbbApA.rst @@ -0,0 +1,4 @@ +The free-threaded build no longer allocates space for the ``PyGC_Head`` +structure in objects that support cyclic garbage collection. A number of +other fields and data structures are used as replacements, including +``ob_gc_bits``, ``ob_tid``, and mimalloc internal data structures. diff --git a/Misc/NEWS.d/next/Core and Builtins/2024-02-01-18-16-52.gh-issue-114806.wrH2J6.rst b/Misc/NEWS.d/next/Core and Builtins/2024-02-01-18-16-52.gh-issue-114806.wrH2J6.rst new file mode 100644 index 00000000000000..795f2529df8207 --- /dev/null +++ b/Misc/NEWS.d/next/Core and Builtins/2024-02-01-18-16-52.gh-issue-114806.wrH2J6.rst @@ -0,0 +1,3 @@ +No longer specialize calls to classes, if those classes have metaclasses. +Fixes bug where the ``__call__`` method of the metaclass was not being +called. diff --git a/Misc/NEWS.d/next/Core and Builtins/2024-02-03-04-07-18.gh-issue-114887.uLSFmN.rst b/Misc/NEWS.d/next/Core and Builtins/2024-02-03-04-07-18.gh-issue-114887.uLSFmN.rst new file mode 100644 index 00000000000000..b4d8cf4089d723 --- /dev/null +++ b/Misc/NEWS.d/next/Core and Builtins/2024-02-03-04-07-18.gh-issue-114887.uLSFmN.rst @@ -0,0 +1,2 @@ +Changed socket type validation in :meth:`~asyncio.loop.create_datagram_endpoint` to accept all non-stream sockets. +This fixes a regression in compatibility with raw sockets. diff --git a/Misc/NEWS.d/next/IDLE/2023-04-25-03-01-23.gh-issue-103820.LCSpza.rst b/Misc/NEWS.d/next/IDLE/2023-04-25-03-01-23.gh-issue-103820.LCSpza.rst new file mode 100644 index 00000000000000..b9d7faf047b28e --- /dev/null +++ b/Misc/NEWS.d/next/IDLE/2023-04-25-03-01-23.gh-issue-103820.LCSpza.rst @@ -0,0 +1,2 @@ +Revise IDLE bindings so that events from mouse button 4/5 on non-X11 +windowing systems (i.e. Win32 and Aqua) are not mistaken for scrolling. diff --git a/Misc/NEWS.d/next/Library/2023-04-08-11-41-07.gh-issue-101599.PaWNFh.rst b/Misc/NEWS.d/next/Library/2023-04-08-11-41-07.gh-issue-101599.PaWNFh.rst new file mode 100644 index 00000000000000..a1608a1ae0d2fa --- /dev/null +++ b/Misc/NEWS.d/next/Library/2023-04-08-11-41-07.gh-issue-101599.PaWNFh.rst @@ -0,0 +1 @@ +Changed argparse flag options formatting to remove redundancy. diff --git a/Misc/NEWS.d/next/Library/2024-01-12-09-35-07.gh-issue-112202.t_0V1m.rst b/Misc/NEWS.d/next/Library/2024-01-12-09-35-07.gh-issue-112202.t_0V1m.rst new file mode 100644 index 00000000000000..9abde13bbf8571 --- /dev/null +++ b/Misc/NEWS.d/next/Library/2024-01-12-09-35-07.gh-issue-112202.t_0V1m.rst @@ -0,0 +1 @@ +Ensure that a :func:`asyncio.Condition.notify` call does not get lost if the awakened ``Task`` is simultaneously cancelled or encounters any other error. diff --git a/Misc/NEWS.d/next/Library/2024-01-15-18-42-44.gh-issue-109534.wYaLMZ.rst b/Misc/NEWS.d/next/Library/2024-01-15-18-42-44.gh-issue-109534.wYaLMZ.rst new file mode 100644 index 00000000000000..fc9a765a230037 --- /dev/null +++ b/Misc/NEWS.d/next/Library/2024-01-15-18-42-44.gh-issue-109534.wYaLMZ.rst @@ -0,0 +1,3 @@ +Fix a reference leak in +:class:`asyncio.selector_events.BaseSelectorEventLoop` when SSL handshakes +fail. Patch contributed by Jamie Phan. diff --git a/Misc/NEWS.d/next/Library/2024-01-23-11-04-21.gh-issue-113267.xe_Pxe.rst b/Misc/NEWS.d/next/Library/2024-01-23-11-04-21.gh-issue-113267.xe_Pxe.rst new file mode 100644 index 00000000000000..ad8aaf9250f6d8 --- /dev/null +++ b/Misc/NEWS.d/next/Library/2024-01-23-11-04-21.gh-issue-113267.xe_Pxe.rst @@ -0,0 +1,2 @@ +Revert changes in :gh:`106584` which made calls of ``TestResult`` methods +``startTest()`` and ``stopTest()`` unbalanced. diff --git a/Misc/NEWS.d/next/Library/2024-01-24-17-25-18.gh-issue-69893.PQq5fR.rst b/Misc/NEWS.d/next/Library/2024-01-24-17-25-18.gh-issue-69893.PQq5fR.rst new file mode 100644 index 00000000000000..1ebf434c33187b --- /dev/null +++ b/Misc/NEWS.d/next/Library/2024-01-24-17-25-18.gh-issue-69893.PQq5fR.rst @@ -0,0 +1,2 @@ +Add the :meth:`!close` method for the iterator returned by +:func:`xml.etree.ElementTree.iterparse`. diff --git a/Misc/NEWS.d/next/Library/2024-01-25-19-22-17.gh-issue-83383.3GwO9v.rst b/Misc/NEWS.d/next/Library/2024-01-25-19-22-17.gh-issue-83383.3GwO9v.rst new file mode 100644 index 00000000000000..e6336204dfa236 --- /dev/null +++ b/Misc/NEWS.d/next/Library/2024-01-25-19-22-17.gh-issue-83383.3GwO9v.rst @@ -0,0 +1,5 @@ +Synchronization of the :mod:`dbm.dumb` database is now no-op if there was no +modification since opening or last synchronization. +The directory file for a newly created empty :mod:`dbm.dumb` database is now +created immediately after opening instead of deferring this until +synchronizing or closing. diff --git a/Misc/NEWS.d/next/Library/2024-01-27-20-11-24.gh-issue-113280.CZPQMf.rst b/Misc/NEWS.d/next/Library/2024-01-27-20-11-24.gh-issue-113280.CZPQMf.rst new file mode 100644 index 00000000000000..3dcdbcf0995616 --- /dev/null +++ b/Misc/NEWS.d/next/Library/2024-01-27-20-11-24.gh-issue-113280.CZPQMf.rst @@ -0,0 +1,2 @@ +Fix a leak of open socket in rare cases when error occurred in +:class:`ssl.SSLSocket` creation. diff --git a/Misc/NEWS.d/next/Library/2024-02-01-10-19-11.gh-issue-114071.vkm2G_.rst b/Misc/NEWS.d/next/Library/2024-02-01-10-19-11.gh-issue-114071.vkm2G_.rst new file mode 100644 index 00000000000000..587ce4d2157637 --- /dev/null +++ b/Misc/NEWS.d/next/Library/2024-02-01-10-19-11.gh-issue-114071.vkm2G_.rst @@ -0,0 +1 @@ +Support tuple subclasses using auto() for enum member value. diff --git a/Misc/NEWS.d/next/Library/2024-02-03-16-59-25.gh-issue-114959.dCfAG2.rst b/Misc/NEWS.d/next/Library/2024-02-03-16-59-25.gh-issue-114959.dCfAG2.rst new file mode 100644 index 00000000000000..5c6eaa7525e3b0 --- /dev/null +++ b/Misc/NEWS.d/next/Library/2024-02-03-16-59-25.gh-issue-114959.dCfAG2.rst @@ -0,0 +1,2 @@ +:mod:`tarfile` no longer ignores errors when trying to extract a directory on +top of a file. diff --git a/Misc/NEWS.d/next/Library/2024-02-03-17-54-17.gh-issue-114965.gHksCK.rst b/Misc/NEWS.d/next/Library/2024-02-03-17-54-17.gh-issue-114965.gHksCK.rst new file mode 100644 index 00000000000000..d59ff991993792 --- /dev/null +++ b/Misc/NEWS.d/next/Library/2024-02-03-17-54-17.gh-issue-114965.gHksCK.rst @@ -0,0 +1 @@ +Update bundled pip to 24.0 diff --git a/Misc/NEWS.d/next/Tests/2023-06-02-05-04-15.gh-issue-105089.KaZFtU.rst b/Misc/NEWS.d/next/Tests/2023-06-02-05-04-15.gh-issue-105089.KaZFtU.rst new file mode 100644 index 00000000000000..d04ef435dd572d --- /dev/null +++ b/Misc/NEWS.d/next/Tests/2023-06-02-05-04-15.gh-issue-105089.KaZFtU.rst @@ -0,0 +1,4 @@ +Fix +``test.test_zipfile.test_core.TestWithDirectory.test_create_directory_with_write`` +test in AIX by doing a bitwise AND of 0xFFFF on mode , so that it will be in +sync with ``zinfo.external_attr`` diff --git a/Misc/NEWS.d/next/Windows/2024-02-01-14-35-05.gh-issue-111239.SO7SUF.rst b/Misc/NEWS.d/next/Windows/2024-02-01-14-35-05.gh-issue-111239.SO7SUF.rst new file mode 100644 index 00000000000000..ea82c3b941f802 --- /dev/null +++ b/Misc/NEWS.d/next/Windows/2024-02-01-14-35-05.gh-issue-111239.SO7SUF.rst @@ -0,0 +1 @@ +Update Windows builds to use zlib v1.3.1. diff --git a/Misc/sbom.spdx.json b/Misc/sbom.spdx.json index 94566772338b10..e94dcb83dd4e40 100644 --- a/Misc/sbom.spdx.json +++ b/Misc/sbom.spdx.json @@ -1570,18 +1570,18 @@ "fileName": "Modules/_decimal/libmpdec/vcdiv64.asm" }, { - "SPDXID": "SPDXRef-FILE-Lib-ensurepip-bundled-pip-23.3.2-py3-none-any.whl", + "SPDXID": "SPDXRef-FILE-Lib-ensurepip-bundled-pip-24.0-py3-none-any.whl", "checksums": [ { "algorithm": "SHA1", - "checksumValue": "8e48f55ab2965ee64bd55cc91a8077d184a33e30" + "checksumValue": "e44313ae1e6af3c2bd3b60ab2fa8c34308d00555" }, { "algorithm": "SHA256", - "checksumValue": "5052d7889c1f9d05224cd41741acb7c5d6fa735ab34e339624a614eaaa7e7d76" + "checksumValue": "ba0d021a166865d2265246961bec0152ff124de910c5cc39f1156ce3fa7c69dc" } ], - "fileName": "Lib/ensurepip/_bundled/pip-23.3.2-py3-none-any.whl" + "fileName": "Lib/ensurepip/_bundled/pip-24.0-py3-none-any.whl" } ], "packages": [ @@ -1742,21 +1742,21 @@ "checksums": [ { "algorithm": "SHA256", - "checksumValue": "f35c4b692542ca110de7ef0bea44d73981caeb34ca0b9b6b2e6d7790dda8f80e" + "checksumValue": "034db59a0b96f8ca18035f36290806a9a6e6bd9d1ff91e45a7f172eb17e51784" } ], - "downloadLocation": "https://files.pythonhosted.org/packages/76/cb/6bbd2b10170ed991cf64e8c8b85e01f2fb38f95d1bc77617569e0b0b26ac/distlib-0.3.6-py2.py3-none-any.whl", + "downloadLocation": "https://files.pythonhosted.org/packages/8e/41/9307e4f5f9976bc8b7fea0b66367734e8faf3ec84bc0d412d8cfabbb66cd/distlib-0.3.8-py2.py3-none-any.whl", "externalRefs": [ { "referenceCategory": "PACKAGE_MANAGER", - "referenceLocator": "pkg:pypi/distlib@0.3.6", + "referenceLocator": "pkg:pypi/distlib@0.3.8", "referenceType": "purl" } ], "licenseConcluded": "MIT", "name": "distlib", "primaryPackagePurpose": "SOURCE", - "versionInfo": "0.3.6" + "versionInfo": "0.3.8" }, { "SPDXID": "SPDXRef-PACKAGE-distro", @@ -2204,19 +2204,19 @@ "checksums": [ { "algorithm": "SHA256", - "checksumValue": "5052d7889c1f9d05224cd41741acb7c5d6fa735ab34e339624a614eaaa7e7d76" + "checksumValue": "ba0d021a166865d2265246961bec0152ff124de910c5cc39f1156ce3fa7c69dc" } ], - "downloadLocation": "https://files.pythonhosted.org/packages/15/aa/3f4c7bcee2057a76562a5b33ecbd199be08cdb4443a02e26bd2c3cf6fc39/pip-23.3.2-py3-none-any.whl", + "downloadLocation": "https://files.pythonhosted.org/packages/8a/6a/19e9fe04fca059ccf770861c7d5721ab4c2aebc539889e97c7977528a53b/pip-24.0-py3-none-any.whl", "externalRefs": [ { "referenceCategory": "SECURITY", - "referenceLocator": "cpe:2.3:a:pypa:pip:23.3.2:*:*:*:*:*:*:*", + "referenceLocator": "cpe:2.3:a:pypa:pip:24.0:*:*:*:*:*:*:*", "referenceType": "cpe23Type" }, { "referenceCategory": "PACKAGE_MANAGER", - "referenceLocator": "pkg:pypi/pip@23.3.2", + "referenceLocator": "pkg:pypi/pip@24.0", "referenceType": "purl" } ], @@ -2224,7 +2224,7 @@ "name": "pip", "originator": "Organization: Python Packaging Authority", "primaryPackagePurpose": "SOURCE", - "versionInfo": "23.3.2" + "versionInfo": "24.0" } ], "relationships": [ @@ -2909,7 +2909,7 @@ "spdxElementId": "SPDXRef-PACKAGE-mpdecimal" }, { - "relatedSpdxElement": "SPDXRef-FILE-Lib-ensurepip-bundled-pip-23.3.2-py3-none-any.whl", + "relatedSpdxElement": "SPDXRef-FILE-Lib-ensurepip-bundled-pip-24.0-py3-none-any.whl", "relationshipType": "CONTAINS", "spdxElementId": "SPDXRef-PACKAGE-pip" } diff --git a/Misc/stable_abi.toml b/Misc/stable_abi.toml index ae19d25809ec86..a9875f6ffd1a56 100644 --- a/Misc/stable_abi.toml +++ b/Misc/stable_abi.toml @@ -2487,3 +2487,5 @@ abi_only = true [data.PyExc_IncompleteInputError] added = '3.13' +[function.PyList_GetItemRef] + added = '3.13' diff --git a/Modules/_testcapi/list.c b/Modules/_testcapi/list.c index 10e18699f01bc1..2cb6499e28336d 100644 --- a/Modules/_testcapi/list.c +++ b/Modules/_testcapi/list.c @@ -59,6 +59,18 @@ list_get_item(PyObject *Py_UNUSED(module), PyObject *args) return Py_XNewRef(PyList_GET_ITEM(obj, i)); } +static PyObject * +list_get_item_ref(PyObject *Py_UNUSED(module), PyObject *args) +{ + PyObject *obj; + Py_ssize_t i; + if (!PyArg_ParseTuple(args, "On", &obj, &i)) { + return NULL; + } + NULLABLE(obj); + return PyList_GetItemRef(obj, i); +} + static PyObject * list_setitem(PyObject *Py_UNUSED(module), PyObject *args) { @@ -191,6 +203,7 @@ static PyMethodDef test_methods[] = { {"list_get_size", list_get_size, METH_O}, {"list_getitem", list_getitem, METH_VARARGS}, {"list_get_item", list_get_item, METH_VARARGS}, + {"list_get_item_ref", list_get_item_ref, METH_VARARGS}, {"list_setitem", list_setitem, METH_VARARGS}, {"list_set_item", list_set_item, METH_VARARGS}, {"list_insert", list_insert, METH_VARARGS}, diff --git a/Modules/_testcapi/watchers.c b/Modules/_testcapi/watchers.c index a763ff46a3c290..1eb0db2c2e6576 100644 --- a/Modules/_testcapi/watchers.c +++ b/Modules/_testcapi/watchers.c @@ -15,8 +15,8 @@ module _testcapi /*[clinic end generated code: output=da39a3ee5e6b4b0d input=6361033e795369fc]*/ // Test dict watching -static PyObject *g_dict_watch_events; -static int g_dict_watchers_installed; +static PyObject *g_dict_watch_events = NULL; +static int g_dict_watchers_installed = 0; static int dict_watch_callback(PyDict_WatchEvent event, diff --git a/Modules/_testinternalcapi.c b/Modules/_testinternalcapi.c index c4a648a1816392..0bb739b5398b11 100644 --- a/Modules/_testinternalcapi.c +++ b/Modules/_testinternalcapi.c @@ -1752,8 +1752,18 @@ module_exec(PyObject *module) return 1; } + Py_ssize_t sizeof_gc_head = 0; +#ifndef Py_GIL_DISABLED + sizeof_gc_head = sizeof(PyGC_Head); +#endif + if (PyModule_Add(module, "SIZEOF_PYGC_HEAD", - PyLong_FromSsize_t(sizeof(PyGC_Head))) < 0) { + PyLong_FromSsize_t(sizeof_gc_head)) < 0) { + return 1; + } + + if (PyModule_Add(module, "SIZEOF_MANAGED_PRE_HEADER", + PyLong_FromSsize_t(2 * sizeof(PyObject*))) < 0) { return 1; } diff --git a/Modules/mathmodule.c b/Modules/mathmodule.c index 0be46b1574c1fe..a877bfcd6afb68 100644 --- a/Modules/mathmodule.c +++ b/Modules/mathmodule.c @@ -2074,11 +2074,6 @@ math_trunc(PyObject *module, PyObject *x) return PyFloat_Type.tp_as_number->nb_int(x); } - if (!_PyType_IsReady(Py_TYPE(x))) { - if (PyType_Ready(Py_TYPE(x)) < 0) - return NULL; - } - math_module_state *state = get_math_module_state(module); trunc = _PyObject_LookupSpecial(x, state->str___trunc__); if (trunc == NULL) { diff --git a/Objects/dictobject.c b/Objects/dictobject.c index 23d7e9b5e38a35..4bb818b90a4a72 100644 --- a/Objects/dictobject.c +++ b/Objects/dictobject.c @@ -118,6 +118,7 @@ As a consequence of this, split keys have a maximum size of 16. #include "pycore_ceval.h" // _PyEval_GetBuiltin() #include "pycore_code.h" // stats #include "pycore_dict.h" // export _PyDict_SizeOf() +#include "pycore_freelist.h" // _PyFreeListState_GET() #include "pycore_gc.h" // _PyObject_GC_IS_TRACKED() #include "pycore_object.h" // _PyObject_GC_TRACK(), _PyDebugAllocatorStats() #include "pycore_pyerrors.h" // _PyErr_GetRaisedException() @@ -242,40 +243,44 @@ static PyObject* dict_iter(PyObject *dict); #include "clinic/dictobject.c.h" -#if PyDict_MAXFREELIST > 0 -static struct _Py_dict_state * -get_dict_state(PyInterpreterState *interp) +#ifdef WITH_FREELISTS +static struct _Py_dict_freelist * +get_dict_state(void) { - return &interp->dict_state; + _PyFreeListState *state = _PyFreeListState_GET(); + return &state->dicts; } #endif void -_PyDict_ClearFreeList(PyInterpreterState *interp) +_PyDict_ClearFreeList(_PyFreeListState *freelist_state, int is_finalization) { -#if PyDict_MAXFREELIST > 0 - struct _Py_dict_state *state = &interp->dict_state; - while (state->numfree) { +#ifdef WITH_FREELISTS + struct _Py_dict_freelist *state = &freelist_state->dicts; + while (state->numfree > 0) { PyDictObject *op = state->free_list[--state->numfree]; assert(PyDict_CheckExact(op)); PyObject_GC_Del(op); } - while (state->keys_numfree) { + while (state->keys_numfree > 0) { PyMem_Free(state->keys_free_list[--state->keys_numfree]); } + if (is_finalization) { + state->numfree = -1; + state->keys_numfree = -1; + } #endif } - void -_PyDict_Fini(PyInterpreterState *interp) +_PyDict_Fini(PyInterpreterState *Py_UNUSED(interp)) { - _PyDict_ClearFreeList(interp); -#if defined(Py_DEBUG) && PyDict_MAXFREELIST > 0 - struct _Py_dict_state *state = &interp->dict_state; - state->numfree = -1; - state->keys_numfree = -1; + // With Py_GIL_DISABLED: + // the freelists for the current thread state have already been cleared. +#ifndef Py_GIL_DISABLED + _PyFreeListState *state = _PyFreeListState_GET(); + _PyDict_ClearFreeList(state, 1); #endif } @@ -290,9 +295,8 @@ unicode_get_hash(PyObject *o) void _PyDict_DebugMallocStats(FILE *out) { -#if PyDict_MAXFREELIST > 0 - PyInterpreterState *interp = _PyInterpreterState_GET(); - struct _Py_dict_state *state = get_dict_state(interp); +#ifdef WITH_FREELISTS + struct _Py_dict_freelist *state = get_dict_state(); _PyDebugAllocatorStats(out, "free PyDictObject", state->numfree, sizeof(PyDictObject)); #endif @@ -300,7 +304,7 @@ _PyDict_DebugMallocStats(FILE *out) #define DK_MASK(dk) (DK_SIZE(dk)-1) -static void free_keys_object(PyInterpreterState *interp, PyDictKeysObject *keys); +static void free_keys_object(PyDictKeysObject *keys); /* PyDictKeysObject has refcounts like PyObject does, so we have the following two functions to mirror what Py_INCREF() and Py_DECREF() do. @@ -348,7 +352,7 @@ dictkeys_decref(PyInterpreterState *interp, PyDictKeysObject *dk) Py_XDECREF(entries[i].me_value); } } - free_keys_object(interp, dk); + free_keys_object(dk); } } @@ -643,12 +647,8 @@ new_keys_object(PyInterpreterState *interp, uint8_t log2_size, bool unicode) log2_bytes = log2_size + 2; } -#if PyDict_MAXFREELIST > 0 - struct _Py_dict_state *state = get_dict_state(interp); -#ifdef Py_DEBUG - // new_keys_object() must not be called after _PyDict_Fini() - assert(state->keys_numfree != -1); -#endif +#ifdef WITH_FREELISTS + struct _Py_dict_freelist *state = get_dict_state(); if (log2_size == PyDict_LOG_MINSIZE && unicode && state->keys_numfree > 0) { dk = state->keys_free_list[--state->keys_numfree]; OBJECT_STAT_INC(from_freelist); @@ -680,16 +680,13 @@ new_keys_object(PyInterpreterState *interp, uint8_t log2_size, bool unicode) } static void -free_keys_object(PyInterpreterState *interp, PyDictKeysObject *keys) +free_keys_object(PyDictKeysObject *keys) { -#if PyDict_MAXFREELIST > 0 - struct _Py_dict_state *state = get_dict_state(interp); -#ifdef Py_DEBUG - // free_keys_object() must not be called after _PyDict_Fini() - assert(state->keys_numfree != -1); -#endif +#ifdef WITH_FREELISTS + struct _Py_dict_freelist *state = get_dict_state(); if (DK_LOG_SIZE(keys) == PyDict_LOG_MINSIZE && state->keys_numfree < PyDict_MAXFREELIST + && state->keys_numfree >= 0 && DK_IS_UNICODE(keys)) { state->keys_free_list[state->keys_numfree++] = keys; OBJECT_STAT_INC(to_freelist); @@ -730,13 +727,9 @@ new_dict(PyInterpreterState *interp, { PyDictObject *mp; assert(keys != NULL); -#if PyDict_MAXFREELIST > 0 - struct _Py_dict_state *state = get_dict_state(interp); -#ifdef Py_DEBUG - // new_dict() must not be called after _PyDict_Fini() - assert(state->numfree != -1); -#endif - if (state->numfree) { +#ifdef WITH_FREELISTS + struct _Py_dict_freelist *state = get_dict_state(); + if (state->numfree > 0) { mp = state->free_list[--state->numfree]; assert (mp != NULL); assert (Py_IS_TYPE(mp, &PyDict_Type)); @@ -1547,7 +1540,7 @@ dictresize(PyInterpreterState *interp, PyDictObject *mp, #endif assert(oldkeys->dk_kind != DICT_KEYS_SPLIT); assert(oldkeys->dk_refcnt == 1); - free_keys_object(interp, oldkeys); + free_keys_object(oldkeys); } } @@ -2458,13 +2451,10 @@ dict_dealloc(PyObject *self) assert(keys->dk_refcnt == 1 || keys == Py_EMPTY_KEYS); dictkeys_decref(interp, keys); } -#if PyDict_MAXFREELIST > 0 - struct _Py_dict_state *state = get_dict_state(interp); -#ifdef Py_DEBUG - // new_dict() must not be called after _PyDict_Fini() - assert(state->numfree != -1); -#endif - if (state->numfree < PyDict_MAXFREELIST && Py_IS_TYPE(mp, &PyDict_Type)) { +#ifdef WITH_FREELISTS + struct _Py_dict_freelist *state = get_dict_state(); + if (state->numfree < PyDict_MAXFREELIST && state->numfree >=0 && + Py_IS_TYPE(mp, &PyDict_Type)) { state->free_list[state->numfree++] = mp; OBJECT_STAT_INC(to_freelist); } @@ -5953,7 +5943,8 @@ PyDict_AddWatcher(PyDict_WatchCallback callback) { PyInterpreterState *interp = _PyInterpreterState_GET(); - for (int i = 0; i < DICT_MAX_WATCHERS; i++) { + /* Start at 2, as 0 and 1 are reserved for CPython */ + for (int i = 2; i < DICT_MAX_WATCHERS; i++) { if (!interp->dict_state.watchers[i]) { interp->dict_state.watchers[i] = callback; return i; diff --git a/Objects/floatobject.c b/Objects/floatobject.c index b7611d5f96ac3b..c440e0dab0e79f 100644 --- a/Objects/floatobject.c +++ b/Objects/floatobject.c @@ -2013,7 +2013,11 @@ _PyFloat_ClearFreeList(_PyFreeListState *freelist_state, int is_finalization) void _PyFloat_Fini(_PyFreeListState *state) { + // With Py_GIL_DISABLED: + // the freelists for the current thread state have already been cleared. +#ifndef Py_GIL_DISABLED _PyFloat_ClearFreeList(state, 1); +#endif } void diff --git a/Objects/frameobject.c b/Objects/frameobject.c index cafe4ef6141d9a..a914c61aac2fd5 100644 --- a/Objects/frameobject.c +++ b/Objects/frameobject.c @@ -926,6 +926,7 @@ frame_tp_clear(PyFrameObject *f) Py_CLEAR(locals[i]); } f->f_frame->stacktop = 0; + Py_CLEAR(f->f_frame->f_locals); return 0; } diff --git a/Objects/genobject.c b/Objects/genobject.c index f47197330fdd80..ab523e46cceaa3 100644 --- a/Objects/genobject.c +++ b/Objects/genobject.c @@ -1685,7 +1685,11 @@ _PyAsyncGen_ClearFreeLists(_PyFreeListState *freelist_state, int is_finalization void _PyAsyncGen_Fini(_PyFreeListState *state) { + // With Py_GIL_DISABLED: + // the freelists for the current thread state have already been cleared. +#ifndef Py_GIL_DISABLED _PyAsyncGen_ClearFreeLists(state, 1); +#endif } diff --git a/Objects/listobject.c b/Objects/listobject.c index 56785e5f37a450..82a4ba952de07d 100644 --- a/Objects/listobject.c +++ b/Objects/listobject.c @@ -30,7 +30,6 @@ get_list_state(void) } #endif - /* Ensure ob_item has room for at least newsize elements, and set * ob_size to newsize. If newsize > ob_size on entry, the content * of the new slots at exit is undefined heap trash; it's the caller's @@ -139,7 +138,11 @@ _PyList_ClearFreeList(_PyFreeListState *freelist_state, int is_finalization) void _PyList_Fini(_PyFreeListState *state) { + // With Py_GIL_DISABLED: + // the freelists for the current thread state have already been cleared. +#ifndef Py_GIL_DISABLED _PyList_ClearFreeList(state, 1); +#endif } /* Print summary info about the state of the optimized allocator */ @@ -221,8 +224,9 @@ PyList_Size(PyObject *op) PyErr_BadInternalCall(); return -1; } - else - return Py_SIZE(op); + else { + return PyList_GET_SIZE(op); + } } static inline int @@ -253,6 +257,21 @@ PyList_GetItem(PyObject *op, Py_ssize_t i) return ((PyListObject *)op) -> ob_item[i]; } +PyObject * +PyList_GetItemRef(PyObject *op, Py_ssize_t i) +{ + if (!PyList_Check(op)) { + PyErr_SetString(PyExc_TypeError, "expected a list"); + return NULL; + } + if (!valid_index(i, Py_SIZE(op))) { + _Py_DECLARE_STR(list_err, "list index out of range"); + PyErr_SetObject(PyExc_IndexError, &_Py_STR(list_err)); + return NULL; + } + return Py_NewRef(PyList_GET_ITEM(op, i)); +} + int PyList_SetItem(PyObject *op, Py_ssize_t i, PyObject *newitem) @@ -328,7 +347,7 @@ PyList_Insert(PyObject *op, Py_ssize_t where, PyObject *newitem) int _PyList_AppendTakeRefListResize(PyListObject *self, PyObject *newitem) { - Py_ssize_t len = PyList_GET_SIZE(self); + Py_ssize_t len = Py_SIZE(self); assert(self->allocated == -1 || self->allocated == len); if (list_resize(self, len + 1) < 0) { Py_DECREF(newitem); @@ -342,7 +361,11 @@ int PyList_Append(PyObject *op, PyObject *newitem) { if (PyList_Check(op) && (newitem != NULL)) { - return _PyList_AppendTakeRef((PyListObject *)op, Py_NewRef(newitem)); + int ret; + Py_BEGIN_CRITICAL_SECTION(op); + ret = _PyList_AppendTakeRef((PyListObject *)op, Py_NewRef(newitem)); + Py_END_CRITICAL_SECTION(); + return ret; } PyErr_BadInternalCall(); return -1; @@ -473,7 +496,7 @@ static PyObject * list_item(PyObject *aa, Py_ssize_t i) { PyListObject *a = (PyListObject *)aa; - if (!valid_index(i, Py_SIZE(a))) { + if (!valid_index(i, PyList_GET_SIZE(a))) { PyErr_SetObject(PyExc_IndexError, &_Py_STR(list_err)); return NULL; } @@ -511,6 +534,8 @@ PyList_GetSlice(PyObject *a, Py_ssize_t ilow, Py_ssize_t ihigh) PyErr_BadInternalCall(); return NULL; } + PyObject *ret; + Py_BEGIN_CRITICAL_SECTION(a); if (ilow < 0) { ilow = 0; } @@ -523,7 +548,9 @@ PyList_GetSlice(PyObject *a, Py_ssize_t ilow, Py_ssize_t ihigh) else if (ihigh > Py_SIZE(a)) { ihigh = Py_SIZE(a); } - return list_slice((PyListObject *)a, ilow, ihigh); + ret = list_slice((PyListObject *)a, ilow, ihigh); + Py_END_CRITICAL_SECTION(); + return ret; } static PyObject * diff --git a/Objects/object.c b/Objects/object.c index 587c5528c01345..bbf7f98ae3daf9 100644 --- a/Objects/object.c +++ b/Objects/object.c @@ -2671,7 +2671,12 @@ _PyTrash_thread_deposit_object(struct _py_trashcan *trash, PyObject *op) _PyObject_ASSERT(op, _PyObject_IS_GC(op)); _PyObject_ASSERT(op, !_PyObject_GC_IS_TRACKED(op)); _PyObject_ASSERT(op, Py_REFCNT(op) == 0); +#ifdef Py_GIL_DISABLED + _PyObject_ASSERT(op, op->ob_tid == 0); + op->ob_tid = (uintptr_t)trash->delete_later; +#else _PyGCHead_SET_PREV(_Py_AS_GC(op), (PyGC_Head*)trash->delete_later); +#endif trash->delete_later = op; } @@ -2697,8 +2702,12 @@ _PyTrash_thread_destroy_chain(struct _py_trashcan *trash) PyObject *op = trash->delete_later; destructor dealloc = Py_TYPE(op)->tp_dealloc; - trash->delete_later = - (PyObject*) _PyGCHead_PREV(_Py_AS_GC(op)); +#ifdef Py_GIL_DISABLED + trash->delete_later = (PyObject*) op->ob_tid; + op->ob_tid = 0; +#else + trash->delete_later = (PyObject*) _PyGCHead_PREV(_Py_AS_GC(op)); +#endif /* Call the deallocator directly. This used to try to * fool Py_DECREF into calling it indirectly, but diff --git a/PC/python3dll.c b/PC/python3dll.c index 09ecf98fe56ea6..aa6bfe2c4022db 100755 --- a/PC/python3dll.c +++ b/PC/python3dll.c @@ -324,6 +324,7 @@ EXPORT_FUNC(PyIter_Send) EXPORT_FUNC(PyList_Append) EXPORT_FUNC(PyList_AsTuple) EXPORT_FUNC(PyList_GetItem) +EXPORT_FUNC(PyList_GetItemRef) EXPORT_FUNC(PyList_GetSlice) EXPORT_FUNC(PyList_Insert) EXPORT_FUNC(PyList_New) diff --git a/PCbuild/get_externals.bat b/PCbuild/get_externals.bat index 3919c0592ec00d..de73d923d8f4df 100644 --- a/PCbuild/get_externals.bat +++ b/PCbuild/get_externals.bat @@ -58,7 +58,7 @@ set libraries=%libraries% sqlite-3.44.2.0 if NOT "%IncludeTkinterSrc%"=="false" set libraries=%libraries% tcl-core-8.6.13.1 if NOT "%IncludeTkinterSrc%"=="false" set libraries=%libraries% tk-8.6.13.1 set libraries=%libraries% xz-5.2.5 -set libraries=%libraries% zlib-1.2.13 +set libraries=%libraries% zlib-1.3.1 for %%e in (%libraries%) do ( if exist "%EXTERNALS_DIR%\%%e" ( diff --git a/PCbuild/python.props b/PCbuild/python.props index e8796081c4eaf3..2cb16693e546b1 100644 --- a/PCbuild/python.props +++ b/PCbuild/python.props @@ -78,7 +78,7 @@ $(ExternalsDir)openssl-bin-3.0.11\$(ArchName)\ $(opensslOutDir)include $(ExternalsDir)\nasm-2.11.06\ - $(ExternalsDir)\zlib-1.2.13\ + $(ExternalsDir)\zlib-1.3.1\ diff --git a/Python/bltinmodule.c b/Python/bltinmodule.c index e54d5cbacdc96f..31c1bf07e8fb91 100644 --- a/Python/bltinmodule.c +++ b/Python/bltinmodule.c @@ -2382,11 +2382,6 @@ builtin_round_impl(PyObject *module, PyObject *number, PyObject *ndigits) { PyObject *round, *result; - if (!_PyType_IsReady(Py_TYPE(number))) { - if (PyType_Ready(Py_TYPE(number)) < 0) - return NULL; - } - round = _PyObject_LookupSpecial(number, &_Py_ID(__round__)); if (round == NULL) { if (!PyErr_Occurred()) diff --git a/Python/bytecodes.c b/Python/bytecodes.c index ebd5b06abb2d4e..6fb4d719e43991 100644 --- a/Python/bytecodes.c +++ b/Python/bytecodes.c @@ -4071,11 +4071,35 @@ dummy_func( } op(_LOAD_CONST_INLINE, (ptr/4 -- value)) { + TIER_TWO_ONLY value = Py_NewRef(ptr); } op(_LOAD_CONST_INLINE_BORROW, (ptr/4 -- value)) { + TIER_TWO_ONLY + value = ptr; + } + + op(_LOAD_CONST_INLINE_WITH_NULL, (ptr/4 -- value, null)) { + TIER_TWO_ONLY + value = Py_NewRef(ptr); + null = NULL; + } + + op(_LOAD_CONST_INLINE_BORROW_WITH_NULL, (ptr/4 -- value, null)) { + TIER_TWO_ONLY value = ptr; + null = NULL; + } + + op(_CHECK_GLOBALS, (dict/4 -- )) { + TIER_TWO_ONLY + DEOPT_IF(GLOBALS() != dict); + } + + op(_CHECK_BUILTINS, (dict/4 -- )) { + TIER_TWO_ONLY + DEOPT_IF(BUILTINS() != dict); } /* Internal -- for testing executors */ diff --git a/Python/ceval_gil.c b/Python/ceval_gil.c index f3b169241535f3..ad90359318761a 100644 --- a/Python/ceval_gil.c +++ b/Python/ceval_gil.c @@ -610,8 +610,16 @@ PyEval_SaveThread(void) void PyEval_RestoreThread(PyThreadState *tstate) { +#ifdef MS_WINDOWS + int err = GetLastError(); +#endif + _Py_EnsureTstateNotNULL(tstate); _PyThreadState_Attach(tstate); + +#ifdef MS_WINDOWS + SetLastError(err); +#endif } diff --git a/Python/condvar.h b/Python/condvar.h index d54db94f2c871d..dcabed6d55928c 100644 --- a/Python/condvar.h +++ b/Python/condvar.h @@ -260,13 +260,13 @@ PyMUTEX_UNLOCK(PyMUTEX_T *cs) return 0; } - Py_LOCAL_INLINE(int) PyCOND_INIT(PyCOND_T *cv) { InitializeConditionVariable(cv); return 0; } + Py_LOCAL_INLINE(int) PyCOND_FINI(PyCOND_T *cv) { @@ -279,27 +279,32 @@ PyCOND_WAIT(PyCOND_T *cv, PyMUTEX_T *cs) return SleepConditionVariableSRW(cv, cs, INFINITE, 0) ? 0 : -1; } -/* This implementation makes no distinction about timeouts. Signal - * 2 to indicate that we don't know. - */ +/* return 0 for success, 1 on timeout, -1 on error */ Py_LOCAL_INLINE(int) PyCOND_TIMEDWAIT(PyCOND_T *cv, PyMUTEX_T *cs, long long us) { - return SleepConditionVariableSRW(cv, cs, (DWORD)(us/1000), 0) ? 2 : -1; + BOOL success = SleepConditionVariableSRW(cv, cs, (DWORD)(us/1000), 0); + if (!success) { + if (GetLastError() == ERROR_TIMEOUT) { + return 1; + } + return -1; + } + return 0; } Py_LOCAL_INLINE(int) PyCOND_SIGNAL(PyCOND_T *cv) { - WakeConditionVariable(cv); - return 0; + WakeConditionVariable(cv); + return 0; } Py_LOCAL_INLINE(int) PyCOND_BROADCAST(PyCOND_T *cv) { - WakeAllConditionVariable(cv); - return 0; + WakeAllConditionVariable(cv); + return 0; } diff --git a/Python/context.c b/Python/context.c index 294485e5b407df..793dfa2b72c7e3 100644 --- a/Python/context.c +++ b/Python/context.c @@ -1287,7 +1287,11 @@ _PyContext_ClearFreeList(_PyFreeListState *freelist_state, int is_finalization) void _PyContext_Fini(_PyFreeListState *state) { + // With Py_GIL_DISABLED: + // the freelists for the current thread state have already been cleared. +#ifndef Py_GIL_DISABLED _PyContext_ClearFreeList(state, 1); +#endif } diff --git a/Python/executor_cases.c.h b/Python/executor_cases.c.h index 241b9056207715..2d914b82dbf88f 100644 --- a/Python/executor_cases.c.h +++ b/Python/executor_cases.c.h @@ -3393,6 +3393,7 @@ case _LOAD_CONST_INLINE: { PyObject *value; PyObject *ptr = (PyObject *)CURRENT_OPERAND(); + TIER_TWO_ONLY value = Py_NewRef(ptr); stack_pointer[0] = value; stack_pointer += 1; @@ -3402,12 +3403,53 @@ case _LOAD_CONST_INLINE_BORROW: { PyObject *value; PyObject *ptr = (PyObject *)CURRENT_OPERAND(); + TIER_TWO_ONLY value = ptr; stack_pointer[0] = value; stack_pointer += 1; break; } + case _LOAD_CONST_INLINE_WITH_NULL: { + PyObject *value; + PyObject *null; + PyObject *ptr = (PyObject *)CURRENT_OPERAND(); + TIER_TWO_ONLY + value = Py_NewRef(ptr); + null = NULL; + stack_pointer[0] = value; + stack_pointer[1] = null; + stack_pointer += 2; + break; + } + + case _LOAD_CONST_INLINE_BORROW_WITH_NULL: { + PyObject *value; + PyObject *null; + PyObject *ptr = (PyObject *)CURRENT_OPERAND(); + TIER_TWO_ONLY + value = ptr; + null = NULL; + stack_pointer[0] = value; + stack_pointer[1] = null; + stack_pointer += 2; + break; + } + + case _CHECK_GLOBALS: { + PyObject *dict = (PyObject *)CURRENT_OPERAND(); + TIER_TWO_ONLY + if (GLOBALS() != dict) goto deoptimize; + break; + } + + case _CHECK_BUILTINS: { + PyObject *dict = (PyObject *)CURRENT_OPERAND(); + TIER_TWO_ONLY + if (BUILTINS() != dict) goto deoptimize; + break; + } + case _INTERNAL_INCREMENT_OPT_COUNTER: { PyObject *opt; opt = stack_pointer[-1]; diff --git a/Python/flowgraph.c b/Python/flowgraph.c index bfc23a298ff492..1a648edf0880c0 100644 --- a/Python/flowgraph.c +++ b/Python/flowgraph.c @@ -212,14 +212,14 @@ basicblock_add_jump(basicblock *b, int opcode, basicblock *target, location loc) } static inline int -basicblock_append_instructions(basicblock *target, basicblock *source) +basicblock_append_instructions(basicblock *to, basicblock *from) { - for (int i = 0; i < source->b_iused; i++) { - int n = basicblock_next_instr(target); + for (int i = 0; i < from->b_iused; i++) { + int n = basicblock_next_instr(to); if (n < 0) { return ERROR; } - target->b_instr[n] = source->b_instr[i]; + to->b_instr[n] = from->b_instr[i]; } return SUCCESS; } @@ -292,9 +292,9 @@ static void dump_basicblock(const basicblock *b) { const char *b_return = basicblock_returns(b) ? "return " : ""; - fprintf(stderr, "%d: [EH=%d CLD=%d WRM=%d NO_FT=%d %p] used: %d, depth: %d, %s\n", + fprintf(stderr, "%d: [EH=%d CLD=%d WRM=%d NO_FT=%d %p] used: %d, depth: %d, preds: %d %s\n", b->b_label.id, b->b_except_handler, b->b_cold, b->b_warm, BB_NO_FALLTHROUGH(b), b, b->b_iused, - b->b_startdepth, b_return); + b->b_startdepth, b->b_predecessors, b_return); if (b->b_instr) { int i; for (i = 0; i < b->b_iused; i++) { @@ -1165,15 +1165,26 @@ remove_redundant_jumps(cfg_builder *g) { return changes; } +static inline bool +basicblock_has_no_lineno(basicblock *b) { + for (int i = 0; i < b->b_iused; i++) { + if (b->b_instr[i].i_loc.lineno >= 0) { + return false; + } + } + return true; +} + /* Maximum size of basic block that should be copied in optimizer */ #define MAX_COPY_SIZE 4 -/* If this block ends with an unconditional jump to a small exit block, then +/* If this block ends with an unconditional jump to a small exit block or + * a block that has no line numbers (and no fallthrough), then * remove the jump and extend this block with the target. * Returns 1 if extended, 0 if no change, and -1 on error. */ static int -inline_small_exit_blocks(basicblock *bb) { +basicblock_inline_small_or_no_lineno_blocks(basicblock *bb) { cfg_instr *last = basicblock_last_instr(bb); if (last == NULL) { return 0; @@ -1182,14 +1193,46 @@ inline_small_exit_blocks(basicblock *bb) { return 0; } basicblock *target = last->i_target; - if (basicblock_exits_scope(target) && target->b_iused <= MAX_COPY_SIZE) { + bool small_exit_block = (basicblock_exits_scope(target) && + target->b_iused <= MAX_COPY_SIZE); + bool no_lineno_no_fallthrough = (basicblock_has_no_lineno(target) && + !BB_HAS_FALLTHROUGH(target)); + if (small_exit_block || no_lineno_no_fallthrough) { + assert(is_jump(last)); + int removed_jump_opcode = last->i_opcode; INSTR_SET_OP0(last, NOP); RETURN_IF_ERROR(basicblock_append_instructions(bb, target)); + if (no_lineno_no_fallthrough) { + last = basicblock_last_instr(bb); + if (IS_UNCONDITIONAL_JUMP_OPCODE(last->i_opcode) && + removed_jump_opcode == JUMP) + { + /* Make sure we don't lose eval breaker checks */ + last->i_opcode = JUMP; + } + } + target->b_predecessors--; return 1; } return 0; } +static int +inline_small_or_no_lineno_blocks(basicblock *entryblock) { + bool changes; + do { + changes = false; + for (basicblock *b = entryblock; b != NULL; b = b->b_next) { + int res = basicblock_inline_small_or_no_lineno_blocks(b); + RETURN_IF_ERROR(res); + if (res) { + changes = true; + } + } + } while(changes); /* every change removes a jump, ensuring convergence */ + return changes; +} + // Attempt to eliminate jumps to jumps by updating inst to jump to // target->i_target using the provided opcode. Return whether or not the // optimization was successful. @@ -1804,9 +1847,7 @@ optimize_cfg(cfg_builder *g, PyObject *consts, PyObject *const_cache, int firstl { assert(PyDict_CheckExact(const_cache)); RETURN_IF_ERROR(check_cfg(g)); - for (basicblock *b = g->g_entryblock; b != NULL; b = b->b_next) { - RETURN_IF_ERROR(inline_small_exit_blocks(b)); - } + RETURN_IF_ERROR(inline_small_or_no_lineno_blocks(g->g_entryblock)); RETURN_IF_ERROR(remove_unreachable(g->g_entryblock)); RETURN_IF_ERROR(resolve_line_numbers(g, firstlineno)); RETURN_IF_ERROR(optimize_load_const(const_cache, g, consts)); @@ -1814,9 +1855,6 @@ optimize_cfg(cfg_builder *g, PyObject *consts, PyObject *const_cache, int firstl RETURN_IF_ERROR(optimize_basic_block(const_cache, b, consts)); } RETURN_IF_ERROR(remove_redundant_nops_and_pairs(g->g_entryblock)); - for (basicblock *b = g->g_entryblock; b != NULL; b = b->b_next) { - RETURN_IF_ERROR(inline_small_exit_blocks(b)); - } RETURN_IF_ERROR(remove_unreachable(g->g_entryblock)); int removed_nops, removed_jumps; @@ -2333,12 +2371,7 @@ convert_pseudo_ops(cfg_builder *g) static inline bool is_exit_or_eval_check_without_lineno(basicblock *b) { if (basicblock_exits_scope(b) || basicblock_has_eval_break(b)) { - for (int i = 0; i < b->b_iused; i++) { - if (b->b_instr[i].i_loc.lineno >= 0) { - return false; - } - } - return true; + return basicblock_has_no_lineno(b); } else { return false; diff --git a/Python/frame.c b/Python/frame.c index 2865b2eab603c2..ddf6ef6ba5465c 100644 --- a/Python/frame.c +++ b/Python/frame.c @@ -139,7 +139,6 @@ _PyFrame_ClearExceptCode(_PyInterpreterFrame *frame) for (int i = 0; i < frame->stacktop; i++) { Py_XDECREF(frame->localsplus[i]); } - Py_XDECREF(frame->frame_obj); Py_XDECREF(frame->f_locals); Py_DECREF(frame->f_funcobj); } diff --git a/Python/gc_free_threading.c b/Python/gc_free_threading.c index f2cd84981461a4..8fbcdb15109b76 100644 --- a/Python/gc_free_threading.c +++ b/Python/gc_free_threading.c @@ -25,7 +25,10 @@ typedef struct _gc_runtime_state GCState; // Automatically choose the generation that needs collecting. #define GENERATION_AUTO (-1) -// A linked-list of objects using the `ob_tid` field as the next pointer. +// A linked list of objects using the `ob_tid` field as the next pointer. +// The linked list pointers are distinct from any real thread ids, because the +// thread ids returned by _Py_ThreadId() are also pointers to distinct objects. +// No thread will confuse its own id with a linked list pointer. struct worklist { uintptr_t head; }; @@ -46,6 +49,7 @@ struct collection_state { GCState *gcstate; Py_ssize_t collected; Py_ssize_t uncollectable; + Py_ssize_t long_lived_total; struct worklist unreachable; struct worklist legacy_finalizers; struct worklist wrcb_to_call; @@ -220,7 +224,7 @@ gc_visit_heaps_lock_held(PyInterpreterState *interp, mi_block_visit_fun *visitor struct visitor_args *arg) { // Offset of PyObject header from start of memory block. - Py_ssize_t offset_base = sizeof(PyGC_Head); + Py_ssize_t offset_base = 0; if (_PyMem_DebugEnabled()) { // The debug allocator adds two words at the beginning of each block. offset_base += 2 * sizeof(size_t); @@ -330,8 +334,14 @@ update_refs(const mi_heap_t *heap, const mi_heap_area_t *area, Py_ssize_t refcount = Py_REFCNT(op); _PyObject_ASSERT(op, refcount >= 0); - // Add the actual refcount to ob_tid. + // We repurpose ob_tid to compute "gc_refs", the number of external + // references to the object (i.e., from outside the GC heaps). This means + // that ob_tid is no longer a valid thread id until it is restored by + // scan_heap_visitor(). Until then, we cannot use the standard reference + // counting functions or allow other threads to run Python code. gc_maybe_init_refs(op); + + // Add the actual refcount to ob_tid. gc_add_refs(op, refcount); // Subtract internal references from ob_tid. Objects with ob_tid > 0 @@ -443,7 +453,7 @@ scan_heap_visitor(const mi_heap_t *heap, const mi_heap_area_t *area, else { // object is reachable, restore `ob_tid`; we're done with these objects gc_restore_tid(op); - state->gcstate->long_lived_total++; + state->long_lived_total++; } return true; @@ -605,6 +615,8 @@ get_gc_state(void) void _PyGC_InitState(GCState *gcstate) { + // TODO: move to pycore_runtime_init.h once the incremental GC lands. + gcstate->generations[0].threshold = 2000; } @@ -885,62 +897,6 @@ invoke_gc_callback(PyThreadState *tstate, const char *phase, assert(!_PyErr_Occurred(tstate)); } - -/* Find the oldest generation (highest numbered) where the count - * exceeds the threshold. Objects in the that generation and - * generations younger than it will be collected. */ -static int -gc_select_generation(GCState *gcstate) -{ - for (int i = NUM_GENERATIONS-1; i >= 0; i--) { - if (gcstate->generations[i].count > gcstate->generations[i].threshold) { - /* Avoid quadratic performance degradation in number - of tracked objects (see also issue #4074): - - To limit the cost of garbage collection, there are two strategies; - - make each collection faster, e.g. by scanning fewer objects - - do less collections - This heuristic is about the latter strategy. - - In addition to the various configurable thresholds, we only trigger a - full collection if the ratio - - long_lived_pending / long_lived_total - - is above a given value (hardwired to 25%). - - The reason is that, while "non-full" collections (i.e., collections of - the young and middle generations) will always examine roughly the same - number of objects -- determined by the aforementioned thresholds --, - the cost of a full collection is proportional to the total number of - long-lived objects, which is virtually unbounded. - - Indeed, it has been remarked that doing a full collection every - of object creations entails a dramatic performance - degradation in workloads which consist in creating and storing lots of - long-lived objects (e.g. building a large list of GC-tracked objects would - show quadratic performance, instead of linear as expected: see issue #4074). - - Using the above ratio, instead, yields amortized linear performance in - the total number of objects (the effect of which can be summarized - thusly: "each full garbage collection is more and more costly as the - number of objects grows, but we do fewer and fewer of them"). - - This heuristic was suggested by Martin von Löwis on python-dev in - June 2008. His original analysis and proposal can be found at: - http://mail.python.org/pipermail/python-dev/2008-June/080579.html - */ - if (i == NUM_GENERATIONS - 1 - && gcstate->long_lived_pending < gcstate->long_lived_total / 4) - { - continue; - } - return i; - } - } - return -1; -} - static void cleanup_worklist(struct worklist *worklist) { @@ -952,6 +908,21 @@ cleanup_worklist(struct worklist *worklist) } } +static bool +gc_should_collect(GCState *gcstate) +{ + int count = _Py_atomic_load_int_relaxed(&gcstate->generations[0].count); + int threshold = gcstate->generations[0].threshold; + if (count <= threshold || threshold == 0 || !gcstate->enabled) { + return false; + } + // Avoid quadratic behavior by scaling threshold to the number of live + // objects. A few tests rely on immediate scheduling of the GC so we ignore + // the scaled threshold if generations[1].threshold is set to zero. + return (count > gcstate->long_lived_total / 4 || + gcstate->generations[1].threshold == 0); +} + static void gc_collect_internal(PyInterpreterState *interp, struct collection_state *state) { @@ -1029,15 +1000,10 @@ gc_collect_main(PyThreadState *tstate, int generation, _PyGC_Reason reason) return 0; } - if (generation == GENERATION_AUTO) { - // Select the oldest generation that needs collecting. We will collect - // objects from that generation and all generations younger than it. - generation = gc_select_generation(gcstate); - if (generation < 0) { - // No generation needs to be collected. - _Py_atomic_store_int(&gcstate->collecting, 0); - return 0; - } + if (reason == _Py_GC_REASON_HEAP && !gc_should_collect(gcstate)) { + // Don't collect if the threshold is not exceeded. + _Py_atomic_store_int(&gcstate->collecting, 0); + return 0; } assert(generation >= 0 && generation < NUM_GENERATIONS); @@ -1082,6 +1048,7 @@ gc_collect_main(PyThreadState *tstate, int generation, _PyGC_Reason reason) m = state.collected; n = state.uncollectable; + gcstate->long_lived_total = state.long_lived_total; if (gcstate->debug & _PyGC_DEBUG_STATS) { double d = _PyTime_AsSecondsDouble(_PyTime_GetPerfCounter() - t1); @@ -1523,12 +1490,10 @@ _PyObject_GC_Link(PyObject *op) { PyThreadState *tstate = _PyThreadState_GET(); GCState *gcstate = &tstate->interp->gc; - gcstate->generations[0].count++; /* number of allocated GC objects */ - if (gcstate->generations[0].count > gcstate->generations[0].threshold && - gcstate->enabled && - gcstate->generations[0].threshold && - !_Py_atomic_load_int_relaxed(&gcstate->collecting) && - !_PyErr_Occurred(tstate)) + gcstate->generations[0].count++; + + if (gc_should_collect(gcstate) && + !_Py_atomic_load_int_relaxed(&gcstate->collecting)) { _Py_ScheduleGC(tstate->interp); } @@ -1537,7 +1502,7 @@ _PyObject_GC_Link(PyObject *op) void _Py_RunGC(PyThreadState *tstate) { - gc_collect_main(tstate, GENERATION_AUTO, _Py_GC_REASON_HEAP); + gc_collect_main(tstate, 0, _Py_GC_REASON_HEAP); } static PyObject * @@ -1552,8 +1517,10 @@ gc_alloc(PyTypeObject *tp, size_t basicsize, size_t presize) if (mem == NULL) { return _PyErr_NoMemory(tstate); } - ((PyObject **)mem)[0] = NULL; - ((PyObject **)mem)[1] = NULL; + if (presize) { + ((PyObject **)mem)[0] = NULL; + ((PyObject **)mem)[1] = NULL; + } PyObject *op = (PyObject *)(mem + presize); _PyObject_GC_Link(op); return op; @@ -1709,8 +1676,6 @@ PyUnstable_GC_VisitObjects(gcvisitobjects_t callback, void *arg) void _PyGC_ClearAllFreeLists(PyInterpreterState *interp) { - _PyDict_ClearFreeList(interp); - HEAD_LOCK(&_PyRuntime); _PyThreadStateImpl *tstate = (_PyThreadStateImpl *)interp->threads.head; while (tstate != NULL) { diff --git a/Python/gc_gil.c b/Python/gc_gil.c index 04c1c184250c60..4e2aa8f7af746c 100644 --- a/Python/gc_gil.c +++ b/Python/gc_gil.c @@ -11,8 +11,6 @@ void _PyGC_ClearAllFreeLists(PyInterpreterState *interp) { - _PyDict_ClearFreeList(interp); - _Py_ClearFreeLists(&interp->freelist_state, 0); } diff --git a/Python/optimizer.c b/Python/optimizer.c index 0d04b09fef1e84..d71ca0aef0e11a 100644 --- a/Python/optimizer.c +++ b/Python/optimizer.c @@ -108,16 +108,14 @@ PyUnstable_Replace_Executor(PyCodeObject *code, _Py_CODEUNIT *instr, _PyExecutor } static int -error_optimize( +never_optimize( _PyOptimizerObject* self, - PyCodeObject *code, + _PyInterpreterFrame *frame, _Py_CODEUNIT *instr, _PyExecutorObject **exec, int Py_UNUSED(stack_entries)) { - assert(0); - PyErr_Format(PyExc_SystemError, "Should never call error_optimize"); - return -1; + return 0; } PyTypeObject _PyDefaultOptimizer_Type = { @@ -130,7 +128,7 @@ PyTypeObject _PyDefaultOptimizer_Type = { _PyOptimizerObject _PyOptimizer_Default = { PyObject_HEAD_INIT(&_PyDefaultOptimizer_Type) - .optimize = error_optimize, + .optimize = never_optimize, .resume_threshold = INT16_MAX, .backedge_threshold = INT16_MAX, }; @@ -174,7 +172,7 @@ _PyOptimizer_Optimize(_PyInterpreterFrame *frame, _Py_CODEUNIT *start, PyObject } _PyOptimizerObject *opt = interp->optimizer; _PyExecutorObject *executor = NULL; - int err = opt->optimize(opt, code, start, &executor, (int)(stack_pointer - _PyFrame_Stackbase(frame))); + int err = opt->optimize(opt, frame, start, &executor, (int)(stack_pointer - _PyFrame_Stackbase(frame))); if (err <= 0) { assert(executor == NULL); return err; @@ -363,7 +361,8 @@ BRANCH_TO_GUARD[4][2] = { ADD_TO_TRACE(_EXIT_TRACE, 0, 0, 0); \ goto done; \ } \ - trace_stack[trace_stack_depth].code = code; \ + assert(func->func_code == (PyObject *)code); \ + trace_stack[trace_stack_depth].func = func; \ trace_stack[trace_stack_depth].instr = instr; \ trace_stack_depth++; #define TRACE_STACK_POP() \ @@ -371,7 +370,8 @@ BRANCH_TO_GUARD[4][2] = { Py_FatalError("Trace stack underflow\n"); \ } \ trace_stack_depth--; \ - code = trace_stack[trace_stack_depth].code; \ + func = trace_stack[trace_stack_depth].func; \ + code = (PyCodeObject *)trace_stack[trace_stack_depth].func->func_code; \ instr = trace_stack[trace_stack_depth].instr; /* Returns 1 on success, @@ -380,20 +380,23 @@ BRANCH_TO_GUARD[4][2] = { */ static int translate_bytecode_to_trace( - PyCodeObject *code, + _PyInterpreterFrame *frame, _Py_CODEUNIT *instr, _PyUOpInstruction *trace, int buffer_size, _PyBloomFilter *dependencies) { bool progress_needed = true; + PyCodeObject *code = (PyCodeObject *)frame->f_executable; + PyFunctionObject *func = (PyFunctionObject *)frame->f_funcobj; + assert(PyFunction_Check(func)); PyCodeObject *initial_code = code; _Py_BloomFilter_Add(dependencies, initial_code); _Py_CODEUNIT *initial_instr = instr; int trace_length = 0; int max_length = buffer_size; struct { - PyCodeObject *code; + PyFunctionObject *func; _Py_CODEUNIT *instr; } trace_stack[TRACE_STACK_SIZE]; int trace_stack_depth = 0; @@ -593,9 +596,9 @@ translate_bytecode_to_trace( ADD_TO_TRACE(uop, oparg, operand, target); if (uop == _POP_FRAME) { TRACE_STACK_POP(); - /* Set the operand to the code object returned to, + /* Set the operand to the function object returned to, * to assist optimization passes */ - trace[trace_length-1].operand = (uintptr_t)code; + trace[trace_length-1].operand = (uintptr_t)func; DPRINTF(2, "Returning to %s (%s:%d) at byte offset %d\n", PyUnicode_AsUTF8(code->co_qualname), @@ -611,10 +614,10 @@ translate_bytecode_to_trace( // Add one to account for the actual opcode/oparg pair: + 1; uint32_t func_version = read_u32(&instr[func_version_offset].cache); - PyFunctionObject *func = _PyFunction_LookupByVersion(func_version); + PyFunctionObject *new_func = _PyFunction_LookupByVersion(func_version); DPRINTF(3, "Function object: %p\n", func); - if (func != NULL) { - PyCodeObject *new_code = (PyCodeObject *)PyFunction_GET_CODE(func); + if (new_func != NULL) { + PyCodeObject *new_code = (PyCodeObject *)PyFunction_GET_CODE(new_func); if (new_code == code) { // Recursive call, bail (we could be here forever). DPRINTF(2, "Bailing on recursive call to %s (%s:%d)\n", @@ -639,8 +642,9 @@ translate_bytecode_to_trace( _Py_BloomFilter_Add(dependencies, new_code); /* Set the operand to the callee's code object, * to assist optimization passes */ - trace[trace_length-1].operand = (uintptr_t)new_code; + trace[trace_length-1].operand = (uintptr_t)new_func; code = new_code; + func = new_func; instr = _PyCode_CODE(code); DPRINTF(2, "Continuing in %s (%s:%d) at byte offset %d\n", @@ -808,7 +812,7 @@ make_executor_from_uops(_PyUOpInstruction *buffer, _PyBloomFilter *dependencies) static int uop_optimize( _PyOptimizerObject *self, - PyCodeObject *code, + _PyInterpreterFrame *frame, _Py_CODEUNIT *instr, _PyExecutorObject **exec_ptr, int curr_stackentries) @@ -816,7 +820,7 @@ uop_optimize( _PyBloomFilter dependencies; _Py_BloomFilter_Init(&dependencies); _PyUOpInstruction buffer[UOP_MAX_TRACE_LENGTH]; - int err = translate_bytecode_to_trace(code, instr, buffer, UOP_MAX_TRACE_LENGTH, &dependencies); + int err = translate_bytecode_to_trace(frame, instr, buffer, UOP_MAX_TRACE_LENGTH, &dependencies); if (err <= 0) { // Error or nothing translated return err; @@ -824,9 +828,10 @@ uop_optimize( OPT_STAT_INC(traces_created); char *uop_optimize = Py_GETENV("PYTHONUOPSOPTIMIZE"); if (uop_optimize == NULL || *uop_optimize > '0') { - err = _Py_uop_analyze_and_optimize(code, buffer, UOP_MAX_TRACE_LENGTH, curr_stackentries); - if (err < 0) { - return -1; + err = _Py_uop_analyze_and_optimize(frame, buffer, + UOP_MAX_TRACE_LENGTH, curr_stackentries, &dependencies); + if (err <= 0) { + return err; } } _PyExecutorObject *executor = make_executor_from_uops(buffer, &dependencies); @@ -887,12 +892,13 @@ PyTypeObject _PyCounterExecutor_Type = { static int counter_optimize( _PyOptimizerObject* self, - PyCodeObject *code, + _PyInterpreterFrame *frame, _Py_CODEUNIT *instr, _PyExecutorObject **exec_ptr, int Py_UNUSED(curr_stackentries) ) { + PyCodeObject *code = (PyCodeObject *)frame->f_executable; int oparg = instr->op.arg; while (instr->op.code == EXTENDED_ARG) { instr++; diff --git a/Python/optimizer_analysis.c b/Python/optimizer_analysis.c index d1225997e10be2..2cfbf4b349d0f5 100644 --- a/Python/optimizer_analysis.c +++ b/Python/optimizer_analysis.c @@ -1,10 +1,12 @@ #include "Python.h" #include "opcode.h" +#include "pycore_dict.h" #include "pycore_interp.h" #include "pycore_opcode_metadata.h" #include "pycore_opcode_utils.h" #include "pycore_pystate.h" // _PyInterpreterState_GET() #include "pycore_uop_metadata.h" +#include "pycore_dict.h" #include "pycore_long.h" #include "cpython/optimizer.h" #include @@ -12,9 +14,210 @@ #include #include "pycore_optimizer.h" +static int +get_mutations(PyObject* dict) { + assert(PyDict_CheckExact(dict)); + PyDictObject *d = (PyDictObject *)dict; + return (d->ma_version_tag >> DICT_MAX_WATCHERS) & ((1 << DICT_WATCHED_MUTATION_BITS)-1); +} + static void -peephole_opt(PyCodeObject *co, _PyUOpInstruction *buffer, int buffer_size) +increment_mutations(PyObject* dict) { + assert(PyDict_CheckExact(dict)); + PyDictObject *d = (PyDictObject *)dict; + d->ma_version_tag += (1 << DICT_MAX_WATCHERS); +} + +static int +globals_watcher_callback(PyDict_WatchEvent event, PyObject* dict, + PyObject* key, PyObject* new_value) +{ + if (event == PyDict_EVENT_CLONED) { + return 0; + } + uint64_t watched_mutations = get_mutations(dict); + if (watched_mutations < _Py_MAX_ALLOWED_GLOBALS_MODIFICATIONS) { + _Py_Executors_InvalidateDependency(_PyInterpreterState_GET(), dict); + increment_mutations(dict); + } + else { + PyDict_Unwatch(1, dict); + } + return 0; +} + + +static void +global_to_const(_PyUOpInstruction *inst, PyObject *obj) +{ + assert(inst->opcode == _LOAD_GLOBAL_MODULE || inst->opcode == _LOAD_GLOBAL_BUILTINS); + assert(PyDict_CheckExact(obj)); + PyDictObject *dict = (PyDictObject *)obj; + assert(dict->ma_keys->dk_kind == DICT_KEYS_UNICODE); + PyDictUnicodeEntry *entries = DK_UNICODE_ENTRIES(dict->ma_keys); + assert(inst->operand <= UINT16_MAX); + PyObject *res = entries[inst->operand].me_value; + if (res == NULL) { + return; + } + if (_Py_IsImmortal(res)) { + inst->opcode = (inst->oparg & 1) ? _LOAD_CONST_INLINE_BORROW_WITH_NULL : _LOAD_CONST_INLINE_BORROW; + } + else { + inst->opcode = (inst->oparg & 1) ? _LOAD_CONST_INLINE_WITH_NULL : _LOAD_CONST_INLINE; + } + inst->operand = (uint64_t)res; +} + +static int +incorrect_keys(_PyUOpInstruction *inst, PyObject *obj) { + if (!PyDict_CheckExact(obj)) { + return 1; + } + PyDictObject *dict = (PyDictObject *)obj; + if (dict->ma_keys->dk_version != inst->operand) { + return 1; + } + return 0; +} + +/* The first two dict watcher IDs are reserved for CPython, + * so we don't need to check that they haven't been used */ +#define BUILTINS_WATCHER_ID 0 +#define GLOBALS_WATCHER_ID 1 + +/* Returns 1 if successfully optimized + * 0 if the trace is not suitable for optimization (yet) + * -1 if there was an error. */ +static int +remove_globals(_PyInterpreterFrame *frame, _PyUOpInstruction *buffer, + int buffer_size, _PyBloomFilter *dependencies) +{ + PyInterpreterState *interp = _PyInterpreterState_GET(); + PyObject *builtins = frame->f_builtins; + if (builtins != interp->builtins) { + return 1; + } + PyObject *globals = frame->f_globals; + assert(PyFunction_Check(((PyFunctionObject *)frame->f_funcobj))); + assert(((PyFunctionObject *)frame->f_funcobj)->func_builtins == builtins); + assert(((PyFunctionObject *)frame->f_funcobj)->func_globals == globals); + /* In order to treat globals as constants, we need to + * know that the globals dict is the one we expected, and + * that it hasn't changed + * In order to treat builtins as constants, we need to + * know that the builtins dict is the one we expected, and + * that it hasn't changed and that the global dictionary's + * keys have not changed */ + + /* These values represent stacks of booleans (one bool per bit). + * Pushing a frame shifts left, popping a frame shifts right. */ + uint32_t builtins_checked = 0; + uint32_t builtins_watched = 0; + uint32_t globals_checked = 0; + uint32_t globals_watched = 0; + if (interp->dict_state.watchers[1] == NULL) { + interp->dict_state.watchers[1] = globals_watcher_callback; + } + for (int pc = 0; pc < buffer_size; pc++) { + _PyUOpInstruction *inst = &buffer[pc]; + int opcode = inst->opcode; + switch(opcode) { + case _GUARD_BUILTINS_VERSION: + if (incorrect_keys(inst, builtins)) { + return 0; + } + if (interp->rare_events.builtin_dict >= _Py_MAX_ALLOWED_BUILTINS_MODIFICATIONS) { + continue; + } + if ((builtins_watched & 1) == 0) { + PyDict_Watch(BUILTINS_WATCHER_ID, builtins); + builtins_watched |= 1; + } + if (builtins_checked & 1) { + buffer[pc].opcode = NOP; + } + else { + buffer[pc].opcode = _CHECK_BUILTINS; + buffer[pc].operand = (uintptr_t)builtins; + builtins_checked |= 1; + } + break; + case _GUARD_GLOBALS_VERSION: + if (incorrect_keys(inst, globals)) { + return 0; + } + uint64_t watched_mutations = get_mutations(globals); + if (watched_mutations >= _Py_MAX_ALLOWED_GLOBALS_MODIFICATIONS) { + continue; + } + if ((globals_watched & 1) == 0) { + PyDict_Watch(GLOBALS_WATCHER_ID, globals); + _Py_BloomFilter_Add(dependencies, globals); + globals_watched |= 1; + } + if (globals_checked & 1) { + buffer[pc].opcode = NOP; + } + else { + buffer[pc].opcode = _CHECK_GLOBALS; + buffer[pc].operand = (uintptr_t)globals; + globals_checked |= 1; + } + break; + case _LOAD_GLOBAL_BUILTINS: + if (globals_checked & builtins_checked & globals_watched & builtins_watched & 1) { + global_to_const(inst, builtins); + } + break; + case _LOAD_GLOBAL_MODULE: + if (globals_checked & globals_watched & 1) { + global_to_const(inst, globals); + } + break; + case _PUSH_FRAME: + { + globals_checked <<= 1; + globals_watched <<= 1; + builtins_checked <<= 1; + builtins_watched <<= 1; + PyFunctionObject *func = (PyFunctionObject *)buffer[pc].operand; + if (func == NULL) { + return 1; + } + assert(PyFunction_Check(func)); + globals = func->func_globals; + builtins = func->func_builtins; + if (builtins != interp->builtins) { + return 1; + } + break; + } + case _POP_FRAME: + { + globals_checked >>= 1; + globals_watched >>= 1; + builtins_checked >>= 1; + builtins_watched >>= 1; + PyFunctionObject *func = (PyFunctionObject *)buffer[pc].operand; + assert(PyFunction_Check(func)); + globals = func->func_globals; + builtins = func->func_builtins; + break; + } + case _JUMP_TO_TOP: + case _EXIT_TRACE: + return 1; + } + } + return 0; +} + +static void +peephole_opt(_PyInterpreterFrame *frame, _PyUOpInstruction *buffer, int buffer_size) +{ + PyCodeObject *co = (PyCodeObject *)frame->f_executable; for (int pc = 0; pc < buffer_size; pc++) { int opcode = buffer[pc].opcode; switch(opcode) { @@ -36,8 +239,17 @@ peephole_opt(PyCodeObject *co, _PyUOpInstruction *buffer, int buffer_size) } case _PUSH_FRAME: case _POP_FRAME: - co = (PyCodeObject *)buffer[pc].operand; + { + PyFunctionObject *func = (PyFunctionObject *)buffer[pc].operand; + if (func == NULL) { + co = NULL; + } + else { + assert(PyFunction_Check(func)); + co = (PyCodeObject *)func->func_code; + } break; + } case _JUMP_TO_TOP: case _EXIT_TRACE: return; @@ -83,16 +295,20 @@ remove_unneeded_uops(_PyUOpInstruction *buffer, int buffer_size) } } - int _Py_uop_analyze_and_optimize( - PyCodeObject *co, + _PyInterpreterFrame *frame, _PyUOpInstruction *buffer, int buffer_size, - int curr_stacklen + int curr_stacklen, + _PyBloomFilter *dependencies ) { - peephole_opt(co, buffer, buffer_size); + int err = remove_globals(frame, buffer, buffer_size, dependencies); + if (err <= 0) { + return err; + } + peephole_opt(frame, buffer, buffer_size); remove_unneeded_uops(buffer, buffer_size); - return 0; + return 1; } diff --git a/Python/pylifecycle.c b/Python/pylifecycle.c index 372f60602375b6..0cac7109340129 100644 --- a/Python/pylifecycle.c +++ b/Python/pylifecycle.c @@ -32,6 +32,7 @@ #include "pycore_typevarobject.h" // _Py_clear_generic_types() #include "pycore_unicodeobject.h" // _PyUnicode_InitTypes() #include "pycore_weakref.h" // _PyWeakref_GET_REF() +#include "cpython/optimizer.h" // _Py_MAX_ALLOWED_BUILTINS_MODIFICATIONS #include "pycore_obmalloc.h" // _PyMem_init_obmalloc() #include "opcode.h" @@ -609,7 +610,11 @@ init_interp_create_gil(PyThreadState *tstate, int gil) static int builtins_dict_watcher(PyDict_WatchEvent event, PyObject *dict, PyObject *key, PyObject *new_value) { - RARE_EVENT_INC(builtin_dict); + PyInterpreterState *interp = _PyInterpreterState_GET(); + if (event != PyDict_EVENT_CLONED && interp->rare_events.builtin_dict < _Py_MAX_ALLOWED_BUILTINS_MODIFICATIONS) { + _Py_Executors_InvalidateAll(interp); + } + RARE_EVENT_INTERP_INC(interp, builtin_dict); return 0; } @@ -1287,11 +1292,9 @@ init_interp_main(PyThreadState *tstate) } } - if ((interp->rare_events.builtins_dict_watcher_id = PyDict_AddWatcher(&builtins_dict_watcher)) == -1) { - return _PyStatus_ERR("failed to add builtin dict watcher"); - } - if (PyDict_Watch(interp->rare_events.builtins_dict_watcher_id, interp->builtins) != 0) { + interp->dict_state.watchers[0] = &builtins_dict_watcher; + if (PyDict_Watch(0, interp->builtins) != 0) { return _PyStatus_ERR("failed to set builtin dict watcher"); } @@ -1622,8 +1625,13 @@ finalize_modules(PyThreadState *tstate) { PyInterpreterState *interp = tstate->interp; - // Stop collecting stats on __builtin__ modifications during teardown - PyDict_Unwatch(interp->rare_events.builtins_dict_watcher_id, interp->builtins); + // Invalidate all executors and turn off tier 2 optimizer + _Py_Executors_InvalidateAll(interp); + Py_XDECREF(interp->optimizer); + interp->optimizer = &_PyOptimizer_Default; + + // Stop watching __builtin__ modifications + PyDict_Unwatch(0, interp->builtins); PyObject *modules = _PyImport_GetModules(interp); if (modules == NULL) { diff --git a/Python/pystate.c b/Python/pystate.c index 430121a6a35d7f..7836c172bbfb61 100644 --- a/Python/pystate.c +++ b/Python/pystate.c @@ -1461,9 +1461,12 @@ clear_datastack(PyThreadState *tstate) void _Py_ClearFreeLists(_PyFreeListState *state, int is_finalization) { + // In the free-threaded build, freelists are per-PyThreadState and cleared in PyThreadState_Clear() + // In the default build, freelists are per-interpreter and cleared in finalize_interp_types() _PyFloat_ClearFreeList(state, is_finalization); _PyTuple_ClearFreeList(state, is_finalization); _PyList_ClearFreeList(state, is_finalization); + _PyDict_ClearFreeList(state, is_finalization); _PyContext_ClearFreeList(state, is_finalization); _PyAsyncGen_ClearFreeLists(state, is_finalization); _PyObjectStackChunk_ClearFreeList(state, is_finalization); @@ -2485,7 +2488,17 @@ PyGILState_Check(void) return 0; } - return (tstate == gilstate_tss_get(runtime)); +#ifdef MS_WINDOWS + int err = GetLastError(); +#endif + + PyThreadState *tcur = gilstate_tss_get(runtime); + +#ifdef MS_WINDOWS + SetLastError(err); +#endif + + return (tstate == tcur); } PyGILState_STATE diff --git a/Python/specialize.c b/Python/specialize.c index a9efbe0453b94e..e38e3556a6d642 100644 --- a/Python/specialize.c +++ b/Python/specialize.c @@ -540,6 +540,7 @@ _PyCode_Quicken(PyCodeObject *code) #define SPEC_FAIL_CALL_METHOD_WRAPPER 28 #define SPEC_FAIL_CALL_OPERATOR_WRAPPER 29 #define SPEC_FAIL_CALL_INIT_NOT_SIMPLE 30 +#define SPEC_FAIL_CALL_METACLASS 31 /* COMPARE_OP */ #define SPEC_FAIL_COMPARE_OP_DIFFERENT_TYPES 12 @@ -1757,6 +1758,10 @@ specialize_class_call(PyObject *callable, _Py_CODEUNIT *instr, int nargs) SPEC_FAIL_CALL_STR : SPEC_FAIL_CALL_CLASS_NO_VECTORCALL); return -1; } + if (Py_TYPE(tp) != &PyType_Type) { + SPECIALIZATION_FAIL(CALL, SPEC_FAIL_CALL_METACLASS); + return -1; + } if (tp->tp_new == PyBaseObject_Type.tp_new) { PyFunctionObject *init = get_init_for_simple_managed_python_class(tp); if (type_get_version(tp, CALL) == 0) { diff --git a/Python/structmember.c b/Python/structmember.c index 7a5a6a49d23116..18bd486952419b 100644 --- a/Python/structmember.c +++ b/Python/structmember.c @@ -197,45 +197,72 @@ PyMember_SetOne(char *addr, PyMemberDef *l, PyObject *v) WARN("Truncation of value to int"); break; } - case Py_T_UINT:{ - unsigned long ulong_val = PyLong_AsUnsignedLong(v); - if ((ulong_val == (unsigned long)-1) && PyErr_Occurred()) { - /* XXX: For compatibility, accept negative int values - as well. */ - PyErr_Clear(); - ulong_val = PyLong_AsLong(v); - if ((ulong_val == (unsigned long)-1) && - PyErr_Occurred()) + case Py_T_UINT: { + /* XXX: For compatibility, accept negative int values + as well. */ + int overflow; + long long_val = PyLong_AsLongAndOverflow(v, &overflow); + if (long_val == -1 && PyErr_Occurred()) { + return -1; + } + if (overflow < 0) { + PyErr_SetString(PyExc_OverflowError, + "Python int too large to convert to C long"); + } + else if (!overflow) { + *(unsigned int *)addr = (unsigned int)(unsigned long)long_val; + if (long_val < 0) { + WARN("Writing negative value into unsigned field"); + } + else if ((unsigned long)long_val > UINT_MAX) { + WARN("Truncation of value to unsigned short"); + } + } + else { + unsigned long ulong_val = PyLong_AsUnsignedLong(v); + if (ulong_val == (unsigned long)-1 && PyErr_Occurred()) { return -1; - *(unsigned int *)addr = (unsigned int)ulong_val; - WARN("Writing negative value into unsigned field"); - } else - *(unsigned int *)addr = (unsigned int)ulong_val; - if (ulong_val > UINT_MAX) - WARN("Truncation of value to unsigned int"); - break; + } + *(unsigned int*)addr = (unsigned int)ulong_val; + if (ulong_val > UINT_MAX) { + WARN("Truncation of value to unsigned int"); + } } + break; + } case Py_T_LONG:{ *(long*)addr = PyLong_AsLong(v); if ((*(long*)addr == -1) && PyErr_Occurred()) return -1; break; } - case Py_T_ULONG:{ - *(unsigned long*)addr = PyLong_AsUnsignedLong(v); - if ((*(unsigned long*)addr == (unsigned long)-1) - && PyErr_Occurred()) { - /* XXX: For compatibility, accept negative int values - as well. */ - PyErr_Clear(); - *(unsigned long*)addr = PyLong_AsLong(v); - if ((*(unsigned long*)addr == (unsigned long)-1) - && PyErr_Occurred()) + case Py_T_ULONG: { + /* XXX: For compatibility, accept negative int values + as well. */ + int overflow; + long long_val = PyLong_AsLongAndOverflow(v, &overflow); + if (long_val == -1 && PyErr_Occurred()) { + return -1; + } + if (overflow < 0) { + PyErr_SetString(PyExc_OverflowError, + "Python int too large to convert to C long"); + } + else if (!overflow) { + *(unsigned long *)addr = (unsigned long)long_val; + if (long_val < 0) { + WARN("Writing negative value into unsigned field"); + } + } + else { + unsigned long ulong_val = PyLong_AsUnsignedLong(v); + if (ulong_val == (unsigned long)-1 && PyErr_Occurred()) { return -1; - WARN("Writing negative value into unsigned field"); + } + *(unsigned long*)addr = ulong_val; } break; - } + } case Py_T_PYSSIZET:{ *(Py_ssize_t*)addr = PyLong_AsSsize_t(v); if ((*(Py_ssize_t*)addr == (Py_ssize_t)-1) diff --git a/Python/sysmodule.c b/Python/sysmodule.c index f558a00a6916eb..437d7f8dfc4958 100644 --- a/Python/sysmodule.c +++ b/Python/sysmodule.c @@ -1878,7 +1878,15 @@ _PySys_GetSizeOf(PyObject *o) return (size_t)-1; } - return (size_t)size + _PyType_PreHeaderSize(Py_TYPE(o)); + size_t presize = 0; + if (!Py_IS_TYPE(o, &PyType_Type) || + PyType_HasFeature((PyTypeObject *)o, Py_TPFLAGS_HEAPTYPE)) + { + /* Add the size of the pre-header if "o" is not a static type */ + presize = _PyType_PreHeaderSize(Py_TYPE(o)); + } + + return (size_t)size + presize; } static PyObject * diff --git a/Python/thread_nt.h b/Python/thread_nt.h index 14b9cddc24c0ec..044e9fa111e979 100644 --- a/Python/thread_nt.h +++ b/Python/thread_nt.h @@ -444,16 +444,7 @@ PyThread_set_key_value(int key, void *value) void * PyThread_get_key_value(int key) { - /* because TLS is used in the Py_END_ALLOW_THREAD macro, - * it is necessary to preserve the windows error state, because - * it is assumed to be preserved across the call to the macro. - * Ideally, the macro should be fixed, but it is simpler to - * do it here. - */ - DWORD error = GetLastError(); - void *result = TlsGetValue(key); - SetLastError(error); - return result; + return TlsGetValue(key); } void @@ -525,14 +516,5 @@ void * PyThread_tss_get(Py_tss_t *key) { assert(key != NULL); - /* because TSS is used in the Py_END_ALLOW_THREAD macro, - * it is necessary to preserve the windows error state, because - * it is assumed to be preserved across the call to the macro. - * Ideally, the macro should be fixed, but it is simpler to - * do it here. - */ - DWORD error = GetLastError(); - void *result = TlsGetValue(key->_key); - SetLastError(error); - return result; + return TlsGetValue(key->_key); } diff --git a/Tools/cases_generator/README.md b/Tools/cases_generator/README.md index ed802e44f31ad5..7fec8a882336cd 100644 --- a/Tools/cases_generator/README.md +++ b/Tools/cases_generator/README.md @@ -5,16 +5,30 @@ Documentation for the instruction definitions in `Python/bytecodes.c` What's currently here: +- `analyzer.py`: code for converting `AST` generated by `Parser` + to more high-level structure for easier interaction - `lexer.py`: lexer for C, originally written by Mark Shannon - `plexer.py`: OO interface on top of lexer.py; main class: `PLexer` -- `parsing.py`: Parser for instruction definition DSL; main class `Parser` -- `generate_cases.py`: driver script to read `Python/bytecodes.c` and +- `parsing.py`: Parser for instruction definition DSL; main class: `Parser` +- `parser.py` helper for interactions with `parsing.py` +- `tierN_generator.py`: a couple of driver scripts to read `Python/bytecodes.c` and write `Python/generated_cases.c.h` (and several other files) -- `analysis.py`: `Analyzer` class used to read the input files -- `flags.py`: abstractions related to metadata flags for instructions -- `formatting.py`: `Formatter` class used to write the output files -- `instructions.py`: classes to analyze and write instructions -- `stacking.py`: code to handle generalized stack effects +- `stack.py`: code to handle generalized stack effects +- `cwriter.py`: code which understands tokens and how to format C code; + main class: `CWriter` +- `generators_common.py`: helpers for generators +- `opcode_id_generator.py`: generate a list of opcodes and write them to + `Include/opcode_ids.h` +- `opcode_metadata_generator.py`: reads the instruction definitions and + write the metadata to `Include/internal/pycore_opcode_metadata.h` +- `py_metadata_generator.py`: reads the instruction definitions and + write the metadata to `Lib/_opcode_metadata.py` +- `target_generator.py`: generate targets for computed goto dispatch and + write them to `Python/opcode_targets.h` +- `uop_id_generator.py`: generate a list of uop IDs and write them to + `Include/internal/pycore_uop_ids.h` +- `uop_metadata_generator.py`: reads the instruction definitions and + write the metadata to `Include/internal/pycore_uop_metadata.h` Note that there is some dummy C code at the top and bottom of `Python/bytecodes.c` diff --git a/Tools/cases_generator/opcode_id_generator.py b/Tools/cases_generator/opcode_id_generator.py index dbea3d0b622c87..5a3009a5c04c27 100644 --- a/Tools/cases_generator/opcode_id_generator.py +++ b/Tools/cases_generator/opcode_id_generator.py @@ -1,6 +1,6 @@ """Generate the list of opcode IDs. Reads the instruction definitions from bytecodes.c. -Writes the IDs to opcode._ids.h by default. +Writes the IDs to opcode_ids.h by default. """ import argparse diff --git a/Tools/cases_generator/opcode_metadata_generator.py b/Tools/cases_generator/opcode_metadata_generator.py index 1826a0b645c3b8..3e9fa3e26daa53 100644 --- a/Tools/cases_generator/opcode_metadata_generator.py +++ b/Tools/cases_generator/opcode_metadata_generator.py @@ -1,6 +1,6 @@ -"""Generate uop metedata. +"""Generate opcode metadata. Reads the instruction definitions from bytecodes.c. -Writes the metadata to pycore_uop_metadata.h by default. +Writes the metadata to pycore_opcode_metadata.h by default. """ import argparse diff --git a/Tools/cases_generator/py_metadata_generator.py b/Tools/cases_generator/py_metadata_generator.py index 43811fdacc8a9e..0dbcd599f9d4d9 100644 --- a/Tools/cases_generator/py_metadata_generator.py +++ b/Tools/cases_generator/py_metadata_generator.py @@ -1,6 +1,6 @@ -"""Generate uop metedata. +"""Generate opcode metadata for Python. Reads the instruction definitions from bytecodes.c. -Writes the metadata to pycore_uop_metadata.h by default. +Writes the metadata to _opcode_metadata.py by default. """ import argparse diff --git a/Tools/cases_generator/uop_metadata_generator.py b/Tools/cases_generator/uop_metadata_generator.py index d4f3a096d2acc1..9083ecc48bdf5b 100644 --- a/Tools/cases_generator/uop_metadata_generator.py +++ b/Tools/cases_generator/uop_metadata_generator.py @@ -1,4 +1,4 @@ -"""Generate uop metedata. +"""Generate uop metadata. Reads the instruction definitions from bytecodes.c. Writes the metadata to pycore_uop_metadata.h by default. """ diff --git a/Tools/clinic/.ruff.toml b/Tools/clinic/.ruff.toml index cbb3a9a8f3a8c2..c019572d0cb186 100644 --- a/Tools/clinic/.ruff.toml +++ b/Tools/clinic/.ruff.toml @@ -1,5 +1,7 @@ target-version = "py310" fix = true + +[lint] select = [ "F", # Enable all pyflakes rules "UP", # Enable all pyupgrade rules by default diff --git a/Tools/gdb/libpython.py b/Tools/gdb/libpython.py index 5ef55524c11be2..483f28b46dfec7 100755 --- a/Tools/gdb/libpython.py +++ b/Tools/gdb/libpython.py @@ -70,6 +70,14 @@ def _type_unsigned_int_ptr(): def _sizeof_void_p(): return gdb.lookup_type('void').pointer().sizeof +def _managed_dict_offset(): + # See pycore_object.h + pyobj = gdb.lookup_type("PyObject") + if any(field.name == "ob_ref_local" for field in pyobj.fields()): + return -1 * _sizeof_void_p() + else: + return -3 * _sizeof_void_p() + Py_TPFLAGS_MANAGED_DICT = (1 << 4) Py_TPFLAGS_HEAPTYPE = (1 << 9) @@ -457,7 +465,7 @@ def get_attr_dict(self): if dictoffset < 0: if int_from_int(typeobj.field('tp_flags')) & Py_TPFLAGS_MANAGED_DICT: assert dictoffset == -1 - dictoffset = -3 * _sizeof_void_p() + dictoffset = _managed_dict_offset() else: type_PyVarObject_ptr = gdb.lookup_type('PyVarObject').pointer() tsize = int_from_int(self._gdbval.cast(type_PyVarObject_ptr)['ob_size']) @@ -485,9 +493,8 @@ def get_keys_values(self): has_values = int_from_int(typeobj.field('tp_flags')) & Py_TPFLAGS_MANAGED_DICT if not has_values: return None - charptrptr_t = _type_char_ptr().pointer() - ptr = self._gdbval.cast(charptrptr_t) - 3 - char_ptr = ptr.dereference() + ptr = self._gdbval.cast(_type_char_ptr()) + _managed_dict_offset() + char_ptr = ptr.cast(_type_char_ptr().pointer()).dereference() if (int(char_ptr) & 1) == 0: return None char_ptr += 1 diff --git a/Tools/requirements-dev.txt b/Tools/requirements-dev.txt index b89f86a35d6115..c0a63b40ff4155 100644 --- a/Tools/requirements-dev.txt +++ b/Tools/requirements-dev.txt @@ -3,5 +3,5 @@ mypy==1.8.0 # needed for peg_generator: -types-psutil==5.9.5.17 -types-setuptools==69.0.0.0 +types-psutil==5.9.5.20240106 +types-setuptools==69.0.0.20240125 diff --git a/Tools/requirements-hypothesis.txt b/Tools/requirements-hypothesis.txt index 0e6e16ae198162..064731a236ee86 100644 --- a/Tools/requirements-hypothesis.txt +++ b/Tools/requirements-hypothesis.txt @@ -1,4 +1,4 @@ # Requirements file for hypothesis that # we use to run our property-based tests in CI. -hypothesis==6.92.2 +hypothesis==6.97.4 diff --git a/Tools/wasm/README.md b/Tools/wasm/README.md index beb857f69e40da..23b38c8e93638a 100644 --- a/Tools/wasm/README.md +++ b/Tools/wasm/README.md @@ -83,7 +83,7 @@ embuilder --pic build zlib bzip2 MINIMAL_PIC ``` -#### Compile a build Python interpreter +### Compile and build Python interpreter From within the container, run the following command: diff --git a/configure b/configure index 7b2119ff7f4f78..0375565c294552 100755 --- a/configure +++ b/configure @@ -920,6 +920,7 @@ LLVM_AR PROFILE_TASK DEF_MAKE_RULE DEF_MAKE_ALL_RULE +JIT_STENCILS_H REGEN_JIT_COMMAND ABIFLAGS LN @@ -8019,12 +8020,14 @@ then : else $as_nop as_fn_append CFLAGS_NODIST " -D_Py_JIT" REGEN_JIT_COMMAND="\$(PYTHON_FOR_REGEN) \$(srcdir)/Tools/jit/build.py $host" + JIT_STENCILS_H="jit_stencils.h" if test "x$Py_DEBUG" = xtrue then : as_fn_append REGEN_JIT_COMMAND " --debug" fi fi + { printf "%s\n" "$as_me:${as_lineno-$LINENO}: result: $enable_experimental_jit" >&5 printf "%s\n" "$enable_experimental_jit" >&6; } @@ -17475,6 +17478,12 @@ if test "x$ac_cv_func_getgid" = xyes then : printf "%s\n" "#define HAVE_GETGID 1" >>confdefs.h +fi +ac_fn_c_check_func "$LINENO" "getgrent" "ac_cv_func_getgrent" +if test "x$ac_cv_func_getgrent" = xyes +then : + printf "%s\n" "#define HAVE_GETGRENT 1" >>confdefs.h + fi ac_fn_c_check_func "$LINENO" "getgrgid" "ac_cv_func_getgrgid" if test "x$ac_cv_func_getgrgid" = xyes @@ -28968,7 +28977,8 @@ then : if true then : - if test "$ac_cv_func_getgrgid" = yes -o "$ac_cv_func_getgrgid_r" = yes + if test "$ac_cv_func_getgrent" = "yes" && + { test "$ac_cv_func_getgrgid" = "yes" || test "$ac_cv_func_getgrgid_r" = "yes"; } then : py_cv_module_grp=yes else $as_nop diff --git a/configure.ac b/configure.ac index 5bef2351c987e8..e121e893a1d0d9 100644 --- a/configure.ac +++ b/configure.ac @@ -1592,11 +1592,13 @@ AS_VAR_IF([enable_experimental_jit], [AS_VAR_APPEND([CFLAGS_NODIST], [" -D_Py_JIT"]) AS_VAR_SET([REGEN_JIT_COMMAND], ["\$(PYTHON_FOR_REGEN) \$(srcdir)/Tools/jit/build.py $host"]) + AS_VAR_SET([JIT_STENCILS_H], ["jit_stencils.h"]) AS_VAR_IF([Py_DEBUG], [true], [AS_VAR_APPEND([REGEN_JIT_COMMAND], [" --debug"])], [])]) AC_SUBST([REGEN_JIT_COMMAND]) +AC_SUBST([JIT_STENCILS_H]) AC_MSG_RESULT([$enable_experimental_jit]) # Enable optimization flags @@ -4787,7 +4789,7 @@ AC_CHECK_FUNCS([ \ copy_file_range ctermid dup dup3 execv explicit_bzero explicit_memset \ faccessat fchmod fchmodat fchown fchownat fdopendir fdwalk fexecve \ fork fork1 fpathconf fstatat ftime ftruncate futimens futimes futimesat \ - gai_strerror getegid getentropy geteuid getgid getgrgid getgrgid_r \ + gai_strerror getegid getentropy geteuid getgid getgrent getgrgid getgrgid_r \ getgrnam_r getgrouplist getgroups gethostname getitimer getloadavg getlogin \ getpeername getpgid getpid getppid getpriority _getpty \ getpwent getpwnam_r getpwuid getpwuid_r getresgid getresuid getrusage getsid getspent \ @@ -7313,7 +7315,9 @@ PY_STDLIB_MOD([_socket], -a "$ac_cv_header_netinet_in_h" = "yes"])) dnl platform specific extensions -PY_STDLIB_MOD([grp], [], [test "$ac_cv_func_getgrgid" = yes -o "$ac_cv_func_getgrgid_r" = yes]) +PY_STDLIB_MOD([grp], [], + [test "$ac_cv_func_getgrent" = "yes" && + { test "$ac_cv_func_getgrgid" = "yes" || test "$ac_cv_func_getgrgid_r" = "yes"; }]) PY_STDLIB_MOD([pwd], [], [test "$ac_cv_func_getpwuid" = yes -o "$ac_cv_func_getpwuid_r" = yes]) PY_STDLIB_MOD([resource], [], [test "$ac_cv_header_sys_resource_h" = yes]) PY_STDLIB_MOD([_scproxy], diff --git a/pyconfig.h.in b/pyconfig.h.in index b22740710bcbee..2b4bb1a2b52866 100644 --- a/pyconfig.h.in +++ b/pyconfig.h.in @@ -474,6 +474,9 @@ /* Define to 1 if you have the `getgid' function. */ #undef HAVE_GETGID +/* Define to 1 if you have the `getgrent' function. */ +#undef HAVE_GETGRENT + /* Define to 1 if you have the `getgrgid' function. */ #undef HAVE_GETGRGID