Add a test to check the current MAGIC_NUMBER against the
expected number for the release if the current release is
at candidate or final level. On test failure, describe to
the developer the procedure for changing the magic number.

This ensures that pre-merge CI will automatically pick up
on magic number changes in maintenance releases (and
explain why those are problematic), rather than relying on
all core developers to be aware of the implications of
such changes.

PyNumber_Int() creates a new reference: need to decrement result
reference counter.

(cherry picked from commit d79c1d4a)

(cherry picked from commit dc980dfbcfce4695ccde056c3983160ba97b5a36)

The traceback no longer displayed for SystemExit raised in a callback registered by atexit..
(cherry picked from commit 3fd54d4a)

import json before using json module
(cherry picked from commit 1dbce04d)

(cherry-picked from commit 0afbabe2)

Also adds `python.exe-gdb.py` to `.gitignore`, it is created by OS X builds.

Also updates checked-in line endings in several files.

Mention that fnmatchcase does not call normcase, and fnmatch does.
(cherry picked from commit e5f6e86c)

* bpo-30614: testInitNonExistentFile() of test_bz2 leaks references

Extract the code of BZ2File_dealloc and create a new BZ2File_clear()
function. Call BZ2File_clear() in BZ2File_dealloc().

Define BZ2File_clear() as tp_clear.

Move the lock initialization before the "self->file =
PyObject_CallFunction" in BZ2File_init() and check the lock is not
created twice.

Call BZ2File_clear() in BZ2File_init() after the init of the lock
Co-Authored-By: Victor Stinner <victor.stinner@gmail.com>

* Create bz2module.c

Fix after the review of Victor Stinner

(cherry picked from commit 42e3acda)

On Windows, subprocess.Popen.communicate() now also ignore EINVAL
on stdin.write() if the child process is still running but closed the
pipe.
(cherry picked from commit d52aa313)

* Update config.{guess,sub} from gnu.org..
(cherry picked from commit 6f46683a)

* bpo-23787: Change sum() docstring from sequence to iterable

Original patch by Raymond Hettinger.

(cherry picked from commit 167e0fc2)

Also weakens the 'should this be run?' regex to allow all builds when .travis.yml changes.
(cherry picked from commit c53b13b2)

Metaprogramming a list of attributes was excessive, and made the code less readable and slower.

Backport of 5de3a641

Also on the short-lived `buildbot-custom` branch.
(cherry picked from commit d3bedf35)

Initially the macOS builds are allowed to fail until such time that they can be determined to be stable and not add an unacceptable amount of time to the overall Travis-passing process.
(cherry picked from commit 21c2dd7c)

(cherry picked from commit 0737ee20)

The "iterable iterable" phrasing created confusion between the term
reference and the parameter name.

This simplifies the phrasing to just use the parameter name
without linking directly to the term definition..
(cherry picked from commit 08e2f355)

(cherry picked from commit a5aa72ac)

(cherry picked from commit f34c6850)

(cherry picked from commit eba68e2c)

This fixes a faulty backport to the Python 2.7 branch only of http://bugs.python.org/issue28032. Details in the bpo-29766.

(cherry picked from commit 9616a82e)

* bpo-30414: multiprocesing.Queue._feed do not break from main loop on exc

Queue background running thread was not handling exceptions correctly.
Any exception occurred inside thread (putting unpickable object) cause
feeder to finish running. After that every message put into queue is
silently ignored.

* bpo-30414: multiprocesing.Queue._feed do not break from main loop on exc

Queue background running thread was not handling exceptions correctly.
Any exception occurred inside thread (putting unpickable object) cause
feeder to finish running. After that every message put into queue is
silently ignored..
(cherry picked from commit bc50f03d)

Ran the docstrings through spell checker, and fixed spelling issues.

(cherry picked from commit 81755471)