Using stdlib `runpy.run_module()` causes `AttributeError: 'AssertionRewritingHook' object has no attribute 'get_code'`

When using runpy's run_module function inside a pytest test and pointing to the very module of the test being run the execution will crash reporting AttributeError: 'AssertionRewritingHook' object has no attribute 'get_code'.

Minimum reproducible example:

# tests/dummy_test.py
import runpy


def test():
    runpy.run_module("tests.dummy_test")

# requirements.txt
pytest==6.2.4

System info:

• OS: Ubuntu 20.04.2 LTS
• Python Version: 3.8.10

Additional observations:

• Calling an external function from other file in the test which calls runpy::run_module using the test's module as parameter will still yield the same error.
• Calling runpy::run_module passing any module other than the test's very own module will yield no error.
• Using runpy::run_path instead of run_module, passing the test's very own file path, will yield no error.

Start with the minimal reproduction in tests/dummy_test.py and run it with pytest==6.2.4 on Python 3.8.10. Trace the import path used by runpy.run_module for the test's own module and the AssertionRewritingHook, then verify that the reproduction completes without the get_code AttributeError while runpy.run_path remains unaffected.