gh-106168: Update the size only after setting the item, to avoid temporary inconsistencies.
Also remove the "what's new" sentence regarding the size setting since tuples cannot grow after allocation.
* Add mimalloc v2.12
Modified src/alloc.c to remove include of alloc-override.c and not
compile new handler.
Did not include the following files:
- include/mimalloc-new-delete.h
- include/mimalloc-override.h
- src/alloc-override-osx.c
- src/alloc-override.c
- src/static.c
- src/region.c
mimalloc is thread safe and shares a single heap across all runtimes,
therefore finalization and getting global allocated blocks across all
runtimes is different.
* mimalloc: minimal changes for use in Python:
- remove debug spam for freeing large allocations
- use same bytes (0xDD) for freed allocations in CPython and mimalloc
This is important for the test_capi debug memory tests
* Don't export mimalloc symbol in libpython.
* Enable mimalloc as Python allocator option.
* Add mimalloc MIT license.
* Log mimalloc in Lib/test/pythoninfo.py.
* Document new mimalloc support.
* Use macro defs for exports as done in:
https://github.com/python/cpython/pull/31164/
Co-authored-by: Sam Gross <colesbury@gmail.com>
Co-authored-by: Christian Heimes <christian@python.org>
Co-authored-by: Victor Stinner <vstinner@python.org>
* Try to fix asyncio.Server.wait_closed() again
I identified the condition that `wait_closed()` is intended
to wait for: the server is closed *and* there are no more
active connections.
When this condition first becomes true, `_wakeup()` is called
(either from `close()` or from `_detach()`) and it sets `_waiters`
to `None`. So we just check for `self._waiters is None`; if it's
not `None`, we know we have to wait, and do so.
A problem was that the new test introduced in 3.12 explicitly
tested that `wait_closed()` returns immediately when the server
is *not* closed but there are currently no active connections.
This was a mistake (probably a misunderstanding of the intended
semantics). I've fixed the test, and added a separate test that
checks exactly for this scenario.
I also fixed an oddity where in `_wakeup()` the result of the
waiter was set to the waiter itself. This result is not used
anywhere and I changed this to `None`, to avoid a GC cycle.
* Update Lib/asyncio/base_events.py
---------
Co-authored-by: Carol Willing <carolcode@willingconsulting.com>
Fix locale.LC_CTYPE documentation to no longer mention string.lower() et al. Those functions were removed in Python 3.0:
https://docs.python.org/2/library/string.html#deprecated-string-functions
Also, fix a comment in logging about locale-specific behavior of `str.lower()`.
Co-authored-by: Hugo van Kemenade <hugovk@users.noreply.github.com>
Add a guide for how to handle non-UTF-8 text encodings.
Link to that guide from the 'text_factory' docs.
Co-authored-by: Alex Waygood <Alex.Waygood@Gmail.com>
Co-authored-by: C.A.M. Gerlach <CAM.Gerlach@Gerlach.CAM>
Co-authored-by: Corvin <corvin@corvin.dev>
Co-authored-by: Ezio Melotti <ezio.melotti@gmail.com>
Co-authored-by: Serhiy Storchaka <storchaka@gmail.com>
It now always returns an integer if one or less counting options are specified.
Previously it could return a single count as a 1-tuple, an integer (only if
option "update" was specified) or None if no items found.
The result is now the same if wantobjects is set to 0.
Add PyUnicode_AsUTF8() function to the limited C API.
multiprocessing posixshmem now uses PyUnicode_AsUTF8() instead of
PyUnicode_AsUTF8AndSize(): the extension is built with the limited C
API. The function now raises an exception if the filename contains an
embedded null character instead of truncating silently the filename.
* PyUnicode_AsUTF8() now raises an exception if the string contains
embedded null characters.
* Update related C API tests (test_capi.test_unicode).
* type_new_set_doc() uses PyUnicode_AsUTF8AndSize() to silently
truncate doc containing null bytes.
Co-authored-by: Serhiy Storchaka <storchaka@gmail.com>
Build the _multiprocessing.posixshmem extension with the Limited C
API.
* Add <errno.h> include.
* Replace PyUnicode_AsUTF8() with PyUnicode_AsUTF8AndSize().
* Replace PyStructSequence_SET_ITEM() with
PyStructSequence_SetItem().
* Replace PyTuple_GET_SIZE() with PyTuple_Size().
* Replace PyTuple_GET_ITEM() with PyTuple_GetItem().
The PySys_Audit() function was added in Python 3.8 by the PEP 578
"Python Runtime Audit Hooks".
Add also PySys_AuditTuple() to the limited C API, function added
to Python 3.13.
Move non-limited "PerfMap" C API from Include/sysmodule.h to
Include/cpython/sysmodule.h.
Add PyMem_RawMalloc(), PyMem_RawCalloc(), PyMem_RawRealloc() and
PyMem_RawFree() to the limited C API.
These functions were added by Python 3.4 and are needed to port
stdlib extensions to the limited C API, like grp and pwd.
Co-authored-by: Erlend E. Aasland <erlend@python.org>
Clarify the size of arenas
From 3.10.0 alpha 7, the pymalloc allocator uses arenas with a fixed size of 1
MiB on 64-bit platforms instead of 256 KiB on 32-bit platforms.
It already mostly worked, except in the case when invalid keyword
argument with non-ASCII name was passed to function with non-ASCII
parameter names. Then it crashed in the debug mode.
* gh-110803: Reorganize docs for what PyType_Slot doesn't cover
- Cover the offset fields first
- Mention the old alternative for MANAGED flags, which is needed
to support older Pythons
- De-emphasize the internal flags: use an inline list.
- Add a note to PyMemberDef saying what to do with it
* Remove an older draft...
If the Py_LIMITED_API macro is defined, Py_BUILD_CORE,
Py_BUILD_CORE_BUILTIN and Py_BUILD_CORE_MODULE macros are now
undefined by Python.h.
Only undefine these 3 macros after including "exports.h" which uses
them to define PyAPI_FUNC(), PyAPI_DATA() and PyMODINIT_FUNC macros.
Remove hacks (undefine manually the 3 Py_BUILD_CORE macros) in
Modules/_testcapi/parts.h and Modules/_testclinic_limited.c.
This commit removes a ':'. I believe the extra colon causes a display error.
What I believe to be an error:
Above this expression
`round(math.pi, ndigits=2) == round(22 / 7, ndigits=2)`
the page displays `.. doctest::`.
What I observed:
After I remove the extra colon, the page does not display `.. doctest::`
Add wrapper for timerfd_create, timerfd_settime, and timerfd_gettime to os module.
Co-authored-by: Serhiy Storchaka <storchaka@gmail.com>
Co-authored-by: Adam Turner <9087854+AA-Turner@users.noreply.github.com>
Co-authored-by: Erlend E. Aasland <erlend.aasland@protonmail.com>
Co-authored-by: Victor Stinner <vstinner@python.org>
sys.audit() now has assertions to check that the event argument is
not NULL and that the format argument does not use the "N" format.
Add tests on PySys_AuditTuple().
* Use `FindFirstFile` Win32 API to fix a bug where `ntpath.realpath()`
breaks out of traversing a series of paths where a (handled)
`ERROR_ACCESS_DENIED` or `ERROR_SHARING_VIOLATION` occurs.
* Update docs to reflect that `ntpath.realpath()` eliminates MS-DOS
style names.
Add PyThreadState_GetUnchecked() function: similar to
PyThreadState_Get(), but don't issue a fatal error if it is NULL. The
caller is responsible to check if the result is NULL. Previously,
this function was private and known as _PyThreadState_UncheckedGet().
* Enhanced sqlite3 connection context management documentation with contextlib.closing
* 📜🤖 Added by blurb_it.
* Fixed gitignore spelling error from nitignore to gitignore
* Renamed .gitignore to .nitignore
* Added generated doctests
* Deleted sqlite3 generated files
* Removed white-space changes
* Removed News entry from the doc
* Expanded a note that context manager can be used for connection management using contextlib.closing
* Removed repeated contextlib.closing code snippet
* Expanded the note around usage of context manageer for sqlite3 connection management
* Deleted extra white-spaces
* Deleted extra white-space
* re-arranged context manager wording
* Re-arranged word layout on how to use context manager
* Fix whitespace errors
* Remove unneeded change in .gitignore
* Added suggested changes
* Added suggested change redirecting to the contextlib.closing implementation
* Added closing keyword
* Removed line 2473
---------
Co-authored-by: blurb-it[bot] <43283697+blurb-it[bot]@users.noreply.github.com>
Co-authored-by: Erlend E. Aasland <erlend@python.org>
Make PyObject_VisitManagedDict() and PyObject_ClearManagedDict()
functions public in Python 3.13 C API.
* Rename _PyObject_VisitManagedDict() to PyObject_VisitManagedDict().
* Rename _PyObject_ClearManagedDict() to PyObject_ClearManagedDict().
* Document these functions.
If the timeout is greater than PY_TIMEOUT_MAX,
PyThread_acquire_lock_timed() uses a timeout of PY_TIMEOUT_MAX
microseconds, which is around 280.6 years. This case is unlikely and
limiting a timeout to 280.6 years sounds like a reasonable trade-off.
The constant PY_TIMEOUT_MAX is not used in PyPI top 5,000 projects.
* Add new installation path functions subsection
* Add content from install/index to sysconfig
* Fix table
* Update note about installers
* Clean up the list of schemes, remove references to Distutils
This method supports file URIs (including variants) as described in RFC 8089, such as URIs generated by `pathlib.Path.as_uri()` and `urllib.request.pathname2url()`.
The method is added to `Path` rather than `PurePath` because it uses `os.fsdecode()`, and so its results vary from system to system. I intend to deprecate `PurePath.as_uri()` and move it to `Path` for the same reason.
Co-authored-by: Adam Turner <9087854+AA-Turner@users.noreply.github.com>
* Refactor os_sched_getaffinity_impl(): move variable definitions to
their first assignment.
* Fix test_posix.test_sched_getaffinity(): restore the old CPU mask
when the test completes!
* Doc: Specify that os.cpu_count() counts *logicial* CPUs.
* Doc: Specify that os.sched_getaffinity(0) is related to the calling
thread.
* Remove unused <locale.h> includes.
* Remove unused <fcntl.h> include in traceback.h.
* Remove redundant <assert.h> and <stddef.h> includes. They are already
included by "Python.h".
* Remove <object.h> include in faulthandler.c. Python.h already includes it.
* Add missing <stdbool.h> in pycore_pythread.h if HAVE_PTHREAD_STUBS
is defined.
* Fix also warnings in pthread_stubs.h: don't redefine macros if they
are already defined, like the __NEED_pthread_t macro.
* pycore_pythread.h is now the central place to make sure that
_POSIX_THREADS and _POSIX_SEMAPHORES macros are defined if
available.
* Make sure that pycore_pythread.h is included when _POSIX_THREADS
and _POSIX_SEMAPHORES macros are tested.
* PY_TIMEOUT_MAX is now defined as a constant, since its value
depends on _POSIX_THREADS, instead of being defined as a macro.
* Prevent integer overflow in the preprocessor when computing
PY_TIMEOUT_MAX_VALUE on Windows:
replace "0xFFFFFFFELL * 1000 < LLONG_MAX"
with "0xFFFFFFFELL < LLONG_MAX / 1000".
* Document the change and give hints how to fix affected code.
* Add an exception for PY_TIMEOUT_MAX name to smelly.py
* Add PY_TIMEOUT_MAX to the stable ABI
As discussed in comments to #109544, the semantics of this attribute
are somewhat confusing. Add a note explaining its limitations and
steering users towards __required_keys__ and __optional_keys__ instead.
Remove replace_stdout(): call sys.stdout.reconfigure() instead of set
the error handler to backslashreplace.
display_header() logs an empty line and flush stdout.
Remove encoding workaround in display_header() since stdout error
handler is now set to backslashreplace earlier.
* Add --fast-ci and --slow-ci options to libregrtest:
* --fast-ci uses a default timeout of 10 minutes and "-u all,-cpu"
(skip slowest tests).
* --slow-ci uses a default timeout of 20 minues and "-u all" (run
all tests).
* regrtest header now lists test resources.
* Makefile changes:
* "make test", "make hostrunnertest" and "make coverage-report" now
use --fast-ci option and TESTTIMEOUT variable.
* "make buildbottest" now uses "--slow-ci". Remove options which
became redundant with "--slow-ci".
* "make testall" and "make testuniversal" now use --slow-ci option
and TESTTIMEOUT variable.
* "make testall" now uses "find -exec rm ..." instead of
"find ... -print|xargs rm ...", same as "make clean".
* GitHub Actions workflow:
* Ubuntu and Address Sanitizer jobs now use "make test". Remove
options which became redundant with "--fast-ci".
* Windows jobs now use --fast-ci option.
* Use -j0 to detect the number of CPUs.
* Set Makefile TESTTIMEOUT default to an empty string, since
--slow-ci and --fast-ci use different default timeout. It's now
accepted to pass "--timeout=" to regrtest: treated as not timeout.
* Tools/scripts/run_tests.py now uses --fast-ci option.
* Tools/buildbot/test.bat now uses --slow-ci option. Remove
--timeout=1200 option, redundant with --slow-ci.
A post main merge edit to the text was added in the 3.12 backport PR.
e38d7104b8
This includes that in main. It's a minor edit over #109767 to resolve the comment there.
Document the `os.fork` posix threads detected `DeprecationWarning` in 3.12 What's New, os, multiprocessing, and concurrent.futures docs.
Many reviews and doc cleanup edits by Adam & Hugo. 🥳
Co-authored-by: Adam Turner <9087854+AA-Turner@users.noreply.github.com>
Co-authored-by: Hugo van Kemenade <hugovk@users.noreply.github.com>
The previous layout made it look like the other three deprecations are part of the first one, when in fact they are independent.
The new layout is consistent with that used for sqlite3 in 3.12 (https://docs.python.org/3.13/whatsnew/3.12.html#deprecated).
SubprocessProtocol process_exited() method can be called before
pipe_data_received() and pipe_connection_lost() methods. Document it
and adapt the test for that.
Revert commit 282edd7b2a.
_child_watcher_callback() calls immediately _process_exited(): don't
add an additional delay with call_soon(). The reverted change didn't
make _process_exited() more determistic: it can still be called
before pipe_connection_lost() for example.
Co-authored-by: Davide Rizzo <sorcio@gmail.com>
Adds APIs to get the TLS certificate chains, verified or full unverified, from SSLSocket and SSLObject.
Co-authored-by: Gregory P. Smith [Google LLC] <greg@krypto.org>
This feature is off by default via code but on by default via the CLI. The `.gitignore` file contains `*` which causes the entire directory to be ignored.
Co-authored-by: Adam Turner <9087854+AA-Turner@users.noreply.github.com>
Co-authored-by: Hugo van Kemenade <hugovk@users.noreply.github.com>
Document PyMODINIT_FUNC macro.
Remove links to PyAPI_FUNC() and PyAPI_DATA() macros since they are
not documented. These macros should only be used to define the Python
C API. They should not be used outside Python code base.
BINARY_SUBSCR example erroneously uses two different names `key` and `index` to refer to the same variable. STORE_SUBSCR and DELETE_SUBSCR use only `key` in the same context. Changing `index` to `key` for consistency.
partev
Tian Gao
Nikita Sobolev
lefp
scoder
Irit Katriel
Terry Jan Reedy
LoipesMas
Victor Stinner
Yilei Yang
Serhiy Storchaka
Krzysiek Karbowiak
Thomas Grainger
Quentin Peter
Prometheus3375
xzmeng
Karolina Surma
Dino Viehland
Guido van Rossum
Zack Cerza
Jonathan Berthias
Łukasz Langa
Raymond Hettinger
Khalil Mouawad
Erlend E. Aasland
Hugo van Kemenade
Pavel Karateev
InSync
Zachary Ware
James Tocknell
Anthony Shaw
Don Patterson
Pablo Martí Gamboa
Nick
Maciej Olko
Gouvernathor
Unique-Usman
Adam Turner
Matthieu Dartiailh
Mienxiu
paskozdilar
Petr Viktorin
T. Wouters
Ezio Melotti
Radislav Chugunov
Donghee Na
Albert Villanova del Moral
zipperer
Masaru Tsuchiyama
Bradley Reynolds
Harmen Stoppels
Towster15
박문식
Benjamin Peterson
P. L. Lim
Jelle Zijlstra
Lincoln
Alex Waygood
Lele Gaifax
numbermaniac
Steve Dower
Charles Machalow
Ned Batchelder
Quentin Agren
Barney Gale
Jacob Coffee
Yuki K
Davide Rizzo
Sam Gross
lohaswinner
Antoine Pitrou
Gregory P. Smith
AN Long
Heinz-Alexander Fuetterer
James Gerity
Mateusz Nowak
Xuehai Pan
Anthony Sottile
AlberLC
Vinay Sajip
Brett Cannon
Egil Martinsson
Brandt Bucher
Oleksandr Kravets
Carl Meyer
wim glenn
Delgan
Shantanu
Christoph Anton Mitterer
Mark Dickinson
Stanley
Colin Watson