object encapsulation and enable local reasoning about program correctness in object-oriented languages. However, a type system that enforces strict object encapsulation is too constraining: it does not allow e#cient implementation of important constructs like iterators. This paper argues that the right way to solve the problem is to allow objects of classes defined in the same module to have privileged access to each other's representations; we show how to do this for inner classes. This approach allows programmers to express constructs like iterators and yet supports local reasoning about the correctness of the classes, because a class and its inner classes together can be reasoned about as a module. The paper also sketches how we use our variant of ownership types to enable e#cient software upgrades in persistent object stores.

Making software reliable is one of the most important technological challenges facing our society today. This thesis presents a new type system that addresses this problem by statically preventing several important classes of programming errors. If a program type checks, we guarantee at compile time that the program does not contain any of those errors. We designed our type system in the context of a Java-like object-oriented language; we call the resulting system SafeJava. The SafeJava type system offers significant software engineering benefits. Specifically, it provides a statically enforceable way of specifying object encapsulation and enables local reasoning about program correctness; it combines effects clauses with encapsulation to enable modular checking of methods in the presence of subtyping; it statically prevents data races and deadlocks in multithreaded programs, which are known to be some of the most difficult programming errors to detect, reproduce, and

Persistent object stores require a way to automatically upgrade persistent objects, to change their code and storage representation. Automatic upgrades are a challenge for such systems. Upgrades must be performed in a way that is efficient both in space and time, and that does not stop application access to the store. In addition, however, the approach must be modular: it must allow programmers to reason locally about the correctness of their upgrades similar to the way they would reason about regular code. This paper provides solutions to both problems. The paper first defines upgrade...

... object encapsulation and enable local reasoning about program correctness in object-oriented languages. However, a type system that enforces strict object encapsulation is too constraining: it does not allow efficient implementation of important constructs like iterators. This paper argues that the right way to solve the problem is to allow objects of classes defined in the same module to have privileged access to each other's representations; we show how to do this for inner classes. This approach allows programmers to express constructs like iterators and yet supports local reasoning about the correctness of the classes, because a class and its inner classes together can be reasoned about as a module. The paper also sketches how we use our variant of ownership types to enable efficient software upgrades in persistent object stores.

This paper outlines the design and implementation of the PEVM, a new scalable, high-performance implementation of orthogonal persistence for the Java platform (OPJ). The PEVM is based on the Sun Microsystems Laboratories Virtual Machine for Research, which features an optimizing Just-In-Time compiler, exact generational garbage collection, and fast thread synchronization. The PEVM also uses a new, scalable persistent object store designed to manage 80GB of objects. It is approximately ten times faster than previous OPJ implementations and can run signi cantly larger programs. Despite its greater speed and scalability, the PEVM's implementation is much simpler (e.g., just 43 % of the VM source patches needed by our previous OPJ implementation). This is largely due to the pointer swizzling strategy we chose, the ResearchVM's exact memory management, and simple but e ective mechanisms. For example, we implement some key data structures in the Java programming language since this automatically makes them persistent.

Object-oriented databases allow objects that are manipulated by programs to be stored reliably so that they can be used again later and shared with other programs. Since objects in the OODB may live a long time, there may be a need to upgrade them: to change their code and storage representation. This paper describes a technique for upgrading objects in an OODB. The approach preserves the database state by transforming objects to their new classes while retaining their state and their identity. The approach is efficient: we do not interrupt application execution to run an upgrade, but instead run the upgrade incrementally, one transform at a time. Objects are transformed lazily, but just in time; applications never observe non-upgraded objects. Laziness can sometimes lead to problems for the code that transforms objects, however; e.g., a transform might observe broken invariants or interfaces unknown at the time it was written. We define precisely when these problems arise, and we also provide mechanisms for avoiding them. Ours is the first work to provide a full analysis of these problems and to allow safe lazy upgrades even when problems arise. We have implemented our approach on the Thor OODB and we present performance results that show that the overhead of our infrastructure is low.

Orthogonal persistence provides the programmer with persistence for all data types, with minimal impact on the programing model or development process. We motivate the addition of orthogonal persistence to the Java ™ platform, and show how this results in a simple and appealing application development model. The overall goal is to provide the illusion of continuous computation in the face of system shutdowns, planned or unplanned. This is achieved by checkpointing the state of the system periodically to stable memory. We describe how the principles of orthogonal persistence are applied to the Java ™ programming language and specify the small set of changes to the Java language specification and core libraries necessary to fulfill these principles. We describe the rationale for our particular choices, informed by the experience with the PJama prototype implementations. Finally, the programming model for managing state that is external to the Java ™ virtual machine is discussed in detail.

Developing embedded systems software poses unique challenges to Java application developers and virtual machine designers. Chief among these challenges is the memory footprint of both the virtual machine and the applications that run within it. With the rapidly increasing set of features provided by the Java language, virtual machine designers are often forced to build custom implementations that make various tradeoffs between the footprint of the virtual machine and the subset of the Java language and class libraries that are supported. In this paper, we present the ExoVM, a system in which an application is initialized in a fully featured virtual machine, and then the code, data, and virtual machine features necessary to execute it are packaged into a binary image. Key to this process is feature analysis, a technique for computing the reachable code and data of a Java program and its implementation inside the VM simultaneously. The ExoVM reduces the need to develop customized embedded virtual machines by reusing a single VM infrastructure and automatically eliding the implementation of unused Java features on a per-program basis. We present a constraint-based instantiation of the analysis technique, an implementation in IBM’s J9 Java VM, experiments evaluating our technique for the EEMBC benchmark suite, and some discussion of the individual costs of some of Java’s features. Our evaluation shows that our system can reduce the non-heap memory allocation of the virtual machine by as much as 75%. We discuss VM and language design decisions that our work shows are important in targeting embedded systems, supporting the long-term goal of a common VM infrastructure spanning from motes to large servers. 1.