YA-JOP.Yet Another Java Object Profiler On Art Virtual Machine

摘要

In this paper, we design and implement an efficient and low overhead Java object profiler called YA-JOP on Android 6.0 and its ART virtual machine, which uses an AOT (ahead-of-time) compiler. YA-JOP records the allocation site, the class information, the object size, the birth time and death time, the last access time and the access regular pattern for every Java object. The data profiled can help the developers to detect memory leaks, find reusable objects, implement optimizations like pretenuring, etc. The profiler proposed in this paper has reasonable execution time overhead, imposes no overhead on the Java heap and does not modify any existing key data structure of the ART Virtual Machine, including the object layouts, class layouts and any others.