Abstract

Threads (“lightweight ” processes) have become. a common element of new languages and operating systems. This paper examines the performance implications of several data structure and algorithm alternatives for thread management in shared-memory multiprocessors. Both experimental measurements and analytical model projections are presented For applications with fine-grained parallelism, small differences in thread management are shown to have significant performance impact, often posing a tradeoff between throughput and latency. Per-processor data structures can be used to improve throughput, and in some circumstances to avoid locking, improving latency as well. The method used by processors to queue for locks is also shown to affect performance significantly. Normal methods of critical resource waiting can substantially degrade performance with moderate numbers of waiting processors. We present an Ethernet-style backoff algo rithm that largely eliminates~tiis effect. 1.

Citations