A Tool Framework for Static and Dynamic Analysis of Object-Oriented Software with Templates

Abstract

The developers of high-performance scientific applications often work in complex computing environments that place heavy demands on program analysis tools. The developers need tools that interoperate, are portable across machine architectures, and provide source-level feedback. In this paper, we describe a tool framework, the Program Database Toolkit (PDT), that supports the development of program analysis tools meeting these requirements. PDT uses compile-time information to create a complete database of high-level program information that is structured for well-defined and uniform access by tools and applications. PDT's current applications make heavy use of advanced features of C++, in particular, templates. We describe the toolkit, focussing on its most important contribution --its handling of templates --as well as its use in existing applications.

ParaProf: A Portable, Extensible, and Scalable Tool for Parallel Performance Profile Analysis

“…The information needed for this is partly encoded in the profile event names, but ParaProf needs to have a standard means to acquire source mapping metadata (e.g., source files, and line and column position) to associate events to the program. We will apply our PDT [12] source analysis technology to this problem, and also hope to leverage the work in HPCView. In addition, source text display and interaction capabilities are required.…”

The Tau Parallel Performance System

“…To support automatic performance instrumentation at the source level, the TAU project has developed the Program Database Toolkit (PDT) (Lindlan et al 2000). The purpose of PDT, shown in Figure 4 is to parse the application source code and locate the semantic constructs to be instrumented.…”

Generating precise dependencies for large software

“…4) Templates: Template instances are not directly defined in the source code but are instantiated separately at compile time. The instantiation information is usually not available in the AST [5].…”