Object-oriented programming languages provide a rich set of features that provide significant software engineering benefits. The increased productivity provided by these features comes at a justifiable cost in a more sophisticated runtime system whose responsibility is to implement these features e#ciently. However, the virtualization introduced by this sophistication provides a significant challenge to understanding complete system performance, not found in traditionally compiled languages, such as C or C++. Thus, understanding system performance of such a system requires profiling that spans all levels of the execution stack, such as the hardware, operating system, virtual machine, and application.

Modern Java programs, such as middleware and application servers, include many complex software components. Improving the performance of these Java applications requires a better understanding of the interactions between the application, virtual machine, operating system, and architecture. Hardware performance monitors, which are available on most modern processors, provide facilities to obtain detailed performance measurements of long-running applications in real time. However, interpreting the data collected using hardware performance monitors is difficult because of the low-level nature of the data. We have

This paper proposes and implements a rigorous method for studying the dynamic behaviour of AspectJ programs. As part of this methodology several new metrics specific to AspectJ programs are proposed and tools for collecting the relevant metrics are presented. The major tools consist of: (1) a modified version of the AspectJ compiler that tags bytecode instructions with an indication of the cause of their generation, such as a particular feature of AspectJ; and (2) a modified version of the *J dynamic metrics collection tool which is composed of a JVMPI-based trace generator and an analyzer which propagates tags and computes the proposed metrics. This dynamic propagation is essential, and thus this paper contributes not only new metrics, but also non-trivial ways of computing them.

AspectJ, an aspect-oriented extension of Java, is becoming increasingly popular. However, not much work has been directed at optimising compilers for AspectJ. Optimising AOP languages provides many new and interesting challenges for compiler writers, and this paper identifies and addresses three such challenges. First, compiling around advice efficiently is particularly challenging. We provide a new code generation strategy for around advice, which (unlike previous implementations) both avoids the use of excessive inlining and the use of closures. We show it leads to more compact code, and can also improve run-time performance. Second, woven code sometimes includes run-time tests to determine whether advice should execute. One important case is the cflow pointcut which uses information about the dynamic calling context. Previous techniques for cflow were very costly in terms of both time and space. We present new techniques to minimise or eliminate the overhead of cflow using both intra- and interprocedural analyses. Third, we have addressed the general problem of how to structure an optimising compiler so that traditional analyses can be easily adapted to the AOP setting. We have implemented all of the techniques in this paper in abc, our AspectBench Compiler for AspectJ, and we demonstrate significant speedups with empirical results. Some of our techniques have already been integrated into the production AspectJ compiler, ajc 1.2.1.

Abstract. Thread level speculation (TLS) has shown great promise as a strategy for fine to medium grain automatic parallelisation, and in a hardware context techniques to ensure correct TLS behaviour are now well established. Software and virtual machine TLS designs, however, require adherence to high level language semantics, and this can impose many additional constraints on TLS behaviour, as well as open up new opportunities to exploit language-specific information. We present a detailed design for a Java-specific, software TLS system that operates at the bytecode level, and fully addresses the problems and requirements imposed by the Java language and VM environment. Using SableSpMT, our research TLS framework, we provide experimental data on the corresponding costs and benefits; we find that exceptions, GC, and dynamic class loading have only a small impact, but that concurrency, native methods, and memory model concerns do play an important role, as does an appropriate, language-specific runtime TLS support system. Full consideration of language and execution semantics is critical to correct and efficient execution of high level TLS designs, and our work here provides a baseline for future Java or Java virtual machine implementations. 1

Abstract. A trace monitor observes the sequence of events in a system, and takes appropriate action when a given pattern occurs in that sequence. Aspect-oriented programming provides a convenient framework for writing such trace monitors. We provide a brief introduction to aspect-oriented programming in AspectJ. AspectJ only provides support for triggering extra code with single events, and we present a new language feature (named tracematches) that allows one to directly express patterns that range over the whole current trace. Implementing this feature efficiently is challenging, and we report on our work towards that goal. Another drawback of AspectJ is the highly syntactic nature of the event patterns, often requiring the programmer to list all methods that have a certain property, rather than specifying that property itself. We argue that Datalog provides an appropriate notation for describing such properties. Furthermore, all of the existing patterns in AspectJ can be reduced to Datalog via simple rewrite rules. This research is carried out with abc, an extensible optimising compiler for AspectJ, which is freely available for download. 1

Java workloads are becoming more and more prominent on various computing devices. Understanding the behavior of a Java workload which includes the interaction between the application and the virtual machine (VM), is thus of primary importance during performance analysis and optimization. Moreover, as contemporary software projects are increasing in complexity, automatic performance analysis techniques are indispensable. This paper proposes an off-line methodlevel phase analysis approach for Java workloads that consists of three steps. In the first step, the execution time is computed for each method invocation. Using an off-line tool, we subsequently analyze the dynamic call graph (that is annotated with the method invocations ’ execution times) to identify method-level phases. Finally, we measure performance characteristics for each of the selected phases. This is done using hardware performance monitors. As such, our approach allows for linking microprocessor-level information at the individual methods in the Java application’s source code. This is extremely interesting information during performance analysis and optimization as programmers can use this information to optimize their code. We evaluate our approach in the Jikes RVM on an IA-32 platform using the SPECjvm98 and SPECjbb2000 benchmarks. This is done according to a number of important criteria: the overhead during profiling, the variability within and between the phases, its applicability in Java workload characterization (measuring performance characteristics of the various VM components) and application bottleneck identification. Categories and Subject Descriptors C.4 [Performance of Systems]: design studies, measurement

Many large-scale Java applications suffer from runtime bloat. They execute large volumes of methods, and create many temporary objects, all to execute relatively simple operations. There are large opportunities for performance optimizations in these applications, but most are being missed by existing optimization and tooling technology. While JIT optimizations struggle for a few percent, performance experts analyze deployed applications and regularly find gains of 2 × or more. Finding such big gains is difficult, for both humans and compilers, because of the diffuse nature of runtime bloat. Time is spread thinly across calling contexts, making it difficult to judge how to improve performance. Bloat results from a pile-up of seemingly harmless decisions. Each adds temporary objects and method calls, and often copies values between those temporary objects. While data copies are not the entirety of bloat, we have observed that they are excellent indicators of regions of excessive activity. By optimizing copies, one is likely to remove the objects that carry copied values, and the method calls that allocate and populate them. We introduce copy profiling, a technique that summarizes runtime activity in terms of chains of data copies. A flat copy profile counts copies by method. We show how flat profiles alone can be helpful. In many cases, diagnosing a problem requires data flow context. Tracking and making sense of raw copy chains does not scale, so we introduce a summarizing abstraction called the copy graph. We implement three clients analyses that, using the copy graph, expose common patterns of bloat, such as finding hot copy chains and discovering temporary data structures. We demonstrate, with examples from a large-scale commercial application and several benchmarks, that copy profiling can be used by a programmer to quickly find opportunities for large performance gains.

Speculative multithreading (SpMT), or thread level speculation (TLS), is a dynamic parallelisation technique that uses out-of-order execution and memory buffering to achieve speedup. The complexity of implementing software SpMT results in long development lead times, and the lack of reuse between different software SpMT systems makes comparisons difficult. In order to address these problems we have created libspmt, a new language independent library for speculative multithreading. It provides a virtualization of speculative execution components that are unavailable in commercial multiprocessors, and enables method level speculation when fully integrated. We have isolated a clean and minimal library interface, and the corresponding implementation is highly modular. It can accommodate hosts that implement garbage collection, exceptions, and non-speculative multithreading. We created libspmt by refactoring a previous implementation of software SpMT for Java, and by aiming for modularity have exposed several new opportunities for optimization. 1

Currently, more and more research results on measuring class diagrams have been developed in literatures. In order to study these metrics systematically and deeply, this paper analyzes and compares some typical metrics for UML class diagrams from different viewpoints, different types of relationships, different types of metric values, complexity, and theoretical and empirical validation. Finally, the authors discuss their advantages and disadvantages as well as the existing problems and the prospects.