Redundant boxing elimination by a dynamic compiler for Java

Abstract

Auto-boxing improves code readability by eliminating the need for explicit boxing code, but it does not improve performance, because it does not eliminate boxing code but inserts the code implicitly. Current auto-boxing implementations try to improve performance by caching some of the boxed values in order to avoid wrapper class instance allocation on each boxing operation. Such an implementation, however, sometimes suffers greater performance degradation because it prevents traditional optimizations, such as redundant instance allocation elimination. This paper presents a new optimizing technique that eliminates the boxing code inserted by auto-boxing if the code is found redundant. Estimation using the SPECjbb2005 benchmark showed that this optimization improved performance by 3.6%.

One VM to rule them all

“…Since then it has been applied to array-bound check elimination [71], escape analysis and stack allocation [36], object fusing [68], boxing elimination [15] and partial redundancy elimination [44].…”

Da capo con scala

“…We have therefore measured to which degree the Java and Scala benchmarks create boxed values. To this end, we distinguish between the mere request to create a boxed value by using the appropriate valueOf factory method and the actual creation of a new instance using the boxed type's constructor; usage of the factory method allows for caching but may impede JIT compiler optimizations [11]. Figures 17a and 17b show how many boxed values are requested and how many are actually created.…”