The deployment of Java as a concurrent programming language has created a critical need for high-performance, concurrent, and incremental multiprocessor garbage collection. We present the Recycler, a fully concurrent pure reference counting garbage collector that we have implemented in the Jalapeno Java virtual machine running on shared memory multiprocessors.

This paper describes the design and implementation of a concurrent compacting garbage collector for languages that distinguish mutable data from immutable data (e.g., ML) as well as for languages that manipulate only immutable data (e.g., pure functional languages such as Haskell). The collector runs on shared-memory parallel computers and requires minimal mutator/collector synchronization. No special hardware or operating system support is required. Our concurrent collector derives from sequential semi-space copying collectors. The collector requires that a heap object includes a forwarding pointer in addition to its data fields. An access to an immutable object can be satisfied either by the original or the forwarded copy of the object. A mutable object is always accessed through the forwarded copy, if one exists. Measurements of this collector in a Standard ML compiler on a shared-memory computer indicate that it eliminates perceptible garbage-collection pauses by reclaiming storage...

Abstract. Automatic storage reclamation via reference counting has important advantages, but has always suffered from a major weakness due to its inability to reclaim cyclic data structures. We describe a novel cycle collection algorithm that is both concurrent — it is capable of collecting garbage even in the presence of simultaneous mutation — and localized — it never needs to perform a global search of the entire data space. We describe our algorithm in detail and present a proof of correctness. We have implemented our algorithm in the Jalapeño Java virtual machine as part of the Recycler, a concurrent multiprocessor reference counting garbage collector that achieves maximum mutator pause times of only 6 milliseconds. We present measurements of the behavior of the cycle collection algorithm over a set of eight benchmarks that demonstrate the effectiveness of the algorithm at finding garbage cycles, handling concurrent mutation, and eliminating global tracing. 1

The emergence of standards for programming real-time systems in Java has encouraged many developers to consider its use for systems previously only built using C, Ada, or assembly language. However, the RTSJ standard in isolation leaves many important problems unaddressed, and suffers from some serious problems in usability and safety. As a result, the use of Java for real-time programming has continued to be viewed as risky and adoption has been slow. In this paper we provide a description of IBM’s new real-time Java virtual machine product, which combines Metronome realtime garbage collection, ahead-of-time compilation, and a complete implementation of the RTSJ standard, running on top of a custom real-time multiprocessor Linux kernel. We will describe the implementation of each of these components, including how they interacted both positively and negatively, and the extensions to previous work required to move it from research prototype to a system implementing the complete semantics of the Java language. The system has been adopted for hard real-time development of naval weapons systems and soft real-time telecommunications servers. We present measurements showing that the system is able to provide sub-millisecond worst-case garbage collection latencies, 50 microsecond Linux scheduling accuracy, and eliminate non-determinism due to JIT compilation.

Constructing correct concurrent garbage collection algorithms is notoriously hard. Numerous such algorithms have been proposed, implemented, and deployed – and yet the relationship among them in terms of speed and precision is poorly understood, and the validation of one algorithm does not carry over to others. As programs with low latency requirements written in garbagecollected languages become part of society’s mission-critical infrastructure, it is imperative that we raise the level of confidence in the correctness of the underlying system, and that we understand the trade-offs inherent in our algorithmic choice. In this paper we present correctness-preserving transformations that can be applied to an initial abstract concurrent garbage collection algorithm which is simpler, more precise, and easier to prove correct than algorithms used in practice — but also more expensive and with less concurrency. We then show how both pre-existing and new algorithms can be synthesized from the abstract algorithm by a series of our transformations. We relate the algorithms formally using a new definition of precision, and informally with respect to overhead and concurrency. This provides many insights about the nature of concurrent collection, allows the direct synthesis of new and useful algorithms, reduces the burden of proof to a single simple algorithm, and lays the groundwork for the automated synthesis of correct concurrent collectors. 1.

Tracing and reference counting are uniformly viewed as being fundamentally different approaches to garbage collection that possess very distinct performance properties. We have implemented highperformance collectors of both types, and in the process observed that the more we optimized them, the more similarly they behaved — that they seem to share some deep structure. We present a formulation of the two algorithms that shows that they are in fact duals of each other. Intuitively, the difference is that tracing operates on live objects, or “matter”, while reference counting operates on dead objects, or “anti-matter”. For every operation performed by the tracing collector, there is a precisely corresponding anti-operation performed by the reference counting collector. Using this framework, we show that all high-performance collectors (for example, deferred reference counting and generational collection) are in fact hybrids of tracing and reference counting. We develop a uniform cost-model for the collectors to quantify the trade-offs that result from choosing different hybridizations of tracing and reference counting. This allows the correct scheme to be selected based on system performance requirements and the expected properties of the target application.

Real-time Garbage Collection (RTGC) has recently advanced to the point where it is being used in production for financial trading, military command-and-control, and telecommunications. However, among potential users of RTGC, there is enormous diversity in both application requirements and deployment environments. Previously described RTGCs tend to work well in a narrow band of possible environments, leading to fragile systems and limiting adoption of real-time garbage collection technology. This paper introduces a collector scheduling methodology called tax-and-spend and the collector design revisions needed to support it. Tax-and-spend provides a general mechanism which works well across a variety of application, machine, and operating system configurations. Tax-and-spend subsumes the predominant pre-existing RTGC scheduling techniques. It allows different policies to be applied in different contexts depending on the needs of the application. Virtual machines can co-exist compositionally on a single machine. We describe the implementation of our system, Metronome-TS, as an extension of the Metronome collector in IBM’s Real-time J9 virtual machine product, and we evaluate it running on an 8-way SMP blade with a real-time Linux kernel. Compared to the stateof-the-art Metronome system on which it is based, implemented in the identical infrastructure, it achieves almost 3x shorter latencies, comparable utilization at a 2.5x shorter time window, and mean throughput improvements of 10-20%.

Abstract. There are many algorithms for concurrent garbage collection, but they are complex to describe, verify, and implement. This has resulted in a poor understanding of the relationships between the algorithms, and has precluded systematic concurrent garbage collection algorithm, and show how existing snapshot and incremental update collectors, can be derived from the abstract algorithm by reducing precision. We also derive a new hybrid algorithm that reduces floating garbage while terminating quickly. We have implemented a concurrent collector framework and the resulting algorithms in IBM’s J9 Java virtual machine product and compared their performance in terms of space, time, and incrementality. The results show that incremental update algorithms sometimes reduce memory requirements (on 3 of 5 benchmarks) but they also sometimes take longer due to recomputation in the termination phase (on 4 of 5 benchmarks). Our new hybrid algorithm has memory requirements similar to the incremental update collectors while avoiding recomputation in the termination phase. 1