Analysis of input-dependent program behavior using active profiling

Authors:
Xipeng Shen;Chengliang Zhang;Chen Ding;Michael L. Scott;Sandhya Dwarkadas;Mitsunori Ogihara
Affiliations:
Computer Science Department, The College of William and Mary;Computer Science Department, University of Rochester;Computer Science Department, University of Rochester;Computer Science Department, University of Rochester;Computer Science Department, University of Rochester;Computer Science Department, University of Rochester
Venue:
ecs'07 Experimental computer science on Experimental computer science
Year:
2007

Citing 31
Cited 0

Memory access patterns of parallel scientific programs

SIGMETRICS '87 Proceedings of the 1987 ACM SIGMETRICS conference on Measurement and modeling of computer systems
Design of the opportunistic garbage collector

OOPSLA '89 Conference proceedings on Object-oriented programming systems, languages and applications
A static performance estimator to guide data partitioning decisions

PPOPP '91 Proceedings of the third ACM SIGPLAN symposium on Principles and practice of parallel programming
ATOM: a system for building customized program analysis tools

PLDI '94 Proceedings of the ACM SIGPLAN 1994 conference on Programming language design and implementation
A program data flow analysis procedure

Communications of the ACM
Memory hierarchy reconfiguration for energy and performance in general-purpose processor architectures

Proceedings of the 33rd annual ACM/IEEE international symposium on Microarchitecture
Managing multi-configuration hardware via dynamic working set analysis

ISCA '02 Proceedings of the 29th annual international symposium on Computer architecture
Parallel performance prediction using lost cycles analysis

Proceedings of the 1994 ACM/IEEE conference on Supercomputing
SPEC CPU2000: Measuring CPU Performance in the New Millennium

Computer
Scalable analysis techniques for microprocessor performance counter metrics

Proceedings of the 2002 ACM/IEEE conference on Supercomputing
The design, implementation, and evaluation of a compiler algorithm for CPU energy reduction

PLDI '03 Proceedings of the ACM SIGPLAN 2003 conference on Programming language design and implementation
Measurements of major locality phases in symbolic reference strings

SIGMETRICS '76 Proceedings of the 1976 ACM SIGMETRICS conference on Computer performance modeling measurement and evaluation
A measure of program locality and its application

SIGMETRICS '84 Proceedings of the 1984 ACM SIGMETRICS conference on Measurement and modeling of computer systems
An Executable Analytical Performance Evaluation Approach for Early Performance Prediction

IPDPS '03 Proceedings of the 17th International Symposium on Parallel and Distributed Processing
Profile-based dynamic voltage and frequency scaling for a multiple clock domain microprocessor

Proceedings of the 30th annual international symposium on Computer architecture
Positional adaptation of processors: application to energy reduction

Proceedings of the 30th annual international symposium on Computer architecture
Dynamically managing the communication-parallelism trade-off in future clustered processors

Proceedings of the 30th annual international symposium on Computer architecture
Phase tracking and prediction

Proceedings of the 30th annual international symposium on Computer architecture
The Search for Lost Cycles: A New Approach to Parallel Program Performance Evaluation

The Search for Lost Cycles: A New Approach to Parallel Program Performance Evaluation
Characterizing and Predicting Program Behavior and its Variability

Proceedings of the 12th International Conference on Parallel Architectures and Compilation Techniques
EXPERT: expedited simulation exploiting program behavior repetition

Proceedings of the 18th annual international conference on Supercomputing
Low-overhead memory leak detection using adaptive statistical profiling

ASPLOS XI Proceedings of the 11th international conference on Architectural support for programming languages and operating systems
Locality phase prediction

ASPLOS XI Proceedings of the 11th international conference on Architectural support for programming languages and operating systems
Method-level phase behavior in java workloads

OOPSLA '04 Proceedings of the 19th annual ACM SIGPLAN conference on Object-oriented programming, systems, languages, and applications
Gated memory control for memory monitoring, leak detection and garbage collection

Proceedings of the 2005 workshop on Memory system performance
Online Phase Detection Algorithms

Proceedings of the International Symposium on Code Generation and Optimization
Selecting Software Phase Markers with Code Structure Analysis

Proceedings of the International Symposium on Code Generation and Optimization
Generating object lifetime traces with Merlin

ACM Transactions on Programming Languages and Systems (TOPLAS)
Bell: bit-encoding online memory leak detection

Proceedings of the 12th international conference on Architectural support for programming languages and operating systems
Garbage collection hints

HiPEAC'05 Proceedings of the First international conference on High Performance Embedded Architectures and Compilers
An approach to performance prediction for parallel applications

Euro-Par'05 Proceedings of the 11th international Euro-Par conference on Parallel Processing

Quantified Score

Hi-index	0.00

Visualization

Abstract

Utility programs, which perform similar and largely independent operations on a sequence of inputs, include such common applications as compilers, interpreters, and document parsers; databases; and compression and encoding tools. The repetitive behavior of these programs, while often clear to users, has been difficult to capture automatically. We present an active profiling technique in which controlled inputs to utility programs are used to expose execution phases, which are then marked, automatically, through binary instrumentation, enabling us to exploit phase transitions in production runs with arbitrary inputs. Experiments with five programs from the SPEC benchmark suites show that phase behavior is surprisingly predictable in many (though not all) cases. This predictability can in turn be used for optimized memory management leading to significant performance improvement.