Cache Line Boundary Allocation for Garbage Collected Systems

by Wagle, Prasad Ajit

Abstract (Summary)
WAGLE, PRASAD AJIT. Cache Line Boundary Allocation for Garbage Collected Systems. (Under the direction of Dr. Edward F Gehringer.) Garbage collected systems became increasingly popular with the advent of Java programming language. Cache performance of garbage collected systems has been a heavily researched area. Past works have shown that cache line utilization has been poor for garbage collected systems. This work aims at reducing the cache miss rate by aligning objects to cache line boundaries during object allocation. Object alignment to a cache line makes that object use minimum number of cache lines most of the times. This improves the overall utilization of a cache line between the time it is brought in to the cache and evicted out. Improved cache utilization translates into improvement in total program execution time. Both cache performance (cache miss rate) and total execution time is studied on the Java Grande, Dacapo suite, Spec JBB2000, Spec JVM98 and soot benchmarks. Boundary allocation strategy is shown to improve cache performance/execution time for most of the memory intensive benchmarks. The strategy performs well when implemented in the old generation of generational collectors. Boundary allocation in young generation shows only marginal performance improvement. Most objects in young generation die very soon. Hence advantage of boundary allocation is negated by the overhead of aligning objects and more garbage collection cycles due to fragmentation. Performance of a benchmark would improve if fragmentation due to cache line alignment is less. It is shown that fragmentation depends on the object size distribution. Benchmarks that show performance improvement have denser object size distribution for size equal to less than half the cache line size and vice versa. It is also shown that fragmentation would decrease (hence performance would improve) with bigger cache lines.
Bibliographical Information:

Advisor:Dr. Suleyman Sair; Dr. Yan Solihin; Dr. Edward F Gehringer

School:North Carolina State University

School Location:USA - North Carolina

Source Type:Master's Thesis

Keywords:computer engineering


Date of Publication:08/22/2007

© 2009 All Rights Reserved.