SPLASH 2014
Mon 20 - Fri 24 October 2014 Portland, Oregon, United States
Fri 24 Oct 2014 10:52 - 11:15 at Salon F - Dynamic Optimization Chair(s): Zachary Tatlock

Predicting a sequence of upcoming function calls is important for optimizing programs written in modern managed languages (e.g., Java, Javascript, C#.) Existing function call predictions are mainly built on statistical patterns, suitable for predicting a single call but not a sequence of calls. This paper presents a new way to enable call sequence prediction, which exploits program structures through Probabilistic Calling Automata (PCA), a new program representation that captures both the inherent ensuing relations among function calls, and the probabilistic nature of execution paths. It shows that PCA-based prediction outperforms existing predictions, yielding substantial speedup when being applied to guide Just-In-Time compilation. By enabling accurate, efficient call sequence prediction for the first time, PCA-based predictors open up many new opportunities for dynamic program optimizations.

Fri 24 Oct
Times are displayed in time zone: Tijuana, Baja California change

10:30 - 12:00
Dynamic OptimizationOOPSLA at Salon F
Chair(s): Zachary TatlockUniversity of Washington, Seattle
10:30
22m
Talk
Accelerating Iterators in Optimizing AST Interpreters
OOPSLA
Wei ZhangUniversity of California, Irvine, Per LarsenUniversity of California, Irvine, Stefan BrunthalerUniversity of California, Irvine, Michael FranzUniversity of California, Irvine
Link to publication
10:52
22m
Talk
Call Sequence Prediction through Probabilistic Calling Automata
OOPSLA
Zhijia ZhaoCollege of William and Mary / North Carolina State University, Bo WuCollege of William and Mary, Mingzhou ZhouCollege of William and Mary, Yufei DingCollege of William and Mary, Jianhua SunCollege of William and Mary, Xipeng ShenNorth Carolina State University, Youfeng WuIntel Corporation
Link to publication
11:15
22m
Talk
Space-Efficient Multi-Versioning for Input-Adaptive Feedback-Driven Program Optimizations
OOPSLA
Mingzhou ZhouCollege of William and Mary, Xipeng ShenNorth Carolina State University, Yaoqing GaoIBM Toronto Labs, Graham YiuIBM Toronto Labs
Link to publication
11:37
22m
Talk
The HipHop Virtual Machine
OOPSLA
Keith AdamsFacebook, Jason EvansFacebook, Bertrand MaherFacebook, Guilherme OttoniFacebook, Drew ParoskiFacebook, Brett SimmersFacebook, Edwin SmithFacebook, Owen YamauchiFacebook
Link to publication