I expect the issue to be fixed based on https://github.com/google/oss-fuzz/pull/11708#issuecomment-2006442396 and https://github.com/actions/runner-images/issues/9491.
Victor Stinner [Wed, 27 Mar 2024 22:10:14 +0000 (23:10 +0100)]
gh-113317: Change how Argument Clinic lists converters (#116853)
* Add a new create_parser_namespace() function for
PythonParser to pass objects to executed code.
* In run_clinic(), list converters using 'converters' and
'return_converters' dictionarties.
* test_clinic: add 'object()' return converter.
* Use also create_parser_namespace() in eval_ast_expr().
Co-authored-by: Erlend E. Aasland <erlend@python.org>
Mark Shannon [Tue, 26 Mar 2024 11:11:42 +0000 (11:11 +0000)]
GH-117108: Set the "old space bit" to "visited" for all young objects (#117213)
Change old space bit of young objects from 0 to gcstate->visited_space.
This ensures that any object created *and* collected during cycle GC has the bit set correctly.
Python 3.10 changed from using SSL_write() and SSL_read() to SSL_write_ex() and
SSL_read_ex(), but did not update handling of the return value.
Change error handling so that the return value is not examined.
OSError (not EOF) is now returned when retval is 0.
According to *recent* man pages of all functions for which we call
PySSL_SetError, (in OpenSSL 3.0 and 1.1.1), their return value should
be used to determine whether an error happened (i.e. if PySSL_SetError
should be called), but not what kind of error happened (so,
PySSL_SetError shouldn't need retval). To get the error,
we need to use SSL_get_error.
Co-authored-by: Serhiy Storchaka <storchaka@gmail.com> Co-authored-by: Petr Viktorin <encukou@gmail.com>
gh-117187: Fix XML tests for vanilla Expat <2.6.0 (GH-117203)
This fixes XML unittest fallout from the https://github.com/python/cpython/issues/115398 security fix. When configured using `--with-system-expat` on systems with older pre 2.6.0 versions of libexpat, our unittests were failing.
* sax|etree: Simplify Expat version guard where simplifiable
Idea by Matěj Cepl
* sax|etree: Fix reparse deferral tests for vanilla Expat <2.6.0
This *does not fix* the case of distros with an older version of libexpat with the 2.6.0 feature backported as a security fix. (Ubuntu is a known example of this with its libexpat1 2.5.0-2ubunutu0.1 package)
Eric V. Smith [Mon, 25 Mar 2024 23:59:14 +0000 (19:59 -0400)]
gh-109870: Dataclasses: batch up exec calls (gh-110851)
Instead of calling `exec()` once for each function added to a dataclass, only call `exec()` once per dataclass. This can lead to speed improvements of up to 20%.
gh-112571: Move fish venv activation script into the common folder (GH-117169)
pythongh-112571: allow using fish venv activation script on windows
The fish shell can be used on windows under cygwin or msys2.
This change moves the script to the common folder
so the venv module will install it on both posix and nt systems (like the bash script).
Ken Jin [Sat, 23 Mar 2024 22:19:17 +0000 (06:19 +0800)]
gh-117180: Complete call sequence when trace stack overflow (GH-117184)
---------
Co-authored-by: Peter Lazorchak <lazorchakp@gmail.com> Co-authored-by: Guido van Rossum <gvanrossum@users.noreply.github.com> Co-authored-by: Guido van Rossum <gvanrossum@gmail.com>
Jakub Stasiak [Fri, 22 Mar 2024 16:49:56 +0000 (17:49 +0100)]
GH-113171: Fix "private" (non-global) IP address ranges (GH-113179)
* GH-113171: Fix "private" (really non-global) IP address ranges
The _private_networks variables, used by various is_private
implementations, were missing some ranges and at the same time had
overly strict ranges (where there are more specific ranges considered
globally reachable by the IANA registries).
This patch updates the ranges with what was missing or otherwise
incorrect.
I left 100.64.0.0/10 alone, for now, as it's been made special in [1]
and I'm not sure if we want to undo that as I don't quite understand the
motivation behind it.
The _address_exclude_many() call returns 8 networks for IPv4, 121
networks for IPv6.
Tim Peters [Fri, 22 Mar 2024 03:27:25 +0000 (22:27 -0500)]
GH-116939: Rewrite binarysort() (#116940)
Rewrote binarysort() for clarity.
Also changed the signature to be more coherent (it was mixing sortslice with raw pointers).
No change in method or functionality. However, I left some experiments in, disabled for now
via `#if` tricks. Since this code was first written, some kinds of comparisons have gotten
enormously faster (like for lists of floats), which changes the tradeoffs.
For example, plain insertion sort's simpler innermost loop and highly predictable branches
leave it very competitive (even beating, by a bit) binary insertion when comparisons are
very cheap, despite that it can do many more compares. And it wins big on runs that
are already sorted (moving the next one in takes only 1 compare then).
So I left code for a plain insertion sort, to make future experimenting easier.
Also made the maximum value of minrun a `#define` (``MAX_MINRUN`) to make
experimenting with that easier too.
And another bit of `#if``-disabled code rewrites binary insertion's innermost loop to
remove its unpredictable branch. Surprisingly, this doesn't really seem to help
overall. I'm unclear on why not. It certainly adds more instructions, but they're very
simple, and it's hard to be believe they cost as much as a branch miss.
Eric Snow [Fri, 22 Mar 2024 00:20:20 +0000 (18:20 -0600)]
gh-105716: Fix _PyInterpreterState_IsRunningMain() For Embedders (gh-117140)
When I added _PyInterpreterState_IsRunningMain() and friends last year, I tried to accommodate applications that embed Python but don't call _PyInterpreterState_SetRunningMain() (not that they're expected to). That mostly worked fine until my recent changes in gh-117049, where the subtleties with the fallback code led to failures; the change ended up breaking test_tools.test_freeze, which exercises a basic embedding situation.
The simplest fix is to drop the fallback code I originally added to _PyInterpreterState_IsRunningMain() (and later to _PyThreadState_IsRunningMain()). I've kept the fallback in the _xxsubinterpreters module though. I've also updated Py_FrozenMain() to call _PyInterpreterState_SetRunningMain().
Guido van Rossum [Thu, 21 Mar 2024 19:37:41 +0000 (12:37 -0700)]
gh-117045: Add code object to function version cache (#117028)
Changes to the function version cache:
- In addition to the function object, also store the code object,
and allow the latter to be retrieved even if the function has been evicted.
- Stop assigning new function versions after a critical attribute (e.g. `__code__`)
has been modified; the version is permanently reset to zero in this case.
- Changes to `__annotations__` are no longer considered critical. (This fixes gh-109998.)
Changes to the Tier 2 optimization machinery:
- If we cannot map a function version to a function, but it is still mapped to a code object,
we continue projecting the trace.
The operand of the `_PUSH_FRAME` and `_POP_FRAME` opcodes can be either NULL,
a function object, or a code object with the lowest bit set.
This allows us to trace through code that calls an ephemeral function,
i.e., a function that may not be alive when we are constructing the executor,
e.g. a generator expression or certain nested functions.
We will lose globals removal inside such functions,
but we can still do other peephole operations
(and even possibly [call inlining](https://github.com/python/cpython/pull/116290),
if we decide to do it), which only need the code object.
As before, if we cannot retrieve the code object from the cache, we stop projecting.
gh-116333: Relax error string text expectations in SSL-related tests (GH-116334)
* Relax error string text expectations in SSL-related tests
As suggested [here][1], this change relaxes the OpenSSL error string
text expectations in a number of tests. This was specifically done in
support of more easily building CPython [AWS-LC][2], but because AWS-LC
is a fork of [BoringSSL][3], it should increase compatibility with that
library as well.
In addition to the error string relaxations, we also add some guards
around the `tls-unique` channel binding being used with TLSv1.3, as that
feature (described in [RFC 6929][4]) is [not defined][5] for TLSv1.3.
Split `_PyThreadState_DeleteExcept` into two functions:
- `_PyThreadState_RemoveExcept` removes all thread states other than one
passed as an argument. It returns the removed thread states as a
linked list.
- `_PyThreadState_DeleteList` deletes those dead thread states. It may
call destructors, so we want to "start the world" before calling
`_PyThreadState_DeleteList` to avoid potential deadlocks.
Eric Snow [Thu, 21 Mar 2024 17:15:02 +0000 (11:15 -0600)]
gh-76785: Drop PyInterpreterID_Type (gh-117101)
I added it quite a while ago as a strategy for managing interpreter lifetimes relative to the PEP 554 (now 734) implementation. Relatively recently I refactored that implementation to no longer rely on InterpreterID objects. Thus now I'm removing it.
Victor Stinner [Thu, 21 Mar 2024 16:07:00 +0000 (17:07 +0100)]
gh-115754: Add Py_GetConstant() function (#116883)
Add Py_GetConstant() and Py_GetConstantBorrowed() functions.
In the limited C API version 3.13, getting Py_None, Py_False,
Py_True, Py_Ellipsis and Py_NotImplemented singletons is now
implemented as function calls at the stable ABI level to hide
implementation details. Getting these constants still return borrowed
references.
Add _testlimitedcapi/object.c and test_capi/test_object.py to test
Py_GetConstant() and Py_GetConstantBorrowed() functions.
Eric Snow [Thu, 21 Mar 2024 15:56:12 +0000 (09:56 -0600)]
gh-76785: Clean Up Interpreter ID Conversions (gh-117048)
Mostly we unify the two different implementations of the conversion code (from PyObject * to int64_t. We also drop the PyArg_ParseTuple()-style converter function, as well as rename and move PyInterpreterID_LookUp().
Sam Gross [Thu, 21 Mar 2024 14:01:16 +0000 (10:01 -0400)]
gh-116522: Stop the world before fork() and during shutdown (#116607)
This changes the free-threaded build to perform a stop-the-world pause
before deleting other thread states when forking and during shutdown.
This fixes some crashes when using multiprocessing and during shutdown
when running with `PYTHON_GIL=0`.
This also changes `PyOS_BeforeFork` to acquire the runtime lock
(i.e., `HEAD_LOCK(&_PyRuntime)`) before forking to ensure that data
protected by the runtime lock (and not just the GIL or stop-the-world)
is in a consistent state before forking.
Petr Viktorin [Wed, 20 Mar 2024 16:33:08 +0000 (17:33 +0100)]
gh-114314: ctypes: remove stgdict and switch to heap types (GH-116458)
Before this change, ctypes classes used a custom dict subclass, `StgDict`,
as their `tp_dict`. This acts like a regular dict but also includes extra information
about the type.
This replaces stgdict by `StgInfo`, a C struct on the type, accessed by
`PyObject_GetTypeData()` (PEP-697).
All usage of `StgDict` (mainly variables named `stgdict`, `dict`, `edict` etc.) is
converted to `StgInfo` (named `stginfo`, `info`, `einfo`, etc.).
Where the dict is actually used for class attributes (as a regular PyDict), it's now
called `attrdict`.
This change -- not overriding `tp_dict` -- is made to make me comfortable with
the next part of this PR: moving the initialization logic from `tp_new` to `tp_init`.
The `StgInfo` is set up in `__init__` of each class, with a guard that prevents
calling `__init__` more than once. Note that abstract classes (like `Array` or
`Structure`) are created using `PyType_FromMetaclass` and do not have
`__init__` called.
Previously, this was done in `__new__`, which also wasn't called for abstract
classes.
Since `__init__` can be called from Python code or skipped, there is a tested
guard to ensure `StgInfo` is initialized exactly once before it's used.
Co-authored-by: neonene <53406459+neonene@users.noreply.github.com> Co-authored-by: Erlend E. Aasland <erlend.aasland@protonmail.com>
Brett Simmers [Wed, 20 Mar 2024 15:18:26 +0000 (08:18 -0700)]
gh-116908: Only write to `_pending_calls.calls_to_do` with atomic operations (#117044)
These writes to `pending->calls_to_do` need to be atomic, because other threads
can read (atomically) from `calls_to_do` without holding `pending->mutex`.
Sam Gross [Tue, 19 Mar 2024 18:40:20 +0000 (14:40 -0400)]
gh-113964: Don't prevent new threads until all non-daemon threads exit (#116677)
Starting in Python 3.12, we prevented calling fork() and starting new threads
during interpreter finalization (shutdown). This has led to a number of
regressions and flaky tests. We should not prevent starting new threads
(or `fork()`) until all non-daemon threads exit and finalization starts in
earnest.
This changes the checks to use `_PyInterpreterState_GetFinalizing(interp)`,
which is set immediately before terminating non-daemon threads.
Serhiy Storchaka [Tue, 19 Mar 2024 17:26:32 +0000 (19:26 +0200)]
gh-90300: Improve the Python CLI help output (GH-115853)
* document equivalent command-line options for all environment variables
* document equivalent environment variables for all command-line options
* reduce the size of variable and option descriptions to minimum
* remove the ending period in single-sentence descriptions
Co-authored-by: Éric <merwok@netwok.org> Co-authored-by: Hugo van Kemenade <1324225+hugovk@users.noreply.github.com>