An input-centric paradigm for program dynamic optimizations

Abstract

Accurately predicting program behaviors (e.g., locality, dependency, method calling frequency) is fundamental for program optimizations and runtime adaptations. Despite decades of remarkable progress, prior studies have not systematically exploited program inputs, a deciding factor for program behaviors.Triggered by the strong and predictive correlations between program inputs and behaviors that recent studies have uncovered, this work proposes to include program inputs into the focus of program behavior analysis, cultivating a new paradigm named input-centric program behavior analysis. This new approach consists of three components, forming a three-layer pyramid. At the base is program input characterization, a component for resolving the complexity in program raw inputs and the extraction of important features. In the middle is input-behavior modeling, a component for recognizing and modeling the correlations between characterized input features and program behaviors. These two components constitute input-centric program behavior analysis, which (ideally) is able to predict the large-scope behaviors of a program's execution as soon as the execution starts. The top layer of the pyramid is input-centric adaptation, which capitalizes on the novel opportunities that the first two components create to facilitate proactive adaptation for program optimizations.By centering on program inputs, the new approach resolves a proactivity-adaptivity dilemma inherent in previous techniques. Its benefits are demonstrated through proactive dynamic optimizations and version selection, yielding significant performance improvement on a set of Java and C programs.

Microarchitecture-Aware Code Generation for Deep Learning on Single-ISA Heterogeneous Multi-Core Mobile Processors

“…The selector can be realized with a few conditional statements followed by additional function calls. Conventional input-aware FMV schemes also employ a runtime selector in the flow, but the selector must analyze input data in detail to determine an optimal version, which translates to large processing overheads [18], [19]. However, our runtime selector chooses a version by simply looking at the implementer (IMP) and identification (IDCODE) information retrieved from the PMU.…”

Space-efficient multi-versioning for input-adaptive feedback-driven program optimizations

“…Berube and his colleagues provide a compilercentric clustering approach to reduce the number of representative workloads [1]. Tian and others [22] have proposed an input-centric program dynamic optimization framework, which offers a systematic way to include program inputs into the dynamic optimization process. They later [24] eliminated the needs for offline training, making the framework deployable in a way completely transparent to users.…”

“…Often, different inputs may prompt the program to behave very differently. As a consequence, the code produced by FDO on one training run may work inferiorly on a different input [2,10,13,20,21,23,26]. Such input sensitivity is especially prominent in arising data-driven computing (e.g., business analytics), the data in which show increasing variety and complexity.…”

Automatic Generation of Program Affinity Policies Using Machine Learning

“…Some programs have changing behavior across inputs. Techniques exist to model and predict how an input will affect a program's behavior [12]. Such techniques could be integrated into the AutoFinity flow, and enable the reuse of learned program behavior.…”