Mailman 3 January 2025 - Python-checkins

[3.12] gh-127614: Correctly check for ttyname_r() in configure (#128503) (#128599)
by erlend-aasland 07 Jan '25

07 Jan '25

https://github.com/python/cpython/commit/6b3371914c7448373e7dc4bb28020173b1… commit: 6b3371914c7448373e7dc4bb28020173b10dbe02 branch: 3.12 author: Erlend E. Aasland <erlend(a)python.org> committer: erlend-aasland <erlend.aasland(a)protonmail.com> date: 2025-01-07T22:05:42Z summary: [3.12] gh-127614: Correctly check for ttyname_r() in configure (#128503) (#128599) (cherry picked from commit e08b28235a863323ca3a7e444776bb7803e77caf) PR #14868 replaced the ttyname() call with ttyname_r(), but the old check remained. files: M Modules/clinic/posixmodule.c.h M Modules/posixmodule.c M configure M configure.ac M pyconfig.h.in diff --git a/Modules/clinic/posixmodule.c.h b/Modules/clinic/posixmodule.c.h index a33461dc5600dd..57121e1e8903a5 100644 --- a/Modules/clinic/posixmodule.c.h +++ b/Modules/clinic/posixmodule.c.h @@ -304,7 +304,7 @@ os_access(PyObject *module, PyObject *const *args, Py_ssize_t nargs, PyObject *k return return_value; } -#if defined(HAVE_TTYNAME) +#if defined(HAVE_TTYNAME_R) PyDoc_STRVAR(os_ttyname__doc__, "ttyname($module, fd, /)\n" @@ -337,7 +337,7 @@ os_ttyname(PyObject *module, PyObject *arg) return return_value; } -#endif /* defined(HAVE_TTYNAME) */ +#endif /* defined(HAVE_TTYNAME_R) */ #if defined(HAVE_CTERMID) @@ -12032,4 +12032,4 @@ os_waitstatus_to_exitcode(PyObject *module, PyObject *const *args, Py_ssize_t na #ifndef OS_WAITSTATUS_TO_EXITCODE_METHODDEF #define OS_WAITSTATUS_TO_EXITCODE_METHODDEF #endif /* !defined(OS_WAITSTATUS_TO_EXITCODE_METHODDEF) */ -/*[clinic end generated code: output=6d34c4564aca7725 input=a9049054013a1b77]*/ +/*[clinic end generated code: output=efe45f93561db8d3 input=a9049054013a1b77]*/ diff --git a/Modules/posixmodule.c b/Modules/posixmodule.c index 82078a569881a0..deabc103056642 100644 --- a/Modules/posixmodule.c +++ b/Modules/posixmodule.c @@ -3237,7 +3237,7 @@ os_access_impl(PyObject *module, path_t *path, int mode, int dir_fd, #endif -#ifdef HAVE_TTYNAME +#ifdef HAVE_TTYNAME_R /*[clinic input] os.ttyname diff --git a/configure b/configure index 32c9600b2e8cac..89edc42f45c361 100755 --- a/configure +++ b/configure @@ -18553,10 +18553,10 @@ then : printf "%s\n" "#define HAVE_TRUNCATE 1" >>confdefs.h fi -ac_fn_c_check_func "$LINENO" "ttyname" "ac_cv_func_ttyname" -if test "x$ac_cv_func_ttyname" = xyes +ac_fn_c_check_func "$LINENO" "ttyname_r" "ac_cv_func_ttyname_r" +if test "x$ac_cv_func_ttyname_r" = xyes then : - printf "%s\n" "#define HAVE_TTYNAME 1" >>confdefs.h + printf "%s\n" "#define HAVE_TTYNAME_R 1" >>confdefs.h fi ac_fn_c_check_func "$LINENO" "umask" "ac_cv_func_umask" diff --git a/configure.ac b/configure.ac index 0f60c849965719..1a02d19f1b2a3f 100644 --- a/configure.ac +++ b/configure.ac @@ -4941,7 +4941,7 @@ AC_CHECK_FUNCS([ \ sigfillset siginterrupt sigpending sigrelse sigtimedwait sigwait \ sigwaitinfo snprintf splice strftime strlcpy strsignal symlinkat sync \ sysconf system tcgetpgrp tcsetpgrp tempnam timegm times tmpfile \ - tmpnam tmpnam_r truncate ttyname umask uname unlinkat utimensat utimes vfork \ + tmpnam tmpnam_r truncate ttyname_r umask uname unlinkat utimensat utimes vfork \ wait wait3 wait4 waitid waitpid wcscoll wcsftime wcsxfrm wmemcmp writev \ ]) diff --git a/pyconfig.h.in b/pyconfig.h.in index 167c6a11411aa6..df4d29fe5497a4 100644 --- a/pyconfig.h.in +++ b/pyconfig.h.in @@ -1435,8 +1435,8 @@ /* Define to 1 if you have the `truncate' function. */ #undef HAVE_TRUNCATE -/* Define to 1 if you have the `ttyname' function. */ -#undef HAVE_TTYNAME +/* Define to 1 if you have the `ttyname_r' function. */ +#undef HAVE_TTYNAME_R /* Define to 1 if you don't have `tm_zone' but do have the external array `tzname'. */

1 0

gh-127614: Correctly check for ttyname_r() in configure (#128503)
by erlend-aasland 07 Jan '25

07 Jan '25

https://github.com/python/cpython/commit/e08b28235a863323ca3a7e444776bb7803… commit: e08b28235a863323ca3a7e444776bb7803e77caf branch: main author: Erlend E. Aasland <erlend(a)python.org> committer: erlend-aasland <erlend.aasland(a)protonmail.com> date: 2025-01-07T22:42:36+01:00 summary: gh-127614: Correctly check for ttyname_r() in configure (#128503) PR #14868 replaced the ttyname() call with ttyname_r(), but the old check remained. files: M Modules/clinic/posixmodule.c.h M Modules/posixmodule.c M configure M configure.ac M pyconfig.h.in diff --git a/Modules/clinic/posixmodule.c.h b/Modules/clinic/posixmodule.c.h index 554299b8598299..4e6c5b068c42b7 100644 --- a/Modules/clinic/posixmodule.c.h +++ b/Modules/clinic/posixmodule.c.h @@ -309,7 +309,7 @@ os_access(PyObject *module, PyObject *const *args, Py_ssize_t nargs, PyObject *k return return_value; } -#if defined(HAVE_TTYNAME) +#if defined(HAVE_TTYNAME_R) PyDoc_STRVAR(os_ttyname__doc__, "ttyname($module, fd, /)\n" @@ -342,7 +342,7 @@ os_ttyname(PyObject *module, PyObject *arg) return return_value; } -#endif /* defined(HAVE_TTYNAME) */ +#endif /* defined(HAVE_TTYNAME_R) */ #if defined(HAVE_CTERMID) @@ -13140,4 +13140,4 @@ os__emscripten_debugger(PyObject *module, PyObject *Py_UNUSED(ignored)) #ifndef OS__EMSCRIPTEN_DEBUGGER_METHODDEF #define OS__EMSCRIPTEN_DEBUGGER_METHODDEF #endif /* !defined(OS__EMSCRIPTEN_DEBUGGER_METHODDEF) */ -/*[clinic end generated code: output=9c2ca1dbf986c62c input=a9049054013a1b77]*/ +/*[clinic end generated code: output=39b69b279fd637f7 input=a9049054013a1b77]*/ diff --git a/Modules/posixmodule.c b/Modules/posixmodule.c index 151d469983fafb..bb8d698bfed375 100644 --- a/Modules/posixmodule.c +++ b/Modules/posixmodule.c @@ -3347,7 +3347,7 @@ os_access_impl(PyObject *module, path_t *path, int mode, int dir_fd, #endif -#ifdef HAVE_TTYNAME +#ifdef HAVE_TTYNAME_R /*[clinic input] os.ttyname diff --git a/configure b/configure index 08c10ff52ce8d5..6e1b393a3ece68 100755 --- a/configure +++ b/configure @@ -19986,10 +19986,10 @@ then : printf "%s\n" "#define HAVE_TRUNCATE 1" >>confdefs.h fi -ac_fn_c_check_func "$LINENO" "ttyname" "ac_cv_func_ttyname" -if test "x$ac_cv_func_ttyname" = xyes +ac_fn_c_check_func "$LINENO" "ttyname_r" "ac_cv_func_ttyname_r" +if test "x$ac_cv_func_ttyname_r" = xyes then : - printf "%s\n" "#define HAVE_TTYNAME 1" >>confdefs.h + printf "%s\n" "#define HAVE_TTYNAME_R 1" >>confdefs.h fi ac_fn_c_check_func "$LINENO" "umask" "ac_cv_func_umask" diff --git a/configure.ac b/configure.ac index fb1dd77be167b9..6d44be8959865d 100644 --- a/configure.ac +++ b/configure.ac @@ -5152,7 +5152,7 @@ AC_CHECK_FUNCS([ \ sigfillset siginterrupt sigpending sigrelse sigtimedwait sigwait \ sigwaitinfo snprintf splice strftime strlcpy strsignal symlinkat sync \ sysconf tcgetpgrp tcsetpgrp tempnam timegm times tmpfile \ - tmpnam tmpnam_r truncate ttyname umask uname unlinkat unlockpt utimensat utimes vfork \ + tmpnam tmpnam_r truncate ttyname_r umask uname unlinkat unlockpt utimensat utimes vfork \ wait wait3 wait4 waitid waitpid wcscoll wcsftime wcsxfrm wmemcmp writev \ ]) diff --git a/pyconfig.h.in b/pyconfig.h.in index d862966b7de4c8..874b98dc96585a 100644 --- a/pyconfig.h.in +++ b/pyconfig.h.in @@ -1506,8 +1506,8 @@ /* Define to 1 if you have the 'truncate' function. */ #undef HAVE_TRUNCATE -/* Define to 1 if you have the 'ttyname' function. */ -#undef HAVE_TTYNAME +/* Define to 1 if you have the 'ttyname_r' function. */ +#undef HAVE_TTYNAME_R /* Define to 1 if you don't have 'tm_zone' but do have the external array 'tzname'. */

1 0

[3.12] Docs: mark up json.load() using parameter list (GH-128488) (#128597)
by erlend-aasland 07 Jan '25

07 Jan '25

https://github.com/python/cpython/commit/82b061d71157226e127efe037cb3ae1be8… commit: 82b061d71157226e127efe037cb3ae1be89a09c3 branch: 3.12 author: Miss Islington (bot) <31488909+miss-islington(a)users.noreply.github.com> committer: erlend-aasland <erlend.aasland(a)protonmail.com> date: 2025-01-07T21:14:50Z summary: [3.12] Docs: mark up json.load() using parameter list (GH-128488) (#128597) (cherry picked from commit a21e31ec5417374497b0f969ee2b3dc84a0944ff) Co-authored-by: Erlend E. Aasland <erlend(a)python.org> files: M Doc/library/json.rst diff --git a/Doc/library/json.rst b/Doc/library/json.rst index eff8808a440528..88730d0cd25b8d 100644 --- a/Doc/library/json.rst +++ b/Doc/library/json.rst @@ -247,36 +247,82 @@ Basic Usage the original one. That is, ``loads(dumps(x)) != x`` if x has non-string keys. -.. function:: load(fp, *, cls=None, object_hook=None, parse_float=None, parse_int=None, parse_constant=None, object_pairs_hook=None, **kw) +.. function:: load(fp, *, cls=None, object_hook=None, parse_float=None, \ + parse_int=None, parse_constant=None, \ + object_pairs_hook=None, **kw) - Deserialize *fp* (a ``.read()``-supporting :term:`text file` or - :term:`binary file` containing a JSON document) to a Python object using - this :ref:`conversion table <json-to-py-table>`. + Deserialize *fp* to a Python object + using the :ref:`JSON-to-Python conversion table <json-to-py-table>`. - *object_hook* is an optional function that will be called with the result of - any object literal decoded (a :class:`dict`). The return value of - *object_hook* will be used instead of the :class:`dict`. This feature can - be used to implement custom decoders (e.g. `JSON-RPC - <https://www.jsonrpc.org>`_ class hinting). + :param fp: + A ``.read()``-supporting :term:`text file` or :term:`binary file` + containing the JSON document to be deserialized. + :type fp: :term:`file-like object` - *object_pairs_hook* is an optional function that will be called with the - result of any object literal decoded with an ordered list of pairs. The - return value of *object_pairs_hook* will be used instead of the - :class:`dict`. This feature can be used to implement custom decoders. If - *object_hook* is also defined, the *object_pairs_hook* takes priority. + :param cls: + If set, a custom JSON decoder. + Additional keyword arguments to :func:`!load` + will be passed to the constructor of *cls*. + If ``None`` (the default), :class:`!JSONDecoder` is used. + :type cls: a :class:`JSONDecoder` subclass + + :param object_hook: + If set, a function that is called with the result of + any object literal decoded (a :class:`dict`). + The return value of this function will be used + instead of the :class:`dict`. + This feature can be used to implement custom decoders, + for example `JSON-RPC <https://www.jsonrpc.org>`_ class hinting. + Default ``None``. + :type object_hook: :term:`callable` | None + + :param object_pairs_hook: + If set, a function that is called with the result of + any object literal decoded with an ordered list of pairs. + The return value of this function will be used + instead of the :class:`dict`. + This feature can be used to implement custom decoders. + If *object_hook* is also set, *object_pairs_hook* takes priority. + Default ``None``. + :type object_pairs_hook: :term:`callable` | None + + :param parse_float: + If set, a function that is called with + the string of every JSON float to be decoded. + If ``None`` (the default), it is equivalent to ``float(num_str)``. + This can be used to parse JSON floats into custom datatypes, + for example :class:`decimal.Decimal`. + :type parse_float: :term:`callable` | None + + :param parse_int: + If set, a function that is called with + the string of every JSON int to be decoded. + If ``None`` (the default), it is equivalent to ``int(num_str)``. + This can be used to parse JSON integers into custom datatypes, + for example :class:`float`. + :type parse_int: :term:`callable` | None + + :param parse_constant: + If set, a function that is called with one of the following strings: + ``'-Infinity'``, ``'Infinity'``, or ``'NaN'``. + This can be used to raise an exception + if invalid JSON numbers are encountered. + Default ``None``. + :type parse_constant: :term:`callable` | None + + :raises JSONDecodeError: + When the data being deserialized is not a valid JSON document. .. versionchanged:: 3.1 - Added support for *object_pairs_hook*. - *parse_float* is an optional function that will be called with the string of - every JSON float to be decoded. By default, this is equivalent to - ``float(num_str)``. This can be used to use another datatype or parser for - JSON floats (e.g. :class:`decimal.Decimal`). + * Added the optional *object_pairs_hook* parameter. + * *parse_constant* doesn't get called on 'null', 'true', 'false' anymore. - *parse_int* is an optional function that will be called with the string of - every JSON int to be decoded. By default, this is equivalent to - ``int(num_str)``. This can be used to use another datatype or parser for - JSON integers (e.g. :class:`float`). + .. versionchanged:: 3.6 + + * All optional parameters are now :ref:`keyword-only <keyword-only_parameter>`. + * *fp* can now be a :term:`binary file`. + The input encoding should be UTF-8, UTF-16 or UTF-32. .. versionchanged:: 3.11 The default *parse_int* of :func:`int` now limits the maximum length of @@ -284,27 +330,6 @@ Basic Usage conversion length limitation <int_max_str_digits>` to help avoid denial of service attacks. - *parse_constant* is an optional function that will be called with one of the - following strings: ``'-Infinity'``, ``'Infinity'``, ``'NaN'``. This can be - used to raise an exception if invalid JSON numbers are encountered. - - .. versionchanged:: 3.1 - *parse_constant* doesn't get called on 'null', 'true', 'false' anymore. - - To use a custom :class:`JSONDecoder` subclass, specify it with the ``cls`` - kwarg; otherwise :class:`JSONDecoder` is used. Additional keyword arguments - will be passed to the constructor of the class. - - If the data being deserialized is not a valid JSON document, a - :exc:`JSONDecodeError` will be raised. - - .. versionchanged:: 3.6 - All optional parameters are now :ref:`keyword-only <keyword-only_parameter>`. - - .. versionchanged:: 3.6 - *fp* can now be a :term:`binary file`. The input encoding should be - UTF-8, UTF-16 or UTF-32. - .. function:: loads(s, *, cls=None, object_hook=None, parse_float=None, parse_int=None, parse_constant=None, object_pairs_hook=None, **kw) Deserialize *s* (a :class:`str`, :class:`bytes` or :class:`bytearray`

1 0

[3.13] Docs: mark up json.load() using parameter list (GH-128488) (#128596)
by erlend-aasland 07 Jan '25

07 Jan '25

https://github.com/python/cpython/commit/984cdb788ec1be6d5e2c1a7e651ee37d53… commit: 984cdb788ec1be6d5e2c1a7e651ee37d53ad05ed branch: 3.13 author: Miss Islington (bot) <31488909+miss-islington(a)users.noreply.github.com> committer: erlend-aasland <erlend.aasland(a)protonmail.com> date: 2025-01-07T21:13:25Z summary: [3.13] Docs: mark up json.load() using parameter list (GH-128488) (#128596) (cherry picked from commit a21e31ec5417374497b0f969ee2b3dc84a0944ff) Co-authored-by: Erlend E. Aasland <erlend(a)python.org> files: M Doc/library/json.rst diff --git a/Doc/library/json.rst b/Doc/library/json.rst index 21f36cb38f1193..1a92a44f043277 100644 --- a/Doc/library/json.rst +++ b/Doc/library/json.rst @@ -258,36 +258,82 @@ Basic Usage the original one. That is, ``loads(dumps(x)) != x`` if x has non-string keys. -.. function:: load(fp, *, cls=None, object_hook=None, parse_float=None, parse_int=None, parse_constant=None, object_pairs_hook=None, **kw) +.. function:: load(fp, *, cls=None, object_hook=None, parse_float=None, \ + parse_int=None, parse_constant=None, \ + object_pairs_hook=None, **kw) - Deserialize *fp* (a ``.read()``-supporting :term:`text file` or - :term:`binary file` containing a JSON document) to a Python object using - this :ref:`conversion table <json-to-py-table>`. + Deserialize *fp* to a Python object + using the :ref:`JSON-to-Python conversion table <json-to-py-table>`. - *object_hook* is an optional function that will be called with the result of - any object literal decoded (a :class:`dict`). The return value of - *object_hook* will be used instead of the :class:`dict`. This feature can - be used to implement custom decoders (e.g. `JSON-RPC - <https://www.jsonrpc.org>`_ class hinting). + :param fp: + A ``.read()``-supporting :term:`text file` or :term:`binary file` + containing the JSON document to be deserialized. + :type fp: :term:`file-like object` - *object_pairs_hook* is an optional function that will be called with the - result of any object literal decoded with an ordered list of pairs. The - return value of *object_pairs_hook* will be used instead of the - :class:`dict`. This feature can be used to implement custom decoders. If - *object_hook* is also defined, the *object_pairs_hook* takes priority. + :param cls: + If set, a custom JSON decoder. + Additional keyword arguments to :func:`!load` + will be passed to the constructor of *cls*. + If ``None`` (the default), :class:`!JSONDecoder` is used. + :type cls: a :class:`JSONDecoder` subclass + + :param object_hook: + If set, a function that is called with the result of + any object literal decoded (a :class:`dict`). + The return value of this function will be used + instead of the :class:`dict`. + This feature can be used to implement custom decoders, + for example `JSON-RPC <https://www.jsonrpc.org>`_ class hinting. + Default ``None``. + :type object_hook: :term:`callable` | None + + :param object_pairs_hook: + If set, a function that is called with the result of + any object literal decoded with an ordered list of pairs. + The return value of this function will be used + instead of the :class:`dict`. + This feature can be used to implement custom decoders. + If *object_hook* is also set, *object_pairs_hook* takes priority. + Default ``None``. + :type object_pairs_hook: :term:`callable` | None + + :param parse_float: + If set, a function that is called with + the string of every JSON float to be decoded. + If ``None`` (the default), it is equivalent to ``float(num_str)``. + This can be used to parse JSON floats into custom datatypes, + for example :class:`decimal.Decimal`. + :type parse_float: :term:`callable` | None + + :param parse_int: + If set, a function that is called with + the string of every JSON int to be decoded. + If ``None`` (the default), it is equivalent to ``int(num_str)``. + This can be used to parse JSON integers into custom datatypes, + for example :class:`float`. + :type parse_int: :term:`callable` | None + + :param parse_constant: + If set, a function that is called with one of the following strings: + ``'-Infinity'``, ``'Infinity'``, or ``'NaN'``. + This can be used to raise an exception + if invalid JSON numbers are encountered. + Default ``None``. + :type parse_constant: :term:`callable` | None + + :raises JSONDecodeError: + When the data being deserialized is not a valid JSON document. .. versionchanged:: 3.1 - Added support for *object_pairs_hook*. - *parse_float* is an optional function that will be called with the string of - every JSON float to be decoded. By default, this is equivalent to - ``float(num_str)``. This can be used to use another datatype or parser for - JSON floats (e.g. :class:`decimal.Decimal`). + * Added the optional *object_pairs_hook* parameter. + * *parse_constant* doesn't get called on 'null', 'true', 'false' anymore. - *parse_int* is an optional function that will be called with the string of - every JSON int to be decoded. By default, this is equivalent to - ``int(num_str)``. This can be used to use another datatype or parser for - JSON integers (e.g. :class:`float`). + .. versionchanged:: 3.6 + + * All optional parameters are now :ref:`keyword-only <keyword-only_parameter>`. + * *fp* can now be a :term:`binary file`. + The input encoding should be UTF-8, UTF-16 or UTF-32. .. versionchanged:: 3.11 The default *parse_int* of :func:`int` now limits the maximum length of @@ -295,27 +341,6 @@ Basic Usage conversion length limitation <int_max_str_digits>` to help avoid denial of service attacks. - *parse_constant* is an optional function that will be called with one of the - following strings: ``'-Infinity'``, ``'Infinity'``, ``'NaN'``. This can be - used to raise an exception if invalid JSON numbers are encountered. - - .. versionchanged:: 3.1 - *parse_constant* doesn't get called on 'null', 'true', 'false' anymore. - - To use a custom :class:`JSONDecoder` subclass, specify it with the ``cls`` - kwarg; otherwise :class:`JSONDecoder` is used. Additional keyword arguments - will be passed to the constructor of the class. - - If the data being deserialized is not a valid JSON document, a - :exc:`JSONDecodeError` will be raised. - - .. versionchanged:: 3.6 - All optional parameters are now :ref:`keyword-only <keyword-only_parameter>`. - - .. versionchanged:: 3.6 - *fp* can now be a :term:`binary file`. The input encoding should be - UTF-8, UTF-16 or UTF-32. - .. function:: loads(s, *, cls=None, object_hook=None, parse_float=None, parse_int=None, parse_constant=None, object_pairs_hook=None, **kw) Deserialize *s* (a :class:`str`, :class:`bytes` or :class:`bytearray`

1 0

gh-127350: Add more tests for Py_fopen() (GH-128587)
by serhiy-storchaka 07 Jan '25

07 Jan '25

https://github.com/python/cpython/commit/07e6aa2efc5b2c7ecbbdb41636013c64e3… commit: 07e6aa2efc5b2c7ecbbdb41636013c64e335a2ba branch: main author: Serhiy Storchaka <storchaka(a)gmail.com> committer: serhiy-storchaka <storchaka(a)gmail.com> date: 2025-01-07T21:13:08Z summary: gh-127350: Add more tests for Py_fopen() (GH-128587) files: M Lib/test/test_capi/test_file.py M Modules/_testcapi/clinic/file.c.h M Modules/_testcapi/file.c diff --git a/Lib/test/test_capi/test_file.py b/Lib/test/test_capi/test_file.py index 8a08a0a93eb8b7..a67a5121c4588b 100644 --- a/Lib/test/test_capi/test_file.py +++ b/Lib/test/test_capi/test_file.py @@ -5,6 +5,8 @@ _testcapi = import_helper.import_module('_testcapi') +NULL = None + class CAPIFileTest(unittest.TestCase): def test_py_fopen(self): @@ -25,7 +27,9 @@ def test_py_fopen(self): os_helper.TESTFN, os.fsencode(os_helper.TESTFN), ] - # TESTFN_UNDECODABLE cannot be used to create a file on macOS/WASI. + if os_helper.TESTFN_UNDECODABLE is not None: + filenames.append(os_helper.TESTFN_UNDECODABLE) + filenames.append(os.fsdecode(os_helper.TESTFN_UNDECODABLE)) if os_helper.TESTFN_UNENCODABLE is not None: filenames.append(os_helper.TESTFN_UNENCODABLE) for filename in filenames: @@ -33,7 +37,12 @@ def test_py_fopen(self): try: with open(filename, "wb") as fp: fp.write(source) - + except OSError: + # TESTFN_UNDECODABLE cannot be used to create a file + # on macOS/WASI. + filename = None + continue + try: data = _testcapi.py_fopen(filename, "rb") self.assertEqual(data, source[:256]) finally: @@ -47,7 +56,14 @@ def test_py_fopen(self): # non-ASCII mode failing with "Invalid argument" with self.assertRaises(OSError): - _testcapi.py_fopen(__file__, "\xe9") + _testcapi.py_fopen(__file__, b"\xc2\x80") + with self.assertRaises(OSError): + # \x98 is invalid in cp1250, cp1251, cp1257 + # \x9d is invalid in cp1252-cp1255, cp1258 + _testcapi.py_fopen(__file__, b"\xc2\x98\xc2\x9d") + # UnicodeDecodeError can come from the audit hook code + with self.assertRaises((UnicodeDecodeError, OSError)): + _testcapi.py_fopen(__file__, b"\x98\x9d") # invalid filename type for invalid_type in (123, object()): @@ -60,7 +76,8 @@ def test_py_fopen(self): # On Windows, the file mode is limited to 10 characters _testcapi.py_fopen(__file__, "rt+, ccs=UTF-8") - # CRASHES py_fopen(__file__, None) + # CRASHES _testcapi.py_fopen(NULL, 'rb') + # CRASHES _testcapi.py_fopen(__file__, NULL) if __name__ == "__main__": diff --git a/Modules/_testcapi/clinic/file.c.h b/Modules/_testcapi/clinic/file.c.h index 2ca21fffcef680..fddbf48071bd3b 100644 --- a/Modules/_testcapi/clinic/file.c.h +++ b/Modules/_testcapi/clinic/file.c.h @@ -14,7 +14,8 @@ PyDoc_STRVAR(_testcapi_py_fopen__doc__, {"py_fopen", _PyCFunction_CAST(_testcapi_py_fopen), METH_FASTCALL, _testcapi_py_fopen__doc__}, static PyObject * -_testcapi_py_fopen_impl(PyObject *module, PyObject *path, const char *mode); +_testcapi_py_fopen_impl(PyObject *module, PyObject *path, const char *mode, + Py_ssize_t mode_length); static PyObject * _testcapi_py_fopen(PyObject *module, PyObject *const *args, Py_ssize_t nargs) @@ -22,27 +23,15 @@ _testcapi_py_fopen(PyObject *module, PyObject *const *args, Py_ssize_t nargs) PyObject *return_value = NULL; PyObject *path; const char *mode; - - if (!_PyArg_CheckPositional("py_fopen", nargs, 2, 2)) { - goto exit; - } - path = args[0]; - if (!PyUnicode_Check(args[1])) { - _PyArg_BadArgument("py_fopen", "argument 2", "str", args[1]); - goto exit; - } Py_ssize_t mode_length; - mode = PyUnicode_AsUTF8AndSize(args[1], &mode_length); - if (mode == NULL) { - goto exit; - } - if (strlen(mode) != (size_t)mode_length) { - PyErr_SetString(PyExc_ValueError, "embedded null character"); + + if (!_PyArg_ParseStack(args, nargs, "Oz#:py_fopen", + &path, &mode, &mode_length)) { goto exit; } - return_value = _testcapi_py_fopen_impl(module, path, mode); + return_value = _testcapi_py_fopen_impl(module, path, mode, mode_length); exit: return return_value; } -/*[clinic end generated code: output=c9fe964c3e5a0c32 input=a9049054013a1b77]*/ +/*[clinic end generated code: output=c4dc92400306c3eb input=a9049054013a1b77]*/ diff --git a/Modules/_testcapi/file.c b/Modules/_testcapi/file.c index 4bad43010fd440..d15173fc7959e5 100644 --- a/Modules/_testcapi/file.c +++ b/Modules/_testcapi/file.c @@ -14,16 +14,18 @@ module _testcapi _testcapi.py_fopen path: object - mode: str + mode: str(zeroes=True, accept={robuffer, str, NoneType}) / Call Py_fopen(), fread(256) and Py_fclose(). Return read bytes. [clinic start generated code]*/ static PyObject * -_testcapi_py_fopen_impl(PyObject *module, PyObject *path, const char *mode) -/*[clinic end generated code: output=5a900af000f759de input=d7e7b8f0fd151953]*/ +_testcapi_py_fopen_impl(PyObject *module, PyObject *path, const char *mode, + Py_ssize_t mode_length) +/*[clinic end generated code: output=69840d0cfd8b7fbb input=f3a579dd7eb60926]*/ { + NULLABLE(path); FILE *fp = Py_fopen(path, mode); if (fp == NULL) { return NULL;

1 0

Docs: mark up json.load() using parameter list (#128488)
by erlend-aasland 07 Jan '25

07 Jan '25

https://github.com/python/cpython/commit/a21e31ec5417374497b0f969ee2b3dc84a… commit: a21e31ec5417374497b0f969ee2b3dc84a0944ff branch: main author: Erlend E. Aasland <erlend(a)python.org> committer: erlend-aasland <erlend.aasland(a)protonmail.com> date: 2025-01-07T22:07:17+01:00 summary: Docs: mark up json.load() using parameter list (#128488) files: M Doc/library/json.rst diff --git a/Doc/library/json.rst b/Doc/library/json.rst index 169291f74f44a5..cf516cf3dc1d29 100644 --- a/Doc/library/json.rst +++ b/Doc/library/json.rst @@ -258,36 +258,82 @@ Basic Usage the original one. That is, ``loads(dumps(x)) != x`` if x has non-string keys. -.. function:: load(fp, *, cls=None, object_hook=None, parse_float=None, parse_int=None, parse_constant=None, object_pairs_hook=None, **kw) +.. function:: load(fp, *, cls=None, object_hook=None, parse_float=None, \ + parse_int=None, parse_constant=None, \ + object_pairs_hook=None, **kw) - Deserialize *fp* (a ``.read()``-supporting :term:`text file` or - :term:`binary file` containing a JSON document) to a Python object using - this :ref:`conversion table <json-to-py-table>`. + Deserialize *fp* to a Python object + using the :ref:`JSON-to-Python conversion table <json-to-py-table>`. - *object_hook* is an optional function that will be called with the result of - any object literal decoded (a :class:`dict`). The return value of - *object_hook* will be used instead of the :class:`dict`. This feature can - be used to implement custom decoders (e.g. `JSON-RPC - <https://www.jsonrpc.org>`_ class hinting). + :param fp: + A ``.read()``-supporting :term:`text file` or :term:`binary file` + containing the JSON document to be deserialized. + :type fp: :term:`file-like object` - *object_pairs_hook* is an optional function that will be called with the - result of any object literal decoded with an ordered list of pairs. The - return value of *object_pairs_hook* will be used instead of the - :class:`dict`. This feature can be used to implement custom decoders. If - *object_hook* is also defined, the *object_pairs_hook* takes priority. + :param cls: + If set, a custom JSON decoder. + Additional keyword arguments to :func:`!load` + will be passed to the constructor of *cls*. + If ``None`` (the default), :class:`!JSONDecoder` is used. + :type cls: a :class:`JSONDecoder` subclass + + :param object_hook: + If set, a function that is called with the result of + any object literal decoded (a :class:`dict`). + The return value of this function will be used + instead of the :class:`dict`. + This feature can be used to implement custom decoders, + for example `JSON-RPC <https://www.jsonrpc.org>`_ class hinting. + Default ``None``. + :type object_hook: :term:`callable` | None + + :param object_pairs_hook: + If set, a function that is called with the result of + any object literal decoded with an ordered list of pairs. + The return value of this function will be used + instead of the :class:`dict`. + This feature can be used to implement custom decoders. + If *object_hook* is also set, *object_pairs_hook* takes priority. + Default ``None``. + :type object_pairs_hook: :term:`callable` | None + + :param parse_float: + If set, a function that is called with + the string of every JSON float to be decoded. + If ``None`` (the default), it is equivalent to ``float(num_str)``. + This can be used to parse JSON floats into custom datatypes, + for example :class:`decimal.Decimal`. + :type parse_float: :term:`callable` | None + + :param parse_int: + If set, a function that is called with + the string of every JSON int to be decoded. + If ``None`` (the default), it is equivalent to ``int(num_str)``. + This can be used to parse JSON integers into custom datatypes, + for example :class:`float`. + :type parse_int: :term:`callable` | None + + :param parse_constant: + If set, a function that is called with one of the following strings: + ``'-Infinity'``, ``'Infinity'``, or ``'NaN'``. + This can be used to raise an exception + if invalid JSON numbers are encountered. + Default ``None``. + :type parse_constant: :term:`callable` | None + + :raises JSONDecodeError: + When the data being deserialized is not a valid JSON document. .. versionchanged:: 3.1 - Added support for *object_pairs_hook*. - *parse_float* is an optional function that will be called with the string of - every JSON float to be decoded. By default, this is equivalent to - ``float(num_str)``. This can be used to use another datatype or parser for - JSON floats (e.g. :class:`decimal.Decimal`). + * Added the optional *object_pairs_hook* parameter. + * *parse_constant* doesn't get called on 'null', 'true', 'false' anymore. - *parse_int* is an optional function that will be called with the string of - every JSON int to be decoded. By default, this is equivalent to - ``int(num_str)``. This can be used to use another datatype or parser for - JSON integers (e.g. :class:`float`). + .. versionchanged:: 3.6 + + * All optional parameters are now :ref:`keyword-only <keyword-only_parameter>`. + * *fp* can now be a :term:`binary file`. + The input encoding should be UTF-8, UTF-16 or UTF-32. .. versionchanged:: 3.11 The default *parse_int* of :func:`int` now limits the maximum length of @@ -295,27 +341,6 @@ Basic Usage conversion length limitation <int_max_str_digits>` to help avoid denial of service attacks. - *parse_constant* is an optional function that will be called with one of the - following strings: ``'-Infinity'``, ``'Infinity'``, ``'NaN'``. This can be - used to raise an exception if invalid JSON numbers are encountered. - - .. versionchanged:: 3.1 - *parse_constant* doesn't get called on 'null', 'true', 'false' anymore. - - To use a custom :class:`JSONDecoder` subclass, specify it with the ``cls`` - kwarg; otherwise :class:`JSONDecoder` is used. Additional keyword arguments - will be passed to the constructor of the class. - - If the data being deserialized is not a valid JSON document, a - :exc:`JSONDecodeError` will be raised. - - .. versionchanged:: 3.6 - All optional parameters are now :ref:`keyword-only <keyword-only_parameter>`. - - .. versionchanged:: 3.6 - *fp* can now be a :term:`binary file`. The input encoding should be - UTF-8, UTF-16 or UTF-32. - .. function:: loads(s, *, cls=None, object_hook=None, parse_float=None, parse_int=None, parse_constant=None, object_pairs_hook=None, **kw) Deserialize *s* (a :class:`str`, :class:`bytes` or :class:`bytearray`

1 0

[3.12] gh-98188: Fix EmailMessage.get_payload to decode data when CTE value has extra text (GH-127547) (#128529)
by bitdancer 07 Jan '25

07 Jan '25

https://github.com/python/cpython/commit/dae5b16bb3b51702d3506ef21550b22035… commit: dae5b16bb3b51702d3506ef21550b220359ff971 branch: 3.12 author: Miss Islington (bot) <31488909+miss-islington(a)users.noreply.github.com> committer: bitdancer <rdmurray(a)bitdance.com> date: 2025-01-07T12:44:56-05:00 summary: [3.12] gh-98188: Fix EmailMessage.get_payload to decode data when CTE value has extra text (GH-127547) (#128529) gh-98188: Fix EmailMessage.get_payload to decode data when CTE value has extra text (GH-127547) Up to this point message handling has been very strict with regards to content encoding values: mixed case was accepted, but trailing blanks or other text would cause decoding failure, even if the first token was a valid encoding. By Postel's Rule we should go ahead and decode as long as we can recognize that first token. We have not thought of any security or backward compatibility concerns with this fix. This fix does introduce a new technique/pattern to the Message code: we look to see if the header has a 'cte' attribute, and if so we use that. This effectively promotes the header API exposed by HeaderRegistry to an API that any header parser "should" support. This seems like a reasonable thing to do. It is not, however, a requirement, as the string value of the header is still used if there is no cte attribute. The full fix (ignore any trailing blanks or blank-separated trailing text) applies only to the non-compat32 API. compat32 is only fixed to the extent that it now ignores trailing spaces. Note that the HeaderRegistry parsing still records a HeaderDefect if there is extra text. (cherry picked from commit a62ba52f1439c1f878a3ff9b8544caf9aeef9b90) Co-authored-by: RanKKI <hliu86.me(a)gmail.com> Co-authored-by: Bénédikt Tran <10796600+picnixz(a)users.noreply.github.com> files: A Misc/NEWS.d/next/Library/2024-12-03-14-45-16.gh-issue-98188.GX9i2b.rst M Lib/email/message.py M Lib/test/test_email/test_email.py M Lib/test/test_email/test_headerregistry.py M Misc/ACKS diff --git a/Lib/email/message.py b/Lib/email/message.py index 46bb8c21942af8..6b7c3a2377765a 100644 --- a/Lib/email/message.py +++ b/Lib/email/message.py @@ -286,8 +286,12 @@ def get_payload(self, i=None, decode=False): if i is not None and not isinstance(self._payload, list): raise TypeError('Expected list, got %s' % type(self._payload)) payload = self._payload - # cte might be a Header, so for now stringify it. - cte = str(self.get('content-transfer-encoding', '')).lower() + cte = self.get('content-transfer-encoding', '') + if hasattr(cte, 'cte'): + cte = cte.cte + else: + # cte might be a Header, so for now stringify it. + cte = str(cte).strip().lower() # payload may be bytes here. if not decode: if isinstance(payload, str) and utils._has_surrogates(payload): diff --git a/Lib/test/test_email/test_email.py b/Lib/test/test_email/test_email.py index ef8aa0d53c59ac..e33946de350ec0 100644 --- a/Lib/test/test_email/test_email.py +++ b/Lib/test/test_email/test_email.py @@ -810,6 +810,16 @@ def test_unicode_body_defaults_to_utf8_encoding(self): w4kgdGVzdGFiYwo= """)) + def test_string_payload_with_base64_cte(self): + msg = email.message_from_string(textwrap.dedent("""\ + Content-Transfer-Encoding: base64 + + SGVsbG8uIFRlc3Rpbmc= + """), policy=email.policy.default) + self.assertEqual(msg.get_payload(decode=True), b"Hello. Testing") + self.assertDefectsEqual(msg['content-transfer-encoding'].defects, []) + + # Test the email.encoders module class TestEncoders(unittest.TestCase): @@ -2352,6 +2362,40 @@ def test_missing_header_body_separator(self): self.assertDefectsEqual(msg.defects, [errors.MissingHeaderBodySeparatorDefect]) + def test_string_payload_with_extra_space_after_cte(self): + # https://github.com/python/cpython/issues/98188 + cte = "base64 " + msg = email.message_from_string(textwrap.dedent(f"""\ + Content-Transfer-Encoding: {cte} + + SGVsbG8uIFRlc3Rpbmc= + """), policy=email.policy.default) + self.assertEqual(msg.get_payload(decode=True), b"Hello. Testing") + self.assertDefectsEqual(msg['content-transfer-encoding'].defects, []) + + def test_string_payload_with_extra_text_after_cte(self): + msg = email.message_from_string(textwrap.dedent("""\ + Content-Transfer-Encoding: base64 some text + + SGVsbG8uIFRlc3Rpbmc= + """), policy=email.policy.default) + self.assertEqual(msg.get_payload(decode=True), b"Hello. Testing") + cte = msg['content-transfer-encoding'] + self.assertDefectsEqual(cte.defects, [email.errors.InvalidHeaderDefect]) + + def test_string_payload_with_extra_space_after_cte_compat32(self): + cte = "base64 " + msg = email.message_from_string(textwrap.dedent(f"""\ + Content-Transfer-Encoding: {cte} + + SGVsbG8uIFRlc3Rpbmc= + """), policy=email.policy.compat32) + pasted_cte = msg['content-transfer-encoding'] + self.assertEqual(pasted_cte, cte) + self.assertEqual(msg.get_payload(decode=True), b"Hello. Testing") + self.assertDefectsEqual(msg.defects, []) + + # Test RFC 2047 header encoding and decoding class TestRFC2047(TestEmailBase): diff --git a/Lib/test/test_email/test_headerregistry.py b/Lib/test/test_email/test_headerregistry.py index 5a608a033c7e54..a579f9eeb7f4bf 100644 --- a/Lib/test/test_email/test_headerregistry.py +++ b/Lib/test/test_email/test_headerregistry.py @@ -838,6 +838,11 @@ def cte_as_value(self, '7bit', [errors.InvalidHeaderDefect]), + 'extra_space_after_cte': ( + 'base64 ', + 'base64', + []), + } diff --git a/Misc/ACKS b/Misc/ACKS index 0f8b79f26fbf6b..14450325647590 100644 --- a/Misc/ACKS +++ b/Misc/ACKS @@ -1112,6 +1112,7 @@ Gregor Lingl Everett Lipman Mirko Liss Alexander Liu +Hui Liu Yuan Liu Nick Lockwood Stephanie Lockwood diff --git a/Misc/NEWS.d/next/Library/2024-12-03-14-45-16.gh-issue-98188.GX9i2b.rst b/Misc/NEWS.d/next/Library/2024-12-03-14-45-16.gh-issue-98188.GX9i2b.rst new file mode 100644 index 00000000000000..30ab8cfc3f0bc6 --- /dev/null +++ b/Misc/NEWS.d/next/Library/2024-12-03-14-45-16.gh-issue-98188.GX9i2b.rst @@ -0,0 +1,3 @@ +Fix an issue in :meth:`email.message.Message.get_payload` where data +cannot be decoded if the Content Transfer Encoding mechanism contains +trailing whitespaces or additional junk text. Patch by Hui Liu.

1 0

[3.13] gh-98188: Fix EmailMessage.get_payload to decode data when CTE value has extra text (GH-127547) (#128528)
by bitdancer 07 Jan '25

07 Jan '25

https://github.com/python/cpython/commit/ad3bbb6b0df9a83e5e9fc26344468747e8… commit: ad3bbb6b0df9a83e5e9fc26344468747e856d0b6 branch: 3.13 author: Miss Islington (bot) <31488909+miss-islington(a)users.noreply.github.com> committer: bitdancer <rdmurray(a)bitdance.com> date: 2025-01-07T12:43:04-05:00 summary: [3.13] gh-98188: Fix EmailMessage.get_payload to decode data when CTE value has extra text (GH-127547) (#128528) gh-98188: Fix EmailMessage.get_payload to decode data when CTE value has extra text (GH-127547) Up to this point message handling has been very strict with regards to content encoding values: mixed case was accepted, but trailing blanks or other text would cause decoding failure, even if the first token was a valid encoding. By Postel's Rule we should go ahead and decode as long as we can recognize that first token. We have not thought of any security or backward compatibility concerns with this fix. This fix does introduce a new technique/pattern to the Message code: we look to see if the header has a 'cte' attribute, and if so we use that. This effectively promotes the header API exposed by HeaderRegistry to an API that any header parser "should" support. This seems like a reasonable thing to do. It is not, however, a requirement, as the string value of the header is still used if there is no cte attribute. The full fix (ignore any trailing blanks or blank-separated trailing text) applies only to the non-compat32 API. compat32 is only fixed to the extent that it now ignores trailing spaces. Note that the HeaderRegistry parsing still records a HeaderDefect if there is extra text. (cherry picked from commit a62ba52f1439c1f878a3ff9b8544caf9aeef9b90) Co-authored-by: RanKKI <hliu86.me(a)gmail.com> Co-authored-by: Bénédikt Tran <10796600+picnixz(a)users.noreply.github.com> files: A Misc/NEWS.d/next/Library/2024-12-03-14-45-16.gh-issue-98188.GX9i2b.rst M Lib/email/message.py M Lib/test/test_email/test_email.py M Lib/test/test_email/test_headerregistry.py M Misc/ACKS diff --git a/Lib/email/message.py b/Lib/email/message.py index 46bb8c21942af8..6b7c3a2377765a 100644 --- a/Lib/email/message.py +++ b/Lib/email/message.py @@ -286,8 +286,12 @@ def get_payload(self, i=None, decode=False): if i is not None and not isinstance(self._payload, list): raise TypeError('Expected list, got %s' % type(self._payload)) payload = self._payload - # cte might be a Header, so for now stringify it. - cte = str(self.get('content-transfer-encoding', '')).lower() + cte = self.get('content-transfer-encoding', '') + if hasattr(cte, 'cte'): + cte = cte.cte + else: + # cte might be a Header, so for now stringify it. + cte = str(cte).strip().lower() # payload may be bytes here. if not decode: if isinstance(payload, str) and utils._has_surrogates(payload): diff --git a/Lib/test/test_email/test_email.py b/Lib/test/test_email/test_email.py index 65ddbabcaa1997..925a638572d79c 100644 --- a/Lib/test/test_email/test_email.py +++ b/Lib/test/test_email/test_email.py @@ -810,6 +810,16 @@ def test_unicode_body_defaults_to_utf8_encoding(self): w4kgdGVzdGFiYwo= """)) + def test_string_payload_with_base64_cte(self): + msg = email.message_from_string(textwrap.dedent("""\ + Content-Transfer-Encoding: base64 + + SGVsbG8uIFRlc3Rpbmc= + """), policy=email.policy.default) + self.assertEqual(msg.get_payload(decode=True), b"Hello. Testing") + self.assertDefectsEqual(msg['content-transfer-encoding'].defects, []) + + # Test the email.encoders module class TestEncoders(unittest.TestCase): @@ -2352,6 +2362,40 @@ def test_missing_header_body_separator(self): self.assertDefectsEqual(msg.defects, [errors.MissingHeaderBodySeparatorDefect]) + def test_string_payload_with_extra_space_after_cte(self): + # https://github.com/python/cpython/issues/98188 + cte = "base64 " + msg = email.message_from_string(textwrap.dedent(f"""\ + Content-Transfer-Encoding: {cte} + + SGVsbG8uIFRlc3Rpbmc= + """), policy=email.policy.default) + self.assertEqual(msg.get_payload(decode=True), b"Hello. Testing") + self.assertDefectsEqual(msg['content-transfer-encoding'].defects, []) + + def test_string_payload_with_extra_text_after_cte(self): + msg = email.message_from_string(textwrap.dedent("""\ + Content-Transfer-Encoding: base64 some text + + SGVsbG8uIFRlc3Rpbmc= + """), policy=email.policy.default) + self.assertEqual(msg.get_payload(decode=True), b"Hello. Testing") + cte = msg['content-transfer-encoding'] + self.assertDefectsEqual(cte.defects, [email.errors.InvalidHeaderDefect]) + + def test_string_payload_with_extra_space_after_cte_compat32(self): + cte = "base64 " + msg = email.message_from_string(textwrap.dedent(f"""\ + Content-Transfer-Encoding: {cte} + + SGVsbG8uIFRlc3Rpbmc= + """), policy=email.policy.compat32) + pasted_cte = msg['content-transfer-encoding'] + self.assertEqual(pasted_cte, cte) + self.assertEqual(msg.get_payload(decode=True), b"Hello. Testing") + self.assertDefectsEqual(msg.defects, []) + + # Test RFC 2047 header encoding and decoding class TestRFC2047(TestEmailBase): diff --git a/Lib/test/test_email/test_headerregistry.py b/Lib/test/test_email/test_headerregistry.py index 5a608a033c7e54..a579f9eeb7f4bf 100644 --- a/Lib/test/test_email/test_headerregistry.py +++ b/Lib/test/test_email/test_headerregistry.py @@ -838,6 +838,11 @@ def cte_as_value(self, '7bit', [errors.InvalidHeaderDefect]), + 'extra_space_after_cte': ( + 'base64 ', + 'base64', + []), + } diff --git a/Misc/ACKS b/Misc/ACKS index bed3e028d0a18d..206c9b849cf0ea 100644 --- a/Misc/ACKS +++ b/Misc/ACKS @@ -1122,6 +1122,7 @@ Gregor Lingl Everett Lipman Mirko Liss Alexander Liu +Hui Liu Yuan Liu Nick Lockwood Stephanie Lockwood diff --git a/Misc/NEWS.d/next/Library/2024-12-03-14-45-16.gh-issue-98188.GX9i2b.rst b/Misc/NEWS.d/next/Library/2024-12-03-14-45-16.gh-issue-98188.GX9i2b.rst new file mode 100644 index 00000000000000..30ab8cfc3f0bc6 --- /dev/null +++ b/Misc/NEWS.d/next/Library/2024-12-03-14-45-16.gh-issue-98188.GX9i2b.rst @@ -0,0 +1,3 @@ +Fix an issue in :meth:`email.message.Message.get_payload` where data +cannot be decoded if the Content Transfer Encoding mechanism contains +trailing whitespaces or additional junk text. Patch by Hui Liu.

1 0

[3.13] gh-128472: Add `-skip-funcs` to BOLT options to fix computed goto errors (gh-128511) (#128573)
by corona10 07 Jan '25

07 Jan '25

https://github.com/python/cpython/commit/692a955cea019274b16e0132732be761d4… commit: 692a955cea019274b16e0132732be761d48b2606 branch: 3.13 author: Miss Islington (bot) <31488909+miss-islington(a)users.noreply.github.com> committer: corona10 <donghee.na92(a)gmail.com> date: 2025-01-08T01:59:20+09:00 summary: [3.13] gh-128472: Add `-skip-funcs` to BOLT options to fix computed goto errors (gh-128511) (#128573) gh-128472: Add `-skip-funcs` to BOLT options to fix computed goto errors (gh-128511) * Add `-skip-funcs` to BOLT options to fix computed goto errors * NEWS --------- (cherry picked from commit 24b147a19b360c49cb1740aa46211d342aaa071f) Co-authored-by: Zanie Blue <contact(a)zanie.dev> Co-authored-by: Gregory Szorc <gregory.szorc(a)gmail.com> files: A Misc/NEWS.d/next/Build/2025-01-04-22-39-10.gh-issue-128472.Wt5E6M.rst M configure M configure.ac diff --git a/Misc/NEWS.d/next/Build/2025-01-04-22-39-10.gh-issue-128472.Wt5E6M.rst b/Misc/NEWS.d/next/Build/2025-01-04-22-39-10.gh-issue-128472.Wt5E6M.rst new file mode 100644 index 00000000000000..c6233e1f2d8693 --- /dev/null +++ b/Misc/NEWS.d/next/Build/2025-01-04-22-39-10.gh-issue-128472.Wt5E6M.rst @@ -0,0 +1,2 @@ +Skip BOLT optimization of functions using computed gotos, fixing errors on +build with LLVM 19. diff --git a/configure b/configure index 0d8a9ffa8a2605..ff8e29de1439ec 100755 --- a/configure +++ b/configure @@ -9335,7 +9335,7 @@ fi printf %s "checking BOLT_COMMON_FLAGS... " >&6; } if test -z "${BOLT_COMMON_FLAGS}" then - BOLT_COMMON_FLAGS=-update-debug-sections + BOLT_COMMON_FLAGS=" -update-debug-sections -skip-funcs=_PyEval_EvalFrameDefault,sre_ucs1_match/1,sre_ucs2_match/1,sre_ucs4_match/1 " fi diff --git a/configure.ac b/configure.ac index b3e112a0a181d6..dd4e0b4723c1d3 100644 --- a/configure.ac +++ b/configure.ac @@ -2224,7 +2224,14 @@ if test -z "${BOLT_COMMON_FLAGS}" then AS_VAR_SET( [BOLT_COMMON_FLAGS], - [-update-debug-sections] + [m4_normalize(" + [-update-debug-sections] + + dnl At least LLVM 19.x doesn't support computed gotos in PIC compiled code. + dnl Exclude functions containing computed gotos. + dnl TODO this may be fixed in LLVM 20.x via https://github.com/llvm/llvm-project/pull/120267. + [-skip-funcs=_PyEval_EvalFrameDefault,sre_ucs1_match/1,sre_ucs2_match/1,sre_ucs4_match/1] + ")] ) fi

1 0

gh-115999: Add free-threaded specialization for COMPARE_OP (#126410)
by Yhg1s 07 Jan '25

07 Jan '25

https://github.com/python/cpython/commit/8f93dd8a8f237b277abad20d566df90c5c… commit: 8f93dd8a8f237b277abad20d566df90c5cbd7f1e branch: main author: T. Wouters <thomas(a)python.org> committer: Yhg1s <thomas(a)python.org> date: 2025-01-07T06:41:01-08:00 summary: gh-115999: Add free-threaded specialization for COMPARE_OP (#126410) Add free-threaded specialization for COMPARE_OP, and tests for COMPARE_OP specialization in general. Co-authored-by: Donghee Na <donghee.na92(a)gmail.com> files: M Lib/test/test_opcache.py M Python/bytecodes.c M Python/generated_cases.c.h M Python/specialize.c diff --git a/Lib/test/test_opcache.py b/Lib/test/test_opcache.py index 79f452f8068c7f..c7cd4c2e8a3146 100644 --- a/Lib/test/test_opcache.py +++ b/Lib/test/test_opcache.py @@ -1598,6 +1598,39 @@ def __getitem__(self, item): self.assert_specialized(binary_subscr_getitems, "BINARY_SUBSCR_GETITEM") self.assert_no_opcode(binary_subscr_getitems, "BINARY_SUBSCR") + @cpython_only + @requires_specialization_ft + def test_compare_op(self): + def compare_op_int(): + for _ in range(100): + a, b = 1, 2 + c = a == b + self.assertFalse(c) + + compare_op_int() + self.assert_specialized(compare_op_int, "COMPARE_OP_INT") + self.assert_no_opcode(compare_op_int, "COMPARE_OP") + + def compare_op_float(): + for _ in range(100): + a, b = 1.0, 2.0 + c = a == b + self.assertFalse(c) + + compare_op_float() + self.assert_specialized(compare_op_float, "COMPARE_OP_FLOAT") + self.assert_no_opcode(compare_op_float, "COMPARE_OP") + + def compare_op_str(): + for _ in range(100): + a, b = "spam", "ham" + c = a == b + self.assertFalse(c) + + compare_op_str() + self.assert_specialized(compare_op_str, "COMPARE_OP_STR") + self.assert_no_opcode(compare_op_str, "COMPARE_OP") + if __name__ == "__main__": unittest.main() diff --git a/Python/bytecodes.c b/Python/bytecodes.c index 4961693c7e654a..ec1cd00962ac0a 100644 --- a/Python/bytecodes.c +++ b/Python/bytecodes.c @@ -2464,7 +2464,7 @@ dummy_func( }; specializing op(_SPECIALIZE_COMPARE_OP, (counter/1, left, right -- left, right)) { - #if ENABLE_SPECIALIZATION + #if ENABLE_SPECIALIZATION_FT if (ADAPTIVE_COUNTER_TRIGGERS(counter)) { next_instr = this_instr; _Py_Specialize_CompareOp(left, right, next_instr, oparg); diff --git a/Python/generated_cases.c.h b/Python/generated_cases.c.h index b73844ca2d9542..eaa8a563464068 100644 --- a/Python/generated_cases.c.h +++ b/Python/generated_cases.c.h @@ -3229,7 +3229,7 @@ left = stack_pointer[-2]; uint16_t counter = read_u16(&this_instr[1].cache); (void)counter; - #if ENABLE_SPECIALIZATION + #if ENABLE_SPECIALIZATION_FT if (ADAPTIVE_COUNTER_TRIGGERS(counter)) { next_instr = this_instr; _PyFrame_SetStackPointer(frame, stack_pointer); diff --git a/Python/specialize.c b/Python/specialize.c index c918c77779d20d..c9325c39210874 100644 --- a/Python/specialize.c +++ b/Python/specialize.c @@ -2480,23 +2480,23 @@ _Py_Specialize_CompareOp(_PyStackRef lhs_st, _PyStackRef rhs_st, _Py_CODEUNIT *i { PyObject *lhs = PyStackRef_AsPyObjectBorrow(lhs_st); PyObject *rhs = PyStackRef_AsPyObjectBorrow(rhs_st); + uint8_t specialized_op; - assert(ENABLE_SPECIALIZATION); + assert(ENABLE_SPECIALIZATION_FT); assert(_PyOpcode_Caches[COMPARE_OP] == INLINE_CACHE_ENTRIES_COMPARE_OP); // All of these specializations compute boolean values, so they're all valid // regardless of the fifth-lowest oparg bit. - _PyCompareOpCache *cache = (_PyCompareOpCache *)(instr + 1); if (Py_TYPE(lhs) != Py_TYPE(rhs)) { SPECIALIZATION_FAIL(COMPARE_OP, compare_op_fail_kind(lhs, rhs)); goto failure; } if (PyFloat_CheckExact(lhs)) { - instr->op.code = COMPARE_OP_FLOAT; + specialized_op = COMPARE_OP_FLOAT; goto success; } if (PyLong_CheckExact(lhs)) { if (_PyLong_IsCompact((PyLongObject *)lhs) && _PyLong_IsCompact((PyLongObject *)rhs)) { - instr->op.code = COMPARE_OP_INT; + specialized_op = COMPARE_OP_INT; goto success; } else { @@ -2511,19 +2511,16 @@ _Py_Specialize_CompareOp(_PyStackRef lhs_st, _PyStackRef rhs_st, _Py_CODEUNIT *i goto failure; } else { - instr->op.code = COMPARE_OP_STR; + specialized_op = COMPARE_OP_STR; goto success; } } SPECIALIZATION_FAIL(COMPARE_OP, compare_op_fail_kind(lhs, rhs)); failure: - STAT_INC(COMPARE_OP, failure); - instr->op.code = COMPARE_OP; - cache->counter = adaptive_counter_backoff(cache->counter); + unspecialize(instr); return; success: - STAT_INC(COMPARE_OP, success); - cache->counter = adaptive_counter_cooldown(); + specialize(instr, specialized_op); } #ifdef Py_STATS

1 0