ARM Assembly Language, 2nd Edition

Chapter 11

Floating-Point Data-Processing Instructions

11.1 Introduction

Floating-point operations are not unlike their integer counterparts. The basic arithmetic operations are supported, such as add and subtract, multiply and multiply–accumulate, and divide. Three are unique, however, and they are negate, absolute value, and square root. You will also notice that the logic operations are missing. There are no floating-point Boolean instructions, and no bit manipulation instructions. For these operations, should you need them, integer instructions may be used once the floating-point operand is moved into an ARM register with a VMOV instruction.

Floating-point performance is measured in flops, or floating-point operations per second. Only arithmetic ...