Laurent P. Daynes

Task-aware garbage collection in a multi-tasking virtual machine

A multi-tasking virtual machine (MVM) executes multiple programs in isolation, within a single operating system process. The goal of a MVM is to improve startup time, overall system throughput, and performance, by effective reuse and sharing of system resources across programs (tasks). However, multitasking also mandates a memory management system capable of offering a guarantee of isolation with respect to garbage collection costs, accounting of memory usage, and timely reclamation of heap resources upon task termination.To this end, we investigate and evaluate, novel task-aware extensions to a state-of-the-art MVM garbage collector (GC). Our task-aware GC exploits the generational garbage collection hypothesis, in the context of multiple tasks, to provide performance isolation by maintaining task-private young generations. Task aware GC facilitates concurrent per-task allocation and promotion, and minimizes synchronization and scanning overhead. In addition, we efficiently track per-task heap usage to enable GC-free reclamation upon task termination. Moreover, we couple these techniques with a light-weight synchronization mechanism that enables per-task minor collection, concurrently with allocation by other tasks.We empirically evaluate the efficiency, scalability, and through-put that our task-aware GC system enables.

MTM2: Scalable Memory Management for Multi-tasking Managed Runtime Environments

Multi-tasking, managed runtime environments (MREs) for modern type-safe, object-oriented programming languages enable isolated, concurrent execution of multiple applications within a single operating system process. Multi-tasking MREs can potentially extract high-performance on desktop and hand-held systems through aggressive sharing of classes and compiled code, and by exploiting high-level dynamic program information. We investigate the performance of a state-of-the-art multi-taking MRE for concurrent program execution. We find that due to limited support for multi-tasking and performance isolation in the memory management subsystem, multi-tasking performs poorly compared to a production-quality, single-tasking MRE. We present MTM2: a comprehensive memory management system for concurrent multi-tasking. MTM2facilitates performance isolation and efficient heap space usage through on-demand allocation of application-private regions. MTM2mitigates fragmentation using a novel hybrid garbage collector that combines mark-sweep with opportunistic copying. Our evaluation shows that MTM2improves overall performance, scalability, and footprint for concurrent workloads over state-of-the-art, multi- and single-tasking MREs.