Do not spawn ProcessPool workers on demand when they spawn via fork.
This avoids potential deadlocks in the child processes due to forking from
a multithreaded process..
(cherry picked from commit ebb37fc3fd)
Co-authored-by: Gregory P. Smith <greg@krypto.org>
* Remove redundant footnote ref: the footnote has been removed
* Fix footnote ref to match footnote
* Convert footnotes into reST footnotes: will error if missing
(cherry picked from commit 788ef54bc9)
Co-authored-by: Hugo van Kemenade <hugovk@users.noreply.github.com>
Automerge-Triggered-By: GH:serhiy-storchaka
Given that 2.7 has now been end-of-life for two and a half years,
I don't think we need such a detailed explanation here anymore of
the differences between Python 2 and Python 3.
(cherry picked from commit 8efda1e7c6)
Co-authored-by: Alex Waygood <Alex.Waygood@Gmail.com>
Automerge-Triggered-By: GH:serhiy-storchaka
The current wording implied this, but didn't state it explicitly.
(cherry picked from commit 740da8d37a)
Co-authored-by: Matt Wozniski <godlygeek@gmail.com>
merge_freemem(): set keys to NULL do it's harmless to call this again.
(cherry picked from commit 9652900969)
Co-authored-by: Tim Peters <tim.peters@gmail.com>
Fix a crash in subinterpreters related to the garbage collector. When
a subinterpreter is deleted, untrack all objects tracked by its GC.
To prevent a crash in deallocator functions expecting objects to be
tracked by the GC, leak a strong reference to these objects on
purpose, so they are never deleted and their deallocator functions
are not called.
(cherry picked from commit 14243369b5)
Co-authored-by: Victor Stinner <vstinner@python.org>
The function fetch_server_certificate is replaced by get_server_certificate in the module. I reflected the change in the module docstrings.
Co-authored-by: Jelle Zijlstra <jelle.zijlstra@gmail.com>
(cherry picked from commit feca9bbd1f)
Co-authored-by: Kossi GLOKPOR <83467320+glk0@users.noreply.github.com>
`IPv*Network` and `IPv*Interface` constructors accept a 2-tuple of
(address description, netmask) as the address parameter.
When the tuple-based address is used errors are not propagated
correctly through the `ipaddress.ip_*` helper because of the %-formatting now expecting several arguments:
In [7]: ipaddress.ip_network(("192.168.100.0", "fooo"))
...
TypeError: not all arguments converted during string formatting
Compared to:
In [8]: ipaddress.IPv4Network(("192.168.100.0", "foo"))
...
NetmaskValueError: 'foo' is not a valid netmask
Use an f-string to make sure the error is always properly formatted.
Co-authored-by: Jelle Zijlstra <jelle.zijlstra@gmail.com>
(cherry picked from commit 52dc9c3066)
Co-authored-by: Thomas Cellerier <thomascellerier@gmail.com>
Do not store `ProcessPoolExecutor` work item exception traceback that prevents
exception frame locals from being garbage collected.
(cherry picked from commit 9c204b148f)
Co-authored-by: themylogin <themylogin@gmail.com>
Fixed some inconsistencies in the text about relationships
(cherry picked from commit 5f45a9d3c3)
Co-authored-by: David Gilbertson <gilbertson.david@gmail.com>
Remove a confusion for read method in asyncio-subprocess doc for stderr StreamReader instance
(cherry picked from commit bb857a96ef)
Co-authored-by: Harsh <65716674+Harsh-br0@users.noreply.github.com>
If the error handler returns position less or equal than the starting
position of non-encodable characters, most of built-in encoders didn't
properly re-size the output buffer. This led to out-of-bounds writes,
and segfaults.
(cherry picked from commit 18b07d773e)
Co-authored-by: Serhiy Storchaka <storchaka@gmail.com>
This does not alter the `_posixsubprocess.fork_exec()` private API to
avoid issues for anyone relying on that (bad idea) or for anyone who's
`subprocess.py` and `_posixsubprocess.so` upgrades may not become
visible to existing Python 3.10 processes at the same time.
Backports the concept of cd5726fe67.
Provides a fail-safe way to disable vfork for #91401.
I didn't backport the documentation as I don't actually expect this to be used and `.. versionadded: 3.10.5` always looks weird in docs. It's being done more to have a fail-safe in place for people just in case.
