The MIPS64 architecture includes a number of instructions that operate on pairs of single-precision floating-point values. Each pair is packed into a 64-bit floating-point register, with one element being designated the “upper half” and the other being designated the “lower half”.
GCC supports paired-single operations using both the generic vector extensions (see Using Vector Instructions through Built-in Functions) and a collection of MIPS-specific built-in functions. Both kinds of support are enabled by the -mpaired-single command-line option.
The vector type associated with paired-single values is usually
called v2sf. It can be defined in C as follows:
typedef float v2sf __attribute__ ((vector_size (8)));
v2sf values are initialized in the same way as aggregates.
For example:
v2sf a = {1.5, 9.1};
v2sf b;
float e, f;
b = (v2sf) {e, f};
Note: The CPU’s endianness determines which value is stored in
the upper half of a register and which value is stored in the lower half.
On little-endian targets, the first value is the lower one and the second
value is the upper one. The opposite order applies to big-endian targets.
For example, the code above sets the lower half of a to
1.5 on little-endian targets and 9.1 on big-endian targets.