cpython-lru_cache-weakref: Python/pythonrun.c history

Closes #19831: Stop tracemalloc later at Python shutdown to be able to use

Sun, 01 Dec 2013 10:03:26 +0100

Closes #19831: Stop tracemalloc later at Python shutdown to be able to use
tracemalloc in objects destructor

Replace atexit handler with an harcoded C function _PyTraceMalloc_Fini().

Issue #18874: Implement the PEP 454 (tracemalloc)

Sat, 23 Nov 2013 12:27:24 +0100

Issue #18874: Implement the PEP 454 (tracemalloc)

ssue #19183: Implement PEP 456 'secure and interchangeable hash algorithm'.

Wed, 20 Nov 2013 11:46:18 +0100

ssue #19183: Implement PEP 456 'secure and interchangeable hash algorithm'.
Python now uses SipHash24 on all major platforms.

(Merge 3.3) pythonrun.c: fix Py_GetPythonHome(), use Py_ARRAY_LENGTH() to get

Fri, 15 Nov 2013 17:12:14 +0100

(Merge 3.3) pythonrun.c: fix Py_GetPythonHome(), use Py_ARRAY_LENGTH() to get
the size of the env_home buffer, not PATH_MAX+1. env_home is declared using
MAXPATHLEN+1, and PATH_MAX is not declared on IRIX.

Issue #19515: Remove identifiers duplicated in the same file.

Tue, 12 Nov 2013 21:39:02 +0100

Issue #19515: Remove identifiers duplicated in the same file.
Patch written by Andrei Dorian Duma.

Issue #19466: Fix typo. Patch written by Vajrasky Kok.

Tue, 12 Nov 2013 17:18:51 +0100

Issue #19466: Fix typo. Patch written by Vajrasky Kok.

Close #19466: Clear the frames of daemon threads earlier during the Python

Tue, 12 Nov 2013 16:37:55 +0100

Close #19466: Clear the frames of daemon threads earlier during the Python
shutdown to call objects destructors. So "unclosed file" resource warnings are
now corretly emitted for daemon threads.

Issue #19512, #19515: remove shared identifiers, move identifiers where they

Thu, 07 Nov 2013 23:07:29 +0100

Issue #19512, #19515: remove shared identifiers, move identifiers where they
are used.

Move also _Py_IDENTIFIER() defintions to the top in modified files to remove
identifiers duplicated in the same file.

print_error_text() doesn't encode the filename anymore

Thu, 07 Nov 2013 12:37:56 +0100

print_error_text() doesn't encode the filename anymore

Use aslo PyUnicode_FromFormat() to format the line so only one call to
PyFile_WriteObject() is needed. tb_displayline() of Python/traceback.c has
similar implementation.

Issue #19512: add _PyUnicode_CompareWithId() function

Thu, 07 Nov 2013 00:46:04 +0100

Issue #19512: add _PyUnicode_CompareWithId() function

_PyUnicode_CompareWithId() is faster than PyUnicode_CompareWithASCIIString()
when both strings are equal and interned.

Add also _PyId_builtins identifier for "builtins" common string.

print_exception(): don't encode the module name to UTF-8

Thu, 07 Nov 2013 00:12:30 +0100

print_exception(): don't encode the module name to UTF-8

Replace _PyUnicode_AsString()+strcmp() with PyUnicode_CompareWithASCIIString().

Issue #19512: add some common identifiers to only create common strings once,

Wed, 06 Nov 2013 22:41:44 +0100

Issue #19512: add some common identifiers to only create common strings once,
instead of creating temporary Unicode string objects

Add also more identifiers in pythonrun.c to avoid temporary Unicode string
objets for the interactive interpreter.

Issue #19512: Add PyRun_InteractiveOneObject() function

Wed, 06 Nov 2013 18:41:07 +0100

Issue #19512: Add PyRun_InteractiveOneObject() function

Only decode the filename once. PyRun_InteractiveOneObject() uses an identifier
for "" string, so the byte string is only decoded once.

Issue #19512: _print_total_refs() now uses an identifier to get "showrefcount"

Wed, 06 Nov 2013 18:28:21 +0100

Issue #19512: _print_total_refs() now uses an identifier to get "showrefcount"
key from sys._xoptions

Issue #18603: Ensure that PyOS_mystricmp and PyOS_mystrnicmp are in the

Tue, 22 Oct 2013 10:23:12 +0200

Issue #18603: Ensure that PyOS_mystricmp and PyOS_mystrnicmp are in the
Python executable and not removed by the linker's optimizer.

Issue #16129: Py_SetStandardStreamEncoding cleanups

Fri, 18 Oct 2013 23:11:47 +1000

Issue #16129: Py_SetStandardStreamEncoding cleanups

- don't call PyErr_NoMemory with interpreter is not initialised
- note that it's OK to call _PyMem_RawStrDup here
- don't include this in the limited API
- capitalise "IO"
- be explicit that a non-zero return indicates an error
- include versionadded marker in docs

Issue #16129: Add `Py_SetStandardStreamEncoding`

Thu, 17 Oct 2013 22:35:35 +1000

Issue #16129: Add `Py_SetStandardStreamEncoding`

This new pre-initialization API allows embedding
applications like Blender to force a particular
encoding and error handler for the standard IO streams.

Also refactors Modules/_testembed.c to let us start
testing multiple embedding scenarios.

(Initial patch by Bastien Montagne)

Issue #18776: atexit callbacks now display their full traceback when they raise an exception.

Sun, 13 Oct 2013 21:54:15 +0200

Issue #18776: atexit callbacks now display their full traceback when they raise an exception.

Issue #18818: The "encodingname" part of PYTHONIOENCODING is now optional.

Fri, 13 Sep 2013 11:46:24 +0300

Issue #18818: The "encodingname" part of PYTHONIOENCODING is now optional.

Issue #18756: os.urandom() now uses a lazily-opened persistent file descriptor, so as to avoid using many file descriptors when run in parallel from multiple threads.

Sat, 31 Aug 2013 00:26:02 +0200

Issue #18756: os.urandom() now uses a lazily-opened persistent file descriptor, so as to avoid using many file descriptors when run in parallel from multiple threads.

pythonrun.c: use MAXPATHLEN instead of PATH_MAX

Wed, 28 Aug 2013 01:47:46 +0200

pythonrun.c: use MAXPATHLEN instead of PATH_MAX

PATH_MAX is not available on "MIPS IRIX 6.5.30 [SB] 3.x" buildbot

Issue #18571: Implementation of the PEP 446: file descriptors and file handles

Wed, 28 Aug 2013 00:53:59 +0200

Issue #18571: Implementation of the PEP 446: file descriptors and file handles
are now created non-inheritable; add functions os.get/set_inheritable(),
os.get/set_handle_inheritable() and socket.socket.get/set_inheritable().

Close #11619: The parser and the import machinery do not encode Unicode

Mon, 26 Aug 2013 22:28:21 +0200

Close #11619: The parser and the import machinery do not encode Unicode
filenames anymore on Windows.

Issue #18808: Non-daemon threads are now automatically joined when a sub-interpreter is shutdown (it would previously dump a fatal error).

Sun, 25 Aug 2013 19:48:18 +0200

Issue #18808: Non-daemon threads are now automatically joined when a sub-interpreter is shutdown (it would previously dump a fatal error).

Issue #16400: Add command line option for isolated mode.

Sat, 10 Aug 2013 16:36:18 +0200

Issue #16400: Add command line option for isolated mode.

-I

Run Python in isolated mode. This also implies -E and -s. In isolated mode
sys.path contains neither the script’s directory nor the user’s
site-packages directory. All PYTHON* environment variables are ignored,
too. Further restrictions may be imposed to prevent the user from
injecting malicious code.

Fix style

Tue, 30 Jul 2013 21:01:23 +0200

Fix style

Issue #18520: Fix initstdio(), handle PySys_SetObject() failure

Mon, 22 Jul 2013 23:55:19 +0200

Issue #18520: Fix initstdio(), handle PySys_SetObject() failure

Issue #18520: initsite() is a little bit more verbose when import site fails

Mon, 22 Jul 2013 22:53:28 +0200

Issue #18520: initsite() is a little bit more verbose when import site fails

Issue #18520: Add a new PyStructSequence_InitType2() function, same than

Mon, 22 Jul 2013 22:24:54 +0200

Issue #18520: Add a new PyStructSequence_InitType2() function, same than
PyStructSequence_InitType() except that it has a return value (0 on success,
-1 on error).

* PyStructSequence_InitType2() now raises MemoryError on memory allocation failure
* Fix also some calls to PyDict_SetItemString(): handle error

Issue #18520: Fix initsigs(), handle PyOS_InitInterrupts() error

Sun, 21 Jul 2013 13:25:51 +0200

Issue #18520: Fix initsigs(), handle PyOS_InitInterrupts() error

PyOS_InitInterrupts() can raise error when importing the signal module

Close #18469: Replace PyDict_GetItemString() with _PyDict_GetItemId() in structseq.c

Wed, 17 Jul 2013 01:22:45 +0200

Close #18469: Replace PyDict_GetItemString() with _PyDict_GetItemId() in structseq.c

_PyDict_GetItemId() is more efficient: it only builds the Unicode string once.
Identifiers (dictionary keys) are now created at Python initialization, and if
the creation failed, Python does exit with a fatal error.

Before, PyDict_GetItemString() failure was not handled: structseq_new() could
call PyObject_GC_NewVar() with a negative size, and structseq_dealloc() could
also crash.

Issue #18203: Add _PyMem_RawStrdup() and _PyMem_Strdup()

Sun, 07 Jul 2013 23:30:24 +0200

Issue #18203: Add _PyMem_RawStrdup() and _PyMem_Strdup()

Replace strdup() with _PyMem_RawStrdup() or _PyMem_Strdup(), depending if the
GIL is held or not.

Issue #18203: Fix Py_Finalize(): destroy the GIL after the last call to

Sun, 07 Jul 2013 15:50:49 +0200

Issue #18203: Fix Py_Finalize(): destroy the GIL after the last call to
PyMem_Malloc() or PyObject_Malloc().

For example, PyCFunction_Fini() calls PyObject_GC_Del() which calls
PyObject_FREE().

Issue #1545463: Global variables caught in reference cycles are now garbage-collected at shutdown.

Mon, 06 May 2013 21:15:57 +0200

Issue #1545463: Global variables caught in reference cycles are now garbage-collected at shutdown.

Fix crash caused by 8c1385205a35

Sun, 05 May 2013 08:14:53 +0200

Fix crash caused by 8c1385205a35
(thanks Arfrever for reporting).

#17115: I hate you MS for not supporting C99.

Sat, 04 May 2013 17:37:09 -0400

#17115: I hate you MS for not supporting C99.

Issue #17408: Avoid using an obsolete instance of the copyreg module when the interpreter is shutdown and then started again.

Sat, 04 May 2013 20:46:19 +0200

Issue #17408: Avoid using an obsolete instance of the copyreg module when the interpreter is shutdown and then started again.

#17115,17116: Have modules initialize the package and loader

Sat, 04 May 2013 13:56:58 -0400

#17115,17116: Have modules initialize the __package__ and __loader__
attributes to None.

The long-term goal is for people to be able to rely on these
attributes existing and checking for None to see if they have been
set. Since import itself sets these attributes when a loader does not
the only instances when the attributes are None are from someone
overloading __import__() and not using a loader or someone creating a
module from scratch.

This patch also unifies module initialization. Before you could have
different attributes with default values depending on how the module
object was created. Now the only way to not get the same default set
of attributes is to circumvent initialization by calling
ModuleType.__new__() directly.

merge 3.3 (#17863)

Mon, 29 Apr 2013 10:23:31 -0400

merge 3.3 (#17863)

Issue #17832: fix a compilation warning about a function prototype.

Wed, 24 Apr 2013 20:17:53 +0200

Issue #17832: fix a compilation warning about a function prototype.
Also, make the private function static.

#17323: The "[X refs, Y blocks]" printed by debug builds has been disabled by default. It can be re-enabled with the `-X showrefcount` option.

Tue, 26 Mar 2013 01:59:56 +0200

#17323: The "[X refs, Y blocks]" printed by debug builds has been disabled by default. It can be re-enabled with the `-X showrefcount` option.

Issue #13390: New function :func:`sys.getallocatedblocks()` returns the number of memory blocks currently allocated.

Sun, 09 Dec 2012 14:28:26 +0100

Issue #13390: New function :func:`sys.getallocatedblocks()` returns the number of memory blocks currently allocated.
Also, the ``-R`` option to regrtest uses this function to guard against memory allocation leaks.

pythonrun.c: fix Py_GetPythonHome(), use Py_ARRAY_LENGTH() to get the size of

Fri, 15 Nov 2013 17:09:24 +0100

pythonrun.c: fix Py_GetPythonHome(), use Py_ARRAY_LENGTH() to get the size of
the env_home buffer, not PATH_MAX+1. env_home is declared using MAXPATHLEN+1,
and PATH_MAX is not declared on IRIX.

Issue #18603: Ensure that PyOS_mystricmp and PyOS_mystrnicmp are in the

Tue, 22 Oct 2013 10:22:29 +0200

Issue #18603: Ensure that PyOS_mystricmp and PyOS_mystrnicmp are in the
Python executable and not removed by the linker's optimizer.

Issue #18776: atexit callbacks now display their full traceback when they raise an exception.

Sun, 13 Oct 2013 21:53:13 +0200

Issue #18776: atexit callbacks now display their full traceback when they raise an exception.

Fix crash caused by 8c1385205a35

Sun, 05 May 2013 08:12:42 +0200

Fix crash caused by 8c1385205a35
(thanks Arfrever for reporting).

Issue #17408: Avoid using an obsolete instance of the copyreg module when the interpreter is shutdown and then started again.

Sat, 04 May 2013 20:45:02 +0200

Issue #17408: Avoid using an obsolete instance of the copyreg module when the interpreter is shutdown and then started again.

ignore errors when trying to fetch sys.stdin.encoding (closes #17863)

Mon, 29 Apr 2013 10:23:08 -0400

ignore errors when trying to fetch sys.stdin.encoding (closes #17863)

Issue #15001: fix segfault on "del sys.module['main']"

Wed, 07 Nov 2012 09:07:22 +0100

Issue #15001: fix segfault on "del sys.module['__main__']"

Patch by Victor Stinner.

Issue #16218: Support non ascii characters in python launcher.

Thu, 01 Nov 2012 14:51:14 +0200

Issue #16218: Support non ascii characters in python launcher.

Patch by Serhiy Storchaka.

Issue #15895: my analysis was slightly off. The FILE pointer is only leaked when set_main_loader() fails for a pyc file with closeit=0. In the success case run_pyc_file() does its own cleanup of the fp. I've changed the code to use another FILE ptr for pyc files and moved the fclose() to PyRun_SimpleFileExFlags() to make it more obvious what's happening.

Tue, 11 Sep 2012 15:47:28 +0200

Issue #15895: my analysis was slightly off. The FILE pointer is only leaked when set_main_loader() fails for a pyc file with closeit=0. In the success case run_pyc_file() does its own cleanup of the fp. I've changed the code to use another FILE ptr for pyc files and moved the fclose() to PyRun_SimpleFileExFlags() to make it more obvious what's happening.

Issue #15895: Fix FILE pointer leak in PyRun_SimpleFileExFlags() when filename points to a pyc/pyo file and closeit is false.

Tue, 11 Sep 2012 14:11:03 +0200

Issue #15895: Fix FILE pointer leak in PyRun_SimpleFileExFlags() when filename points to a pyc/pyo file and closeit is false.

Null merge (fix already applied to the default branch)

Sat, 04 Aug 2012 01:38:16 +0200

Null merge (fix already applied to the default branch)

Close #13119: use "\r\n" newline for sys.stdout/err on Windows

Sat, 04 Aug 2012 01:28:00 +0200

Close #13119: use "\r\n" newline for sys.stdout/err on Windows

sys.stdout and sys.stderr are now using "\r\n" newline on Windows, as Python 2.

Fix initialization of the faulthandler module

Tue, 31 Jul 2012 02:55:49 +0200

Fix initialization of the faulthandler module

faulthandler requires the importlib if "-X faulthandler" option is present on
the command line, so initialize faulthandler after importlib.

Add also an unit test.

Refcounting fixes

Sun, 15 Jul 2012 23:21:08 +1000

Refcounting fixes

Make set_main_loader static (noticed by Antoine Pitrou)

Sun, 15 Jul 2012 23:18:08 +1000

Make set_main_loader static (noticed by Antoine Pitrou)

Actually initialize main.loader with loader instances, not the corresponding type objects

Sun, 15 Jul 2012 19:10:39 +1000

Actually initialize __main__.__loader__ with loader instances, not the corresponding type objects

Take the first step in resolving the messy pkgutil vs importlib edge cases by basing pkgutil explicitly on importlib, deprecating its internal import emulation and setting main.loader correctly so that runpy still works (Affects #15343, #15314, #15357)

Sun, 15 Jul 2012 18:09:52 +1000

Take the first step in resolving the messy pkgutil vs importlib edge cases by basing pkgutil explicitly on importlib, deprecating its internal import emulation and setting __main__.__loader__ correctly so that runpy still works (Affects #15343, #15314, #15357)

Issue #15020: The program name used to search for Python's path is now python3 under Unix, not python.

Thu, 05 Jul 2012 20:57:33 +0200

Issue #15020: The program name used to search for Python's path is now python3 under Unix, not python.