Available on (x86 or x86-64) and target feature
avx and x86-64 only.Expand description
Copies a to result, then inserts 128 bits (composed of 4 packed
single-precision (32-bit) floating-point elements) from b into result
at the location specified by imm8.