Compiler-based prefetching for recursive data structures

Luk, Chi-Keung; Mowry, Todd C.

doi:10.1145/237090.237190

Cited by 249 publications

(46 citation statements)

References 18 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Neither prefetching technique is able to improve em3d and power because the number of cache misses in very low in these programs so most of the prefetches hit in the L1 cache. Our prefetching results are similar to those reported in related work for linked structures in C programs [18,24].…”

Section: Resultssupporting

confidence: 88%

“…We evaluate our prefetching algorithms using an objectoriented Java version of the Olden benchmark suite that our research group wrote [7]. Other researchers use the C version of the Olden suite to evaluate optimizations for pointerbased programs [8,18,24]. Table 1 lists the benchmarks we use in our experiments, along with characteristics about each program.…”

Section: Resultsmentioning

confidence: 99%

“…Luk and Mowry first developed an effective compiler technique for software prefetching recursive data structures in C programs [18,17]. Their prefetching techniques include greedy prefetching, history-pointer prefetching, and data-linearization.…”

Section: Related Workmentioning

confidence: 99%

“…Given an object o, we know the address of objects that o references, and we cannot prefetch other objects without following pointer chains. Recent pointer prefetching work considers C programs only [16,18,24,15,28]. Objectoriented Java programs pose additional analysis challenges because they mostly allocate data dynamically, contain frequent method invocations, and often implement loops with recursion.…”

Section: Introductionmentioning

confidence: 99%

See 3 more Smart Citations

Data flow analysis for software prefetching linked data structures in Java

Cahoon¹,

McKinley²

Proceedings 2001 International Conference on Parallel Architectures and Compilation Techniques

View full text Add to dashboard Cite

In this paper, we describe an effective compile-time analysis for software prefetching in Java. Previous work in software data prefetching for pointer-based codes uses simple compiler algorithms and does not investigate prefetching for object-oriented language features that make compiletime analysis difficult. We develop a new data flow analysis to detect regular accesses to linked data structures in Java programs. We use intra and interprocedural analysis to identify profitable prefetching opportunities for greedy and jump-pointer prefetching, and we implement these techniques in a compiler for Java. Our results show that both prefetching techniques improve four of our ten programs. The largest performance improvement is 48% with jumppointers, but consistent improvements are difficult to obtain.

show abstract

Section: Resultssupporting

confidence: 88%

Section: Resultsmentioning

confidence: 99%

Section: Related Workmentioning

confidence: 99%

Section: Introductionmentioning

confidence: 99%

See 2 more Smart Citations

Data flow analysis for software prefetching linked data structures in Java

Cahoon¹,

McKinley²

Proceedings 2001 International Conference on Parallel Architectures and Compilation Techniques

View full text Add to dashboard Cite

show abstract

“…With ILP processors, it is unclear how to assign stall time to specific instructions since each instruction's execution may be overlapped with both preceding and following instructions. We use the following convention, similar to previous work (e.g., [5]), to account for stall cycles. At every cycle, we calculate the ratio of the instructions retired from the instruction window in that cycle to the maximum retire rate of the processor and attribute this fraction of the cycle to the busy time.…”

Section: Performance Metricsmentioning

confidence: 99%

The impact of exploiting instruction-level parallelism on shared-memory multiprocessors

Pai

Ranganathan

Abdel-Shafi

et al. 1999

IEEE Trans. Comput.

View full text Add to dashboard Cite

AbstractÐCurrent microprocessors incorporate techniques to aggressively exploit instruction-level parallelism (ILP). This paper evaluates the impact of such processors on the performance of shared-memory multiprocessors, both without and with the latencyhiding optimization of software prefetching. Our results show that, while ILP techniques substantially reduce CPU time in multiprocessors, they are less effective in removing memory stall time. Consequently, despite the inherent latency tolerance features of ILP processors, we find memory system performance to be a larger bottleneck and parallel efficiencies to be generally poorer in ILPbased multiprocessors than in previous generation multiprocessors. The main reasons for these deficiencies are insufficient opportunities in the applications to overlap multiple load misses and increased contention for resources in the system. We also find that software prefetching does not change the memory bound nature of most of our applications on our ILP multiprocessor, mainly due to a large number of late prefetches and resource contention. Our results suggest the need for additional latency hiding or reducing techniques for ILP systems, such as software clustering of load misses and producer-initiated communication.

show abstract

Communication Optimizations for Parallel C Programs

Zhu¹,

Hendren²

1999

Journal of Parallel and Distributed Computing

View full text Add to dashboard Cite

Compiler-based prefetching for recursive data structures

Cited by 249 publications

References 18 publications

Data flow analysis for software prefetching linked data structures in Java

Data flow analysis for software prefetching linked data structures in Java

The impact of exploiting instruction-level parallelism on shared-memory multiprocessors

Communication Optimizations for Parallel C Programs

Contact Info

Product

Resources

About