Very likely the issue is with the test infrastructure and not with the code under test. My own fuzzing doesn't report any inconsistencies. But fuzzing is alway

Very likely the issue is with the test infrastructure and not with the code under test. My own fuzzing doesn't report any inconsistencies. But fuzzing is always necessarily incomplete.