ARM Assembly Language, 2nd Edition

Chapter 5

Loads, Stores, and Addressing

5.1 Introduction

Processor architects spend a great deal of time analyzing typical routines on simulation models of a processor, often to find performance bottlenecks. Dynamic instruction usage gives a good indication of the types of operations that are performed the most while code is running. This differs from static usage that only describes the frequency of an instruction in the code itself. It turns out that while typical code is running, about half of the instructions deal with data movement, including data movement between registers and memory. Therefore, loading and storing data efficiently is critical to optimizing processor performance. As with all RISC processors, dedicated instructions are ...