Effectiveness of cross-platform optimizations for a java just-in-time compiler

Abstract

This paper describes the system overview of our Java Just-In-Time (JIT) compiler, which is the basis for the latest production version of IBM Java JIT compiler that supports a diversity of processor architectures including both 32-bit and 64-bit modes, CISC, RISC, and VLIW architectures. In particular, we focus on the design and evaluation of the cross-platform optimizations that are common across different architectures. We studied the effectiveness of each optimization by selectively disabling it in our JIT compiler on three different platforms: IA-32, IA-64, and PowerPC. Our detailed measurements allowed us to rank the optimizations in terms of the greatest performance improvements with the smallest compilation times. The identified set includes method inlining only for tiny methods, exception check eliminations using forward dataflow analysis and partial redundancy elimination, scalar replacement for instance and class fields using dataflow analysis, optimizations for type inclusion checks, and the elimination of merge points in the control flow graphs. These optimizations can achieve 90% of the peak performance for two industry-standard benchmark programs on these platforms with only 34% of the compilation time compared to the case for using all of the optimizations.

Design and evaluation of dynamic optimizations for a Java just-in-time compiler

“…A number of factors, such as optimization capabilities within the system, the trade-off between performance and compilation overhead for each optimization, the classification of optimizations to each level, and the profiling and method promotion system, can affect the decision. In our case, we conducted a number of experiments for evaluating the performance contribution and the compilation overhead for each optimization [Ishizaki et al 2003]. Figure 4 shows various data for compilation overhead: compilation time, compiled code size, and peak compiler memory use for the three optimization levels.…”

Redundant boxing elimination by a dynamic compiler for Java

“…An instance allocation statement that allocates an instance used only to store values. In this case, we can eliminate the instance allocation statement and all the statements that store a value in the instance, because nothing refers to the stored value [5,6].…”

Adaptive inlining and on-stack replacement in the CACAO virtual machine

“…Inlining can be counted among the most effective optimizing program transformations for a variety of programming languages, with examples of execution time improvements of 5 to 28% [18] for CLU 15% [8] for C code, 24% [4] for intermediate code, and 10 to 44% [13] or up to 40% [2] for Java programs.…”

“…Ishizaki et al [13] reported that inlining only very small methods gave performance improvements to within 15% of the peak performance, while increasing compilation time by only 6% on average. On the other hand, in order to obtain the peak performance, an increase of compile time by up to 50% had to be accepted.…”