Auditory Illusion Benchmark for Large Audio Language Models