JavaTM Virtual Machine Research Articles

Multitasking without compromise

The Multitasking Virtual Machine (called from now on simply MVM) is a modification of the Java™ virtual machine. It enables safe, secure, and scalable multitasking. Safety is achieved by strict isolation of applications from one another. Resource control mechanisms augment security by preventing some denial-ofservice attacks. Improved scalability results from an aggressive application of the main design principle of MVM: share as much of the runtime as possible among applications and replicate everything else. The system can be described as a 'no compromise' approach -- all the known APIs and mechanisms of the Java programming language are available to applications. MVM is implemented as a series of carefully tuned modifications to the Java HotSpot™ virtual machine, including the dynamic compiler. This paper presents the design of MVM, focusing on several novel and general techniques: an in-runtime design of lightweight isolation, an extension of a copying, generational garbage collector to provide best-effort management of a portion of the heap space, and a transparent and automated mechanism for safe execution of user-level native code. MVM demonstrates that multitasking in a safe language can be accomplished with a high degree of protection, without constraining the language, and with competitive performance characteristics.

An object-aware memory architecture

Despite its dominance, object-oriented computation has received scant attention from the architecture community. We propose a novel memory architecture that supports objects and garbage collection (GC). Our architecture is co-designed with a Java™ Virtual Machine (JVM™) 1 1 Java and all Java-based marks are trademarks or registered trademarks of Sun Microsystems, Inc., in the United States and other countries. to improve the functionality and efficiency of heap memory management. The architecture is based on an address space for objects accessed using object IDs mapped by a translator to physical addresses. To support this, the system includes object-addressed caches, a hardware GC barrier to allow in-cache GC of objects, and an exposed cache structure cooperatively managed by the JVM. These extend a conventional architecture, without compromising compatibility or performance for legacy binaries. Our innovations enable various improvements such as: a novel technique for parallel and concurrent garbage collection, without requiring any global synchronization; an in-cache garbage collector, which never accesses main memory; concurrent compaction of objects; and elimination of most GC store barrier overheads. We compare the behavior of our system against that of a conventional generational garbage collector, both with and without an explicit allocate-in-cache operation which eliminates many write misses. Our scheme additionally trades L2 misses for in-cache operations, and provides the mapping indirection required for concurrent compaction.

Stackable thread mechanisms

A complex system environment such as a Java™ Virtual Machine layered on a series of system (e.g. Unix) services layered on a microkernel foundation requires multiple thread models to be supported simultaneously. The essence of the thread synchronization and manipulation mechanisms in these models can be factored into a small set of supporting thread primitives and mechanisms. These thread mechanisms are stackable, supporting multiple layered environments each with its own complex thread semantics.

The design and implementation of a modular and extensible Java Virtual Machine

AbstractThis paper describes the design, implementation, and experimental evaluation of a modular and extensible Java™ Virtual Machine (JVM) infrastructure, called Jupiter. The infrastructure is intended to serve as a vehicle for our research on scalable JVM architectures for a cluster of PC workstations, with support for shared memory in software. Jupiter is constructed, using a building block architecture, out of many modules with small, simple interfaces. This flexible structure, similar to UNIX® shells that build complex command pipelines out of discrete programs, allows the rapid prototyping of our research ideas by confining changes in JVM design to a small number of modules. In spite of this flexibility, Jupiter delivers good performance. Experimental evaluation of the current implementation of Jupiter using the SPECjvm98 and the EPCC Java Grande single‐threaded and multithreaded benchmarks reflects competitive performance. Jupiter is on average about 2.5 times faster than Kaffe and about 2 times slower than the Sun Microsystems JDK (interpreter versions only). By providing a flexible JVM infrastructure that delivers competitive performance, we believe we have developed a framework that supports further research into JVM scalability. Copyright © 2003 John Wiley & Sons, Ltd.

A java virtual machine architecture for very small devices

The smallest complete Java™ virtual machine implementations in use today are based on the CLDC standard and are deployed in mobile phones and PDAs. These implementations require several tens of kilobytes. Smaller Java-like implementations also exist, but these involve compromises in Java semantics. This paper describes a JVM™ architecture designed for very small devices. It supports all the CLDC Java platform semantics, including exact garbage collection, dynamic class loading, and verification. For portability and ease of debugging, the entire system is written in the Java language, with key components automatically translated into C and compiled for the target device. The resulting system will run on the next generation of smart cards, and has performance comparable to the reference CLDC implementation available from Sun™.

Implementation of automated fine-granularity locking in a persistent programming language

To gain broader acceptance, Persistent Programming Languages (PPL) need a better integration of the transaction paradigm. Much like database systems, computations should all run in the context of a transaction; the transactional properties should be applied to all objects irrespective of their type and lifetime, and they must be enforced automatically via safe and efficient mechanisms. This requires lock managers that can cope with the requirement for automated locking in a PPL. This paper reports our design and implementation of such lock managers and the accompanying machinery that a PPL must incorporate to automate locking. Our innovative ingredients for an efficient lock management recipe are a close collaboration with the memory manager of the PPL and the sharing of lock data structures of equal value between locked objects. Sharing such lock values has numerous benefits: it makes the space consumption of locking negligible and independent of the number of locks, which frees transactions of any restriction on the volume of locks that they can use; it eliminates the need for expensive book-keeping of locks by transactions; and it enables the use of both non-blocking synchronization and memorization techniques for whole locking operations. These features combine to achieve fast lock acquisition for the non-blocking cases (6 RISC instructions), and make releasing locks practically free. The impact of lock management techniques and automated locking strategy on the performance of a PPL are analysed based on measurements of various lock managers integrated into PJama, a JavaTM Virtual Machine (JVM) which provides orthogonal persistence. Copyright © 2000 John Wiley & Sons, Ltd.

Java Virtual Machine Profiler Interface

We present the JavaTM Virtual Machine Profiler Interface (JVMPI), which defines a general-purpose and portable mechanism for obtaining comprehensive profiling data from the Java virtual machine. We show that it is extensible, nonintrusive, and powerful enough to suit the needs of different profilers and virtual machine implementations. With the JVMPI, most profiler vendors will not need to build custom instrumentation in the Java virtual machine. In addition, we solve challenges to profiler design and implementation posed by the multithreading and garbage collection support provided by the Java virtual machine. Profilers based on the JVMPI can produce thread-aware CPU time profiles, uncover heavy memory allocation sites, detect unnecessary object retention, pinpoint scalability problems caused by high monitor contention, reveal thread deadlocks, and perform interactive profiling with minimum overhead. We also describe HPROF, a profiler based on JVMPI, developed by us to demonstrate the power of JVMPI.

An efficient meta-lock for implementing ubiquitous synchronization

Programs written in concurrent object-oriented languages, especially ones that employ thread-safe reusable class libraries, can execute synchronization operations (lock, notify, etc.) at an amazing rate. Unless implemented with utmost care, synchronization can become a performance bottleneck. Furthermore, in languages where every object may have its own monitor, per-object space overhead must be minimized. To address these concerns, we have developed a meta-lock to mediate access to synchronization data. The meta-lock is fast (lock + unlock executes in 11 SPARC™ architecture instructions), compact (uses only two bits of space), robust under contention (no busy-waiting), and flexible (supports a variety of higher-level synchronization operations). We have validated the meta-lock with an implementation of the synchronization operations in a high-performance product-quality Java™ virtual machine and report performance data for several large programs.