Managed languages improve programmer productivity with type safety and garbage collection, which eliminate memory errors such as dangling pointers, double frees, and buffer overflows. However, programs may still leak memory if programmers forget to eliminate the last reference to an object that will not be used again. Leaks slow programs by increasing collector workload and frequency. Growing leaks crash programs. Instead of crashing, leak pruning extends program availability by predicting and reclaiming leaked objects at run time. Whereas garbage collection over-approximates live objects using reachability, leak pruning predicts dead objects and reclaims them based on how stale they are and the size of stale data structures. Leak pruning preserves semantics because it waits for heap exhaustion before reclaiming objects and then poisons references to objects it reclaims. If the program later tries to access these objects, the virtual machine (VM) throws an internal error. We implement leak pruning in a Java VM, show its overhead is low, and evaluate it on 10 leaking programs. Leak pruning does not help two programs, executes four substantial programs 1.6-35X longer, and executes four programs, including two leaks in Eclipse, for at least 24 hours. In the worst case, leak pruning defers fatal errors. In the best case, programs with unbounded memory requirements execute indefinitely and correctly in bounded memory with consistent throughput.

A managed runtime environment, such as the Java virtual machine, is non-trivial to benchmark. Java performance is affected in various complex ways by the application and its input, as well as by the virtual machine (JIT optimizer, garbage collector, thread scheduler, etc.). In addition, nondeterminism due to timer-based sampling for JIT optimization, thread scheduling, and various system effects further complicate the Java performance benchmarking process. Replay compilation is a recently introduced Java performance analysis methodology that aims at controlling nondeterminism to improve experimental repeatability. The key idea of replay compilation is to control the compilation load during experimentation by inducing a pre-recorded compilation plan at replay time. Replay compilation also enables teasing apart performance effects of the application versus the virtual machine. This paper argues that in contrast to current practice which uses a single compilation plan at replay time, multiple compilation plans add statistical rigor to the replay compilation methodology. By doing so, replay compilation better accounts for the variability observed in compilation load across compilation plans. In addition, we propose matchedpair comparison for statistical data analysis. Matched-pair comparison considers the performance measurements per compilation plan before and after an innovation of interest as a pair, which enables limiting the number of compilation plans needed for accurate performance analysis compared to statistical analysis assuming unpaired measurements.

Deterministic replay systems, which record and replay non-deterministic events during program execution, have many applications such as bug diagnosis, intrusion analysis and fault tolerance. It is well understood how to replay native (e.g., C) programs on multi-processors, while there is little work for concurrent java applications on multicore. State-of-the-art work for Java either assumes data-race free execution, or relies on static instrumentation, which leads to missing some necessary nondeterministic events. This paper proposes the ORDER framework to record and reproduce non-deterministic events inside Java virtual machine (JVM). Based on observations of good locality at object level for threads and frequent object movements due to garbage collection, ORDER records and replays non-deterministic data accesses by logging and enforcing the order in which threads access objects. This essentially eliminates unnecessary dependencies introduced by address changes of objects during garbage collection and enjoys good locality as well as less contention, which may result in scalable performance on multicore. Further, by dynamically instrumenting Java code in the JVM compilation pipeline, ORDER naturally covers non-determinism in dynamically loaded classes. We have implemented ORDER based on Apache Harmony. Evaluation on SPECjvm2008, PseudoJBB2005, and JRuby shows that ORDER only incurs 108 % performance overhead on average and scales well on a 16-core Xeon testbed. Evaluation with a real-world application, JRuby, shows that several real-world concurrency bugs can be successfully reproduced.