This dissertation shows a new approach to writing object-oriented parallel programs based on design patterns, frameworks, and multiple layers of abstraction.

We present an efficient encoding of core Java constructs in a simple, implementable typed intermediate language. The encoding, after type erasure, has the same operational behavior as a standard implementation using vtables and selfapplication for method invocation. Classes inherit super-class methods with no overhead. We support mutually recursive classes while preserving separate compilation. Our strategy extends naturally to a significant subset of Java, including interfaces and privacy. The formal translation using Featherweight Java allows comprehensible type-preservation proofs and serves as a starting point for extending the translation to new features.

Virtual machines face significant performance challenges beyond those confronted by traditional static optimizers. First, portable program representations and dynamic language features, such as dynamic class loading, force the deferral of most optimizations until runtime, inducing runtime optimization overhead. Second, modular

This article presents an escape analysis framework for Java to determine (1) if an object is not reachable after its method of creation returns, allowing the object to be allocated on the stack, and (2) if an object is reachable only from a single thread during its lifetime, allowing unnecessary synchronization operations on that object to be removed. We introduce a new program abstraction for escape analysis, the connection graph, that is used to establish reachability relationships between objects and object references. We show that the connection graph can be succinctly summarized for each method such that the same summary information may be used in different calling contexts without introducing imprecision into the analysis. We present an interprocedural algorithm that uses the above property to efficiently compute the connection graph and identify the non-escaping objects for methods and threads. The experimental results, from a prototype implementation of our framework in the IBM High Performance Compiler for Java, are very promising. The percentage of objects that may be allocated on the stack exceeds 70 % of all dynamically created objects in the user code in three out of the ten benchmarks (with a median of 19%), 11 % to 92 % of all mutex lock operations are eliminated in those ten programs (with a median of 51%), and the overall execution time reduction ranges from 2 % to 23 % (with a median of 7%) on a 333 MHz PowerPC workstation with 512 MB memory.

Pointer traversals pose significant overhead to the execution of object-oriented programs, since every access to an object?s state requires a pointer dereference. Eliminating redundant pointer traversals reduces both instructions executed as well as redundant memory accesses to relieve pressure on the memory subsystem. We describe an approach to elimination of redundant access expressions that combines partial redundancy elimination (PRE) with type-based alias analysis (TBAA). To explore the potential of this approach we have implemented an optimization framework for Java class files incorporating TBAA-based PRE over pointer access expressions. The framework is implemented as a class-file-to-class-file transformer; optimized classes can then be run in any standard Java execution environment. Our experiments demonstrate improvements in the execution of optimized code for several Java benchmarks running in diverse execution environments: the standard interpreted JDK virtual machine, a virtual machine using ?just-in-time? compilation, and native binaries compiled off-line (?way-ahead-of-time?). Overall, however, our experience is of mixed success with the optimizations, mainly because of the isolation between our optimizer and the underlying execution environments which prevents more effective cooperation between them.We isolate the impact of access path PRE using TBAA, and demonstrate that Java?s requirement of precise exceptions can noticeably impact code-motion optimizations like PRE.

Optimizing exception handling is critical for programs that frequently throw exceptions. We observed that there are many such exception-intensive programs in various categories of Java programs. There are two commonly used exception handling techniques, stack unwinding and stack cutting. Stack unwinding optimizes the normal path, while stack cutting optimizes the exception handling path. However, there has been no single exception handling technique to optimize both paths. We propose a new technique called Exception-Directed Optimization (Edo), which optimizes exception-intensive programs without slowing down exception-minimal programs. Edo, a feedbackdirected dynamic optimization, consists of three steps, exception path profiling, exception path inlining, and throw elimination. Exception path profiling attempts to detect hot exception paths. Exception path inlining compiles the catching method in a hot exception path, inlining the rest of methods in the path. Throw elimination replaces a throw with the explicit control flow to the corresponding catch. We implemented Edo in IBM's production Justin -Time compiler, and obtained the experimental results, which show that, in SPECjvm98, it improved performance of exceptionintensive programs by up to 18.3% without a#ecting performance of exception-minimal programs at all. Categories and Subject Descriptors D.3 [Software]: Programming Languages; D.3.4 [Programming Languages]: Processors---incremental compilers, optimization, runtime environment General Terms Performance, Experimentation, Languages Keywords Feedback-directed dynamic optimization, dynamic compilers, exception handling, inlining Copyright c # 2001 by the Association for Computing Machinery, Inc. Permission to make digital or hard copies of part or a...

The Java language provides exceptions in order to handle errors gracefully. However, the presence of exception handlers complicate the job of a JIT (Just-in-Time) compiler, including optimizations and register allocation, even though exceptions are rarely used in most programs. This paper describes some mechanisms for removing overheads imposed by the existence of exception handlers, including on-demand translation of exception handlers, which expose more optimization opportunities in normal flow. In addition, we also minimize the exception handling overhead for frequently thrown exceptions by jumping directly from the exception throwing point into the exception handler through a technique called exception handler prediction. Experiments show that the existence of exception handlers indeed does not interfere with the translation of normal flow using our exception handling mechanisms. Also, the results reveal that frequently thrown exceptions are efficiently handled with exception handler prediction.

Exception handling is a powerful and widely-used programming language abstraction for constructing robust software systems. Unfortunately, it introduces an inter-procedural flow of control that can be difficult to reason about. Failure to do so correctly can lead to security vulnerabilities, breaches of API encapsulation, and any number of safety policy violations. We present a fully automated tool that statically infers and characterizes exception-causing conditions in Java programs. Our tool is based on an inter-procedural, contextsensitive analysis. The output of this tool is well-suited for use as human-readable documentation of exceptional conditions. We evaluate the output of our tool by comparing it to over 900 instances of existing exception documentation in almost two million lines of code. We find that the output of our tool is at least as good as existing documentation 85% of the time and is better 25 % of the time.

It is difficult to write programs that behave correctly in the presence of run-time errors. Proper behavior in the face of exceptional situations is important to the reliability of long-running programs. Existing programming language features often provide poor support for executing clean-up code and for restoring invariants. We present a dataflow analysis for finding a certain class of mistakes made during exceptional situations. We also present a specification miner for automatically inferring partial notions of what programs should be doing. Finally, we propose and evaluate a new language feature, the compensation stack, to make it easier to write solid code in the presence of run-time errors. We give a dataflow analysis for finding a certain class of exception-handling mis-takes: those that arise from a failure to release resources or to clean up properly along all paths. Many real-world programs violate such resource usage rules because of incorrect exception handling. Our flow-sensitive analysis keeps track of outstanding obligations along program paths and does a precise modeling of control flow in the presence of exceptions. Using it, we have found over 800 exception handling mistakes in almost 4 million lines of Java code. The analysis is unsound and produces false positives, but a few simple filtering rules suffice to remove them in practice. The remaining mistakes were manually verified. These mistakes cause sockets, files and database handles to be leaked along some paths. Specifications are necessary in order to find software bugs using program verification tools. We give a novel automatic specification mining algorithm that uses information about exception handling to learn temporal safety rules. Our algorithm is based on the observation that programs often make mistakes along exceptional control-flow paths, even when they behave correctly on normal execution paths. We show that this focus improves the miner’s effectiveness at discovering specifications beneficial for bug finding. We present quantitative results comparing our technique to four existing miners. We highlight assumptions made by various miners that are not always borne out in practice. Additionally, we apply our algorithm to existing Java programs and analyze its ability to learn specifications that find bugs in those programs. In our experiments, we find filtering candidate specifications to be more important than ranking them. We find 430 bugs in 1 million lines of code. Notably, we find 250 more bugs using per-program specifications learned by our algorithm than with generic specifications that apply to all programs. We present a characterization of the most common causes of those bugs and discuss the limitations of exception handling, finalizers and destructors. Based on that characterization we propose a programming language feature, the compensation stack, that keeps track of obligations at run time and ensures that they are discharged. Finally, we present case studies to demonstrate that this feature is natural, efficient, and can improve reliability; for example, retrofitting a 34,000-line program with compensation stacks resulted in a 0.5% code size decrease, a surprising 17% speed increase (from correctly deallocating resources in the presence of exceptions), and more consistent behavior.

Abstract. Partial redundancy elimination can reduce the number of loads corresponding to field and array accesses in Java programs. The reuse of values loaded from memory at subsequent occurrences of load expressions must be done with care: Precise exceptions and the potential of side effects through method invocations and concurrent modifications in multi-threaded programs must be considered. This work focuses on the effect of concurrency on the load optimization. Unlike previous approaches, our system determines accurate information about side effects and concurrency through a whole-program analysis. Partial redundancy elimination is extended to exploit this information and to broaden the optimization scope. There are three main results: (1) Load elimination is effective even in the most conservative variant without side effect and concurrency analysis (avg. dynamic reduction of loads 23.4%, max. 55.6%). (2) Accurate side effect information can significantly increase the number of optimized expressions (avg. dynamic reduction of loads 28.6%, max. 66.1%). (3) Information about concurrency can make the optimization independent of the memory model, enables aggressive optimization across synchronization statements, and can improve the number of optimization opportunities compared to an uninformed optimizer that is guided by a (weak) memory model (avg. dynamic reduction of loads 28.5%, max. 70.3%). 1