D. Grove, J. Dean, C. Garrett, and C. Chambers. Pro le-guided receiver class prediction. In ACM Conference on Object-Oriented Programming Systems, Languages, and Applications, pages 108-123, Oct. 1995.  Home/Search   Document Not in Database   Summary   Related Articles   Check

....a : DenseProperty storage DenseFormat numRows( numColumns( numRows( numColumns( for (j=1; j =numColumns; j ) end for end for element(i,j) element(i,j) for (i=1; i =numRows; i ) Thread norm1 Figure 11: Sequence diagram for jjAjj 1 implemented at MA level with A dp df. Method inlining [CG94, GDGC95, DGC95, Fer95, AH96, BS96, DA99, IKY 00, SHR 00] can be applied to eliminate the invocations by inserting the code of the invoked methods into the invoking methods. After applying method inlining, 9 the statements inside the nested loop appear as shown in Figure 12. An analysis of the ....

David Grove, Jerey Dean, Charles Garrett, and Craig Chambers. Prole-guided receiver class prediction. In Proceedings of the ACM Object Oriented Programming Systems, Languages and Applications { OOPSLA'95, pages 108-123, 1995.

....case, 50 50) Using this ratio, the total incoming frequency of hotMerge and dup is distributed among their outgoing edges. 6 6. 2 Feedback Directed Method Inlining Using pro ling to improve method inlining decisions can have a signi cant impact on the performance of objectoriented languages [4, 17]. Jikes RVM has the ability to perform inlining based on static heuristics, as well adaptive inlining based on a call edge pro le collected via time based sampling [3] Although Jikes RVM s adaptive inlining shows performance improvements for many benchmarks, the sampled pro le is fairly coarse ....

.... given higher priority for inlining by expanding Jikes RVM s inlining size restrictions while processing those sites (and while processing any new code introduced by inlining at those sites) Using edge counts in this way also has advantage of adding context sensitivity to the inlining decisions [17]. All code inlined into the instrumented method (possibly through multiple levels of inlining) is monitored in the context of its inlined position in that method. It may turn out that certain inlining decisions can be changed for this method even though they were bene cial in other contexts. 6.3 ....

D. Grove, J. Dean, C. Garrett, and C. Chambers. Proleguided receiver class prediction. In ACM Conference on ObjectOriented Programming Systems, Languages, and Applications, pages 108-123, Oct. 1995.

....numRows( numColumns( numRows( numColumns( for (j=1; j =numColumns; j ) end for end for element(i,j) element(i,j) for (i=1; i =numRows; i ) Thread norm1 Fig. 11. Sequence diagram for jjAjj1 implemented at MA level with A dense and stored in dense format. 14 Method inlining [13, 21, 15, 18, 2, 6, 16, 25, 37] can be applied to eliminate the invocations by inserting the code of the invoked methods into the invoking methods. After applying method inlining, 6 , the statements inside the nested loop appear as shown in Figure 12. if (a instanceOf DenseProperty) first guard double aux; try if ....

David Grove, Jerey Dean, Charles Garrett, and Craig Chambers. Prole-guided receiver class prediction. In Proceedings of the ACM Object Oriented Programming Systems, Languages and Applications { OOPSLA'95, pages 108-123, 1995. 32

....performance improvements will come from feedback directed optimizations, where pro ling information is used to decide not only what to optimize, but how to optimize. There exists a large body of work on collecting o ine pro les [3, 10, 11, 15, 26] as well as optimizations based on o ine pro les [6,16,17,19,20,27]. Although some systems [5, 9, 21, 22, 32] apply limited forms of online feedback directed optimizations, most of the o ine work mentioned above has not yet been applied in fully automated online systems. The main diculty in applying these optimizations online is that they often rely on ....

....in applying these optimizations online is that they often rely on instrumenting the code to collect detailed information about program execution, and instrumentation can cause substantial performance degradation. Overheads in the range of 30 1,000 above non instrumented code is not uncommon [3, 10, 11, 16, 17, 27], and overheads in the range of 10,000 (100 times slower) have been reported [16] An online system needs to execute instrumented code for some period of time, prior to performing optimization. The overhead introduced by instrumentation makes this task dif cult to perform for several reasons. ....

D. Grove, J. Dean, C. Garrett, and C. Chambers. Proleguided receiver class prediction. In ACM Conference on ObjectOriented Programming Systems, Languages, and Applications, pages 108-123, Oct. 1995.

....leave at Watson. Rutgers University Technical Report DCS TR 424, November 2000 mation is used to decide not only what to optimize, but how to optimize. There exists a large body of work on collecting of ine pro les [3, 11, 12, 16, 27] as well as optimizations based on o ine pro les [6, 17, 18, 20, 21, 28]. Some systems [5,10,22,23,32] apply limited forms of online feedbackdirected optimizations, however most of the o ine work mentioned above has not yet been applied in fully automated online systems. The main diculty in applying these optimizations online is that they often rely on instrumenting ....

....in applying these optimizations online is that they often rely on instrumenting the code to collect detailed information about program execution, and instrumentation can cause substantial performance degradation. Overheads in the range of 30 1,000 above non instrumented code is not uncommon [3,11,12,17,18,28], and overheads in the range of 10,000 (100 times slower) has been reported [17] An online system needs to execute instrumented code for some period of time, prior to performing optimization. The overhead introduced by instrumentation makes this task dif cult to perform for several reasons. ....

D. Grove, J. Dean, C. Garrett, and C. Chambers. Proleguided receiver class prediction. In ACM Conference on ObjectOriented Programming Systems, Languages, and Applications, pages 108-123, Oct. 1995.

....by program specialization, but without requiring user guidance. To reduce the complexity of performing analysis, simplied type inference algorithms such as Class Hierarchy Analysis are used [12] combined with prole information that guides speculative optimizations such as receiver prediction [18, 20]. Since techniques such as inlining and specialization for types (customization [7] and method argument specialization[10] can cause code explosion, the same proling information is used to focus these optimizations on the critical parts of the program [11, 20] The optimizations ooeered by such ....

D. Grove, J. Dean, C. Garrett, and C. Chambers. Prole-guided receiver class prediction. In Proceedings of OOPSLA '95, pages 108123, Austin, TX, October 1995.

....callees is limited, some compilers replace a virtual call by a conditional that selects at call time which callee to invoke with a direct call. This transformation can be directed using automatically gathered prole information, and enables further optimizations through inlining or customization [17, 20]. However, the cost of a runtime decision remains, and the control AEow is only somewhat simplied. These observations are illustrated by the the benchmarks reported in Section 6. Using state of the art Java compiler technology, we found that programs written using design patterns that operate ....

D. Grove, J. Dean, C. Garrett, and C. Chambers. Prole-guided receiver class prediction. In OOPSLA'95 Conference Proceedings, volume 30, 10, pages 108123, New York, October 1995. ACM Press.

....guarding) since analysis has revealed the exact target. To inline selected potential targets of a virtual call, compilers can perform various forms of guarded inlining. The compiler can decide which targets to speculatively inline at a call site using static heuristics [13, 8] pro le information [24, 21], and or static examination of the program s class hierarchy [7, 11] Jalape no uses both class tests and method tests [12] to perform guarded inlining of virtual calls based on both class hierarchy analysis and on line pro le information. 10 In addition to determining which calls are legal to ....

D. Grove, J. Dean, C. Garrett, and C. Chambers. Prole-guided receiver class prediction. In ACM Conference on Object-Oriented Programming Systems, Languages, and Applications, pages 108-123, Oct. 1995.

No context found.

D. Grove, J. Dean, C. Garrett, and C. Chambers. Pro le-guided receiver class prediction. In ACM Conference on Object-Oriented Programming Systems, Languages, and Applications, pages 108-123, Oct. 1995.

Online articles have much greater impact   More about CiteSeer.IST   Add search form to your site   Submit documents   Feedback  

CiteSeer.IST - Copyright Penn State and NEC