A dynamic-sized nonblocking work stealing deque

Abstract

The non-blocking work-stealing algorithm of Arora, Blumofe, and Plaxton (hencheforth ABP workstealing) is on its way to becoming the multiprocessor load balancing technology of choice in both industry and academia. This highly efficient scheme is based on a collection of array-based double-ended queues (deques) with low cost synchronization among local and stealing processes. Unfortunately, the algorithm's synchronization protocol is strongly based on the use of fixed size arrays, which are prone to overflows, especially in the multiprogrammed environments for which they are designed. This is a significant drawback since, apart from memory inefficiency, it means that the size of the deque must be tailored to accommodate the effects of the hard-to-predict level of multiprogramming, and the implementation must include an expensive and application-specific overflow mechanism. This paper presents the first dynamic memory workstealing algorithm. It is based on a novel way of building non-blocking dynamic-sized work stealing deques by detecting synchronization conflicts based on "pointercrossing" rather than "gaps between indexes" as in the original ABP algorithm. As we show, the new algorithm dramatically increases robustness and memory efficiency, while causing applications no observable performance penalty. We therefore believe it can replace array-based ABP work stealing deques, eliminating the need for applicationspecific overflow mechanisms.

A lock-free algorithm for concurrent bags

“…Arora et al [2] presented an efficient and lock-free work-stealing deque based on arrays. It was later improved to handle dynamic sizes by Hendler et al [16] that used a doubly-linked list of arrays, where each array can be dynamically allocated. In resemblance to [2][16], the new algorithm assign each thread its own data structure, uses arrays as the basic representation for storing items, and CAS is only used for updates that can be concurrently updated by several threads.…”

“…In resemblance to [2][16], the new algorithm assign each thread its own data structure, uses arrays as the basic representation for storing items, and CAS is only used for updates that can be concurrently updated by several threads. In resemblance to [16], additional array blocks are dynamically allocated when needed, although the new algorithm only maintains a single-linked list. In contrast to [2][16], the new algorithm provides linearizable bag semantics; e.g.…”

Lace: Non-blocking Split Deque for Work-Stealing

“…Two unbounded non-blocking deques were proposed, the deque by Hendler et al based on linked lists of small arrays [10], and the Chase-Lev deque that uses dynamic circular arrays [5].…”

“…Tasks are often stored as pointers that are removed from the deque when the task is stolen [9,2,10,5]. To virtually eliminate the overhead of task creation for tasks that are never stolen, Faxén proposed a direct task stack, storing tasks instead of pointers in the work queue, implemented in Wool [7,8].…”

Laws of order

“…Examining known work stealing algorithms that avoid the AWAR pattern (i.e., avoid the use of complex atomic operations) in the common path of the owner's methods [3,16,19,20], reveals that they all contain the RAW pattern in the common path of the take method that succeeds in extracting work items.…”