David R. Kaeli
Miriam E. Leeser, Guevara Noubir
Date of Award
Master of Science
Department or Academic Unit
College of Engineering. Department of Computer and Electrical Engineering.
computer & electrical engineering, binary instrumentation, DSP, embedded system, profile
Embedded computer systems--Programming
Electrical and Computer Engineering
Program analysis and simulation tools have been demonstrated to be valuable in the hardware or software analysis of general purpose microprocessors. Techniques like profiling and instrumentation enable software and hardware designers to make decisions on various design trade-offs in a timely and cost-effective manner. While a large number of program profiling and instrumentation tools have been developed to support hardware and software analysis on general purpose systems, there is a general lack of sophisticated tools available for embedded architectures. Embedded systems are typically sensitive to performance bottlenecks, memory leaks and other software inefficiencies. There is a growing need for better tools in this rapidly growing design space.
In this thesis we describe DSPInst, a binary instrumentation tool for the Analog Device’s Blackfin family of Digital Signal Processors (DSPs). DSPInst provides finegrained control over the execution of programs. Instrumentation tool users are able to gain transparent access to the processor and memory state before or after every executed instruction, without perturbing the architected program state.
DSPInst provides a platform for building a wide range of customized analysis tools at the instruction level granularity. To demonstrate the utility of this toolset, we present three example analysis tools: 1) a utility to count the number of 64-bit instruction executed by the DSP, 2) a profiling tool to collect information for loop execution, and 3) the use of voltage and frequency scaling on any instruction boundary.
Sun, Enqiang, "A binary instrumentation tool for the Blackfin processor" (2009). Electrical and Computer Engineering Master's Theses. Paper 30. http://hdl.handle.net/2047/d20000035
Click button above to open, or right-click to save.