Automatic Array Privatization

Array privatization is one of the most e ective transformations for the exploitation of parallelism. In this paper, we present a technique for automatic array privatization. Our algorithm uses data ow analysis of array references to identify privatizable arrays intraprocedurally as well as interprocedurally. It employs static and dynamic resolution to determine the last value of a lived private array. We compare the result of automatic array privatization with that of manual array privatization and identify directions for future improvement. To enhance the e ectiveness of our algorithm, we develop a goal directly technique to analysis symbolic variables in the present of conditional statements, loops and index arrays.