We tried this before with a dict and for all interned strings. That ran into problems due to interpreter isolation. However, exclusively using a per-interpreter cache caused some inconsistency that can eliminate the benefit of interning. Here we circle back to using a global cache, but only for statically allocated strings. We also use a more-basic _Py_hashtable_t for that global cache instead of a dict.
Ideally we would only have the global cache, but the optional isolation of each interpreter's allocator means that a non-static string object must not outlive its interpreter. Thus we would have to store a copy of each such interned string in the global cache, tied to the main interpreter.
Fix potential unaligned memory access on C APIs involving returned sequences
of `char *` pointers within the :mod:`grp` and :mod:`socket` modules. These
were revealed using a ``-fsaniziter=alignment`` build on ARM macOS.
Fix warnings C4100 in Py_UNUSED() when Python is built with "cl /W4".
Example with this function included by Python.h:
static inline unsigned int
PyUnicode_IS_READY(PyObject* Py_UNUSED(op))
{ return 1; }
Without this change, building a C program with "cl /W4" which just
includes Python.h emits the warning:
Include\cpython/unicodeobject.h(199):
warning C4100: '_unused_op': unreferenced formal parameter
This change fix this warning.
Adding the `rtype_cache` to the `warnings.warn` message improves the
previous, somewhat vague message from
```
/Users/username/cpython/Lib/multiprocessing/resource_tracker.py:224: UserWarning: resource_tracker: There appear to be 6 leaked semaphore objects to clean up at shutdown
```
to
```
/Users/username/cpython/Lib/multiprocessing/resource_tracker.py:224: UserWarning: resource_tracker: There appear to be 6 leaked semaphore objects to clean up at shutdown: {'/mp-yor5cvj8', '/mp-10jx8eqr', '/mp-eobsx9tt', '/mp-0lml23vl', '/mp-9dgtsa_m', '/mp-frntyv4s'}
```
---------
Co-authored-by: blurb-it[bot] <43283697+blurb-it[bot]@users.noreply.github.com>
test_logging: Fix test_udp_reconnection() by increasing the timeout
from 100 ms to 5 minutes (LONG_TIMEOUT).
Replace also blocking wait() with wait(LONG_TIMEOUT) in
test_output() to prevent the test to hang.
Use lowercase `mail from` and `rcpt to` in `smtplib.SMTP`
SMTP commands are case-insensitive. `smtplib` uses lowercase commands,
however it writes `mail FROM` and `rcpt TO`, lacking consistency.
There are 3 paths to use `locale` argument in
`calendar.Locale{Text|HTML}Calendar.__init__(..., locale=None)`:
(1) `locale=None` -- denotes the "default locale"[1]
(2) `locale=""` -- denotes the native environment
(3) `locale=other_valid_locale` -- denotes a custom locale
So far case (2) is covered and case (1) is in 78935daf5a (same branch).
This commit adds a remaining case (3).
[1] In the current implementation, this translates into the following
approach:
GET current locale
IF current locale == "C" THEN
SET current locale TO ""
GET current locale
ENDIF
* Remove unreachable code (and increase test coverage)
This condition cannot be true. `_locale.setlocale()` from the C module
raises `locale.Error` instead of returning `None` for
`different_locale.__enter__` (where `self.oldlocale` is set).
* Expand the try clause to calls to `LocaleTextCalendar.formatmonthname()`.
This method temporarily changes the current locale to the given locale,
so `_locale.setlocale()` may raise `local.Error`.
Co-authored-by: Rohit Mediratta <rohitm@gmail.com>
Co-authored-by: Jessica McKellar <jesstess@mit.edu>
Co-authored-by: Adam Turner <9087854+AA-Turner@users.noreply.github.com>
Co-authored-by: Irit Katriel <1055913+iritkatriel@users.noreply.github.com>
Remove private _PyUnicode_AsString() alias to PyUnicode_AsUTF8(). It
was kept for backward compatibility with Python 3.0 - 3.2.
The PyUnicode_AsUTF8() is available since Python
3.3. The PyUnicode_AsUTF8String() function can be used to keep
compatibility with Python 3.2 and older.
Add test for the 'destination <name> clear' command,
and the 'destination' directive in general.
Fix two bugs in 'destination <name> clear' command:
1. The text attribute of the allocator is called 'text', not '_text'
2. Return after processing the 'clear' command,
instead of proceeding directly to the fail().
* Add PyDict_GetItemRef() and PyDict_GetItemStringRef() functions.
Add these functions to the stable ABI version 3.13.
* Add unit tests on the PyDict C API in test_capi.
There was a slight race in _Py_ClearFileSystemEncoding() (when called from _Py_SetFileSystemEncoding()), between freeing the value and setting the variable to NULL, which occasionally caused crashes when multiple isolated interpreters were used. (Notably, I saw at least 10 different, seemingly unrelated spooky-action-at-a-distance, ways this crashed. Yay, free threading!) We avoid the problem by only setting the global variables with the main interpreter (i.e. runtime init).
A static (process-global) str object must only have its "interned" state cleared when no longer interned in any interpreters. They are the only ones that can be shared by interpreters so we don't have to worry about any other str objects.
We trigger clearing the state with the main interpreter, since no other interpreters may exist at that point and _PyUnicode_ClearInterned() is only called during interpreter finalization.
We do not address here the fact that a string will only be interned in the first interpreter that interns it. In any subsequent interpreters str.state.interned is already set so _PyUnicode_InternInPlace() will skip it. That needs to be addressed separately from fixing the crasher.
* gh-105293: Do not call SSL_CTX_set_session_id_context on client side SSL context
Openssl states this is a "server side only" operation.
Calling this on a client side socket can result in unexpected behavior
* Add news entry on SSL "set session id context" changes
The added file is the coverage default at some point in time + checking branches both ways + IDLE additions, labelled as such and somewhat designed to be unlikely to affect other files. Located in the CPython repository directory, it can be used where it is or copied elsewhere, depending on how one runs coverage.
---------
Co-authored-by: Terry Jan Reedy <tjreedy@udel.edu>
Co-authored-by: Alex Waygood <Alex.Waygood@Gmail.com>
Co-authored-by: Erlend E. Aasland <erlend.aasland@protonmail.com>
Fetch CONFIG_ARGS from the original source directory, instead of from
the copied source tree. When "make clean" is executed in the copied
source tree, the build directory is cleared and the configure argument
lookup fails. However, the original source directory still contains this
information.
gh-86618 assumed a-b-c = a-(b+c) = a-d where d = b+d.
For floats 2.0, 1.0, and 0.9999999999999999, this assumption
is false. The net change of 1.1102230246251565e-16 to 0.0
results in division by 0. Revert the replacement. Add test.
* Convert PyObject_DelAttr() and PyObject_DelAttrString() macros to
functions.
* Add PyObject_DelAttr() and PyObject_DelAttrString() functions to
the stable ABI.
* Replace PyObject_SetAttr(obj, name, NULL) with
PyObject_DelAttr(obj, name).
Detect email address parsing errors and return empty tuple to indicate the parsing error (old API). This fixes or at least ameliorates CVE-2023-27043.
---------
Co-authored-by: Gregory P. Smith <greg@krypto.org>
Fixes a hang in multiprocessing process pool executor when a child process crashes and code could otherwise block on writing to the pipe. See GH-94777 for more details.
Some items remained uninitialized if _sre.template() was called with invalid
indices. Then attempt to clear them in the destructor led to dereferencing
of uninitialized pointer.
Prevent `multiprocessing.spawn` from failing to *import* in environments
where `sys.executable` is `None`. This regressed in 3.11 with the addition
of support for path-like objects in multiprocessing.
Adds a test decorator to have tests only run when part of test_multiprocessing_spawn to `_test_multiprocessing.py` so we can start to avoid re-running the same not-global-state specific test in all 3 modes when there is no need.
We match paths using the `_lines` attribute, which is derived from the
path's string representation. The bug arises because an empty path's string
representation is `'.'` (not `''`), which is matched by the `'*'` wildcard.
mock: Add `ThreadingMock` class
Add a new class that allows to wait for a call to happen by using
`Event` objects. This mock class can be used to test and validate
expectations of multithreading code.
It uses two attributes for events to distinguish calls with any argument
and calls with specific arguments.
The calls with specific arguments need a lock to prevent two calls in
parallel from creating the same event twice.
The timeout is configured at class and constructor level to allow users
to set a timeout, we considered passing it as an argument to the
function but it could collide with a function parameter. Alternatively
we also considered passing it as positional only but from an API
caller perspective it was unclear what the first number meant on the
function call, think `mock.wait_until_called(1, "arg1", "arg2")`, where
1 is the timeout.
Lastly we also considered adding the new attributes to magic mock
directly rather than having a custom mock class for multi threading
scenarios, but we preferred to have specialised class that can be
composed if necessary. Additionally, having added it to `MagicMock`
directly would have resulted in `AsyncMock` having this logic, which
would not work as expected, since when if user "waits" on a
coroutine does not have the same meaning as waiting on a standard
call.
Co-authored-by: Karthikeyan Singaravelan <tir.karthi@gmail.com>
If the output arg to Helper() is a stream rather than the default None, which means 'page to stdout', the ImportError from pydoc.resolve is currently not caught in pydoc.doc. The same error is caught when output is None.
---------
Co-authored-by: Terry Jan Reedy <tjreedy@udel.edu>
Remove the "cpython/pytime.h" header file: it only contained private
functions. Move functions to the internal pycore_time.h header file.
Move tests from _testcapi to _testinternalcapi. Rename also test
methods to have the same name than tested C functions.
No longer export these functions:
* _PyTime_Add()
* _PyTime_As100Nanoseconds()
* _PyTime_FromMicrosecondsClamp()
* _PyTime_FromTimespec()
* _PyTime_FromTimeval()
* _PyTime_GetPerfCounterWithInfo()
* _PyTime_MulDiv()
Command Prompt (CMD Shell) and older versions of PowerShell
require double quotes and single quotes inside the string.
This form also works on linux and macOS.
* EOFError no longer overrides other errors such as MemoryError or OSError at
the start of the object.
* Raise more relevant error when the NULL object occurs as a code object
component.
* Minimize an overhead of calling PyErr_Occurred().
When running the Python test suite with -jN option, if a worker stdout
cannot be decoded from the locale encoding report a failed testn so the
exitcode is non-zero.
PyTuple_SET_ITEM() and PyList_SET_ITEM() now check the index argument
with an assertion if Python is built in debug mode or is built with
assertions.
* list_extend() and _PyList_AppendTakeRef() now set the list size
before calling PyList_SET_ITEM().
* PyStructSequence_GetItem() and PyStructSequence_SetItem() now check
the index argument: must be lesser than REAL_SIZE(op).
* PyStructSequence_GET_ITEM() and PyStructSequence_SET_ITEM() are now
aliases to PyStructSequence_GetItem() and
PyStructSequence_SetItem().
Added a new, experimental, tracing optimizer and interpreter (a.k.a. "tier 2"). This currently pessimizes, so don't use yet -- this is infrastructure so we can experiment with optimizing passes. To enable it, pass ``-Xuops`` or set ``PYTHONUOPS=1``. To get debug output, set ``PYTHONUOPSDEBUG=N`` where ``N`` is a debug level (0-4, where 0 is no debug output and 4 is excessively verbose).
All of this code is likely to change dramatically before the 3.13 feature freeze. But this is a first step.
Brings `pathlib.Path.is_dir()` and `in line with `os.DirEntry.is_dir()`, which
will be important for implementing generic path walking and globbing.
Likewise `is_file()`.
Remove old aliases which were kept backwards compatibility with
Python 3.8:
* _PyObject_CallMethodNoArgs()
* _PyObject_CallMethodOneArg()
* _PyObject_CallOneArg()
* _PyObject_FastCallDict()
* _PyObject_Vectorcall()
* _PyObject_VectorcallMethod()
* _PyVectorcall_Function()
Update code which used these aliases to use new names.
This new exception type is raised instead of `NotImplementedError` when
a path operation is not supported. It can be raised from `Path.readlink()`,
`symlink_to()`, `hardlink_to()`, `owner()` and `group()`. In a future
version of pathlib, it will be raised by `AbstractPath` for these methods
and others, such as `AbstractPath.mkdir()` and `unlink()`.
* Add tests on PyImport_AddModuleRef(), PyImport_AddModule() and
PyImport_AddModuleObject().
* pythonrun.c: Replace Py_XNewRef(PyImport_AddModule(name)) with
PyImport_AddModuleRef(name).
Eric Snow
Illia Volochii
Charlie Zhao
Felipe A. Hernandez
Christopher Chavez
Mark Shannon
Serhiy Storchaka
János Kukovecz
Pablo Galindo Salgado
Victor Stinner
shailshouryya
Paul Moore
Carl Meyer
Erlend E. Aasland
Dong-hee Na
Tomas R
htsedebenham
Gilles Bassière
J. Nick Koston
Lukas van de Wiel
Matthieu Caneill
Bart Skowron
James Hilton-Balfe
Gregory P. Smith
Brandt Bucher
Zach Brantmeier
Pieter Eendebak
Barney Gale
Yonatan Bitton
Tian Gao
Irit Katriel
Nikita Sobolev
Grigoriev Semyon
Kevin Diem
Jason R. Coombs
Inada Naoki
Grant Ramsay
Ammar Askar
Alex Waygood
Chris Brett
CF Bolz-Tereick
Jim Porter
Dennis Sweeney
Prince Roshan
Steve Dower
Ijtaba Hussain
Jelle Zijlstra
Terry Jan Reedy
Ethan Furman
Thomas Dwyer
Louis Paulot
Hugo van Kemenade
Radislav Chugunov
Desmond Cheong
Andrew Geng
Ariel Eizenberg
Lysandros Nikolaou
Mario Corchero
Kirill Podoprigora
Hood Chatham
hms
Guido van Rossum
Łukasz Langa
Gabriel Venberg
James Webber
YAMAMOTO Takashi
Crowthebird
T. Wouters
Erlend E. Aasland