Fix Race conditions in FileSystemBytecodeCache

[dvitek]

Fix race conditions in FileSystemBytecodeCache. See #1654 for a detailed description of the race conditions.

• fixes jinja2.FileSystemBytecodeCache race conditions #1654

Checklist:

• Add tests that demonstrate the correct behavior of the change. Tests should fail without the change.
• Add or update relevant docs, in the docs folder and in code.
• Add an entry in CHANGES.rst summarizing the change and linking to the issue.
• [n/a] Add .. versionchanged:: entries in any relevant code docs.
• Run pre-commit hooks and fix any issues.
• Run pytest and tox, no tests failed.

[davidism]

I'm going to remove the tests because in the past having tests that relied on race conditions was very flaky and caused confusion for contributors when unrelated PRs failed.

[dvitek]

ok

[dvitek]

FWIW, the test tests for the absence of a race, if that makes a difference. It should not fail intermittently, unless the bug is re-introduced.

[davidism]

The problem with that is that there no guarantee that a race happened and was handled vs normal behavior happened. And any such check would then rely on producing the race, which is flaky.