fix: return test runner exit code

[moonluna50]

Issue

The run_all_tests.py test runner called sys.exit() before returning the pytest exit code. This made the remaining code unreachable and caused the unit test for the runner's exit-code behavior to fail.

Approach

Simplified run_tests() to:

• Build the pytest command from the existing TEST_PATHS constant.
• Execute pytest using the current Python interpreter.
• Return result.returncode to the caller.
• Preserve the existing SystemExit behavior through the __main__ entry point.

This also removes duplicated and unreachable test-runner logic.

Testing

Ran the complete production-ready test suite:

• 10 tests passed
• 0 tests failed

The previously failing test_runner_uses_current_python_interpreter test now passes.

Also ran git diff --check successfully.

Considerations / Limitations

This change only fixes the test runner's exit-code handling. It does not modify the underlying eBoot functionality or production bootloader behavior.