This chapter will introduce the floating-point instructions in the non-privileged subset foundation of the LoongArch architecture. It will define the functional aspects of the basic floating-point instructions in accordance with the IEEE 754-2008 standard.
The basic floating-point instruction cannot be implemented separately from the basic integer instruction. Both the basic integer instruction and the basic floating-point instruction need to be implemented simultaneously. Whether the implementation of basic floating-point instructions includes instructions for manipulating double-precision floating-point numbers and doubleword integers is independent of whether the architecture is LA32 or LA64.
8.1.2-Base_Floating-Point_Instructions/8.1.2.1-Floating-Point_Arithmetic_Operation_Instructions.adoc