David L. Detlefs

Lock-free reference counting

Assuming the existence of garbage collection makes it easier to design implementations of concurrent data structures. However, this assumption limits their applicability. We present a methodology that, for a significant class of data structures, allows designers to first tackle the easier problem of designing a garbage-collection-dependent implementation, and then apply our methodology to achieve a garbage-collection-independent one. Our methodology is based on the well-known reference counting technique, and employs the double compare-and-swap operation.

Inlining of Virtual Methods

We discuss aspects of inlining of virtual method invocations. First, we introduce a new method test to guard inlinings of such invocations, with a different set of tradeoffs from the class-equality tests proposed previously in the literature. Second, we consider the problem of inlining virtual methods directly, with no guarding test, in dynamic languages such as Self or the Java™ programming language, whose semantics prohibit a static identification of the complete set of modules that comprise a program. In non-dynamic languages, a whole-program analysis might prove the correctness of a direct virtual inlining. In dynamic languages, however, such analyses can be invalidated by later class loading, and must therefore be treated as assumptions whose later violation must cause recompilation. In the past, such systems have required an on-stack replacement mechanism to update currently-executing invocations of methods containing invalidated inlinings. This paper presents analyses that allow some virtual calls to be inlined directly, while ensuring that invocations in progress may complete safely even if class loading invalidates the inlining for future invocations. This provides the benefits of direct inlining without the need for on-stack replacement, which can be complicated and require space-consuming data structures.

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#8482; 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#8482; virtual machine and report performance data for several large programs.

Garbage collection and local variable type-precision and liveness in Java virtual machines

Full precision in garbage collection implies retaining only those heap allocated objects that will actually be used in the future. Since full precision is not computable in general, garbage collectors use safe (i.e., conservative) approximations such as reachability from a set of root references. Ambiguous roots collectors (commonly called conservative) can be overly conservative because they overestimate the root set, and thereby retain unexpectedly large amounts of garbage. We consider two more precise collection schemes for Java virtual machines (JVMs). One uses a type analysis to obtain a type-precise root set (only those variables that contain references); the other adds a live variable analysis to reduce the root set to only the live reference variables. Even with the Java programming languages strong typing, it turns out that the JVM specification has a feature that makes type-precise root sets difficult to compute. We explain the problem and ways in which it can be solved.Our experimental results include measurements of the costs of the type and liveness analyses at load time, of the incremental benefits at run time of the liveness analysis over the type analysis alone, and of various map sizes and counts. We find that the liveness analysis often produces little or no improvement in heap size, sometimes modest improvements, and occasionally the improvement is dramatic. While further study is in order, we conclude that the main benefit of the liveness analysis is preventing bad surprises.

Even Better DCAS-Based Concurrent Deques

The computer industry is examining the use of strong synchronization operations such as double compare-and-swap (DCAS) as a means of supporting non-blocking synchronization on tomorrows multiprocessor machines. However, before such a primitive will be incorporated into hardware design, its utility needs to be proven by developing a body of effective non-blocking data structures using DCAS.In a previous paper, we presented two linearizable non-blocking implementations of concurrent deques (double-ended queues) using the DCAS operation. These improved on previous algorithms by nearly always allowing unimpeded concurrent access to both ends of the deque while correctly handling the difficult boundary cases when the deque is empty or full. A remaining open question was whether, using DCAS, one can design a non-blocking implementation of concurrent deques that allows dynamic memory allocation but also uses only a single DCAS per push or pop in the best case.This paper answers that question in the affirmative. We present a new non-blocking implementation of concurrent deques using the DCAS operation. This algorithm provides the benefits of our previous techniques while overcoming drawbacks. Like our previous approaches, this implementation relies on automatic storage reclamation to ensure that a storage node is not reclaimed and reused until it can be proved that the node is not reachable from any thread of control. This algorithm uses a linked-list representation with dynamic node allocation and therefore does not impose a fixed maximum capacity on the deque. It does not require the use of a spare bit in pointers. In the best case (no interference), it requires only one DCAS per push and one DCAS per pop. We also sketch a proof of correctness.

DCAS is not a silver bullet for nonblocking algorithm design

Despite years of research, the design of efficient nonblocking algorithms remains difficult. A key reason is that current shared-memory multiprocessor architectures support only single-location synchronisation primitives such as compare-and-swap (CAS) and load-linked/store-conditional (LL/SC). Recently researchers have investigated the utility of double-compare-and-swap (DCAS)--a generalisation of CAS that supports atomic access to two memory locations -- in overcoming these problems. We summarise recent research in this direction and present a detailed case study concerning a previously published nonblocking DCAS-based double-ended queue implementation. Our summary and case study clearly show that DCAS does not provide a silver bullet for nonblocking synchronisation. That is, it does not make the design and verification of even mundane nonblocking data structures with desirable properties easy. Therefore, our position is that while slightly more powerful synchronisation primitives can ave a profound effect on ease of algorithm design and verification, DCAS does not provide sufficient additional power over CAS to justify supporting it in hardware.

DCAS-based concurrent deques

The computer industry is currently examining the use of strong synchronization operations such as double compare-and-swap (DCAS) as a means of supporting non-blocking synchronization on tomorrows multiprocessor machines. However, before such a strong primitive will be incorporated into hardware design, its utility needs to be proven by developing a body of effective non-blocking data structures using DCAS.nAs part of this effort, we present two new linearizable non-blocking implementations of concurrent deques using the DCAS operation. The first uses an array representation, and improves on former algorithms by allowing uninterrupted concurrent access to both ends of the deque while correctly handling the difficult boundary cases when the deque is empty or full. The second uses a linked-list representation, and is the first non-blocking unbounded-memory deque implementation. It too allows uninterrupted concurrent access to both ends of the deque.