Fix Issue15701 - HTTPError info method call raises AttributeError. Fix that to return headers correctly

Patch by Serhiy Storchaka.

Thanks to Inada Naoki

Patch by Serhiy Storchaka.

Fix issue #16646: ftplib.FTP.makeport() might lose socket error details.  (patch by Serhiy Storchaka)

Patch by Ashish Nitin Patil.

Issue #16626: Fix infinite recursion in glob.glob() on Windows when the pattern contains a wildcard in the drive or UNC path.
Patch by Serhiy Storchaka.

Patch by Sebastian Kreft.

Issue #16298: In HTTPResponse.read(), close the socket when there is no Content-Length and the incoming stream is finished.
Patch by Eran Rundstein.

compiler logic will do the right thing with just x as a Py_uhash_t.  This
matches what was already done in the 3.3 version.

cleanup only - no functionality or hash values change.

computation as the overflow behavior of signed integers is undefined.

In practice we require compiling everything with -fwrapv which forces overflow
to be defined as twos compliment but this keeps the code cleaner for checkers
or in the case where someone has compiled it without -fwrapv or their
compiler's equivalent.

Found by Clang trunk's Undefined Behavior Sanitizer (UBSan).

Cleanup only - no functionality or hash values change.

Found by Clang trunk's Undefined-Behavior Sanitizer.  [more to come]

Turns out, the snakebite bots behave also their peculiarities.

I'm really not proud of this stream of commits. :(

The Windows buildbots seem to be unable to agree whether they need them or not.

It cause shutil.rmtree not ignore all errors. Also add a test ensuring that
rmtree fails when being called on a symlink. Patch by Serhiy Storchaka.

This commit updates a test broken by the change made for issue #14783.

Issue #16248: Disable code execution from the user's home directory by tkinter when the -E flag is passed to Python.
Patch by Zachary Ware.

Issue #16248: Disable code execution from the user's home directory by tkinter when the -E flag is passed to Python.
Patch by Zachary Ware.

Original patch by Julien Courteau and Pierre Paul Lefebvre.

Issue #16602: When a weakref's target was part of a long deallocation chain, the object could remain reachable through its weakref even though its refcount had dropped to zero.

Thanks to Eugene Toder for diagnosing and reporting the issue.