Time-Stamping Algorithms for Parallelization of Loops at Run-Time

In this paper, we present two new run-time algorithms for the parallelization of loops that have indirect access patterns. The algorithms can handle any type of loop-carried dependencies. They follow the INSPECTOR/EXECUTOR scheme and improve upon previous algorithms with the same generality by allowing concurrent reads of the same location and by increasing the overlap of dependent iterations. The algorithms are based on time-stamping rules and implemented using multithreading tools. The experimental results on an SMP server with four processors show that our schemes are efficient and outperform their competitors consistently in all test cases. The difference between the two proposed algorithms is that one allows partially concurrent reads without causing extra overhead in its inspector, while the other allows fully concurrent reads at a slight overhead in the dependence analysis. The algorithm allowing fully concurrent reads obtains up to an 80% improvement over its competitor.