Function core::arch::x86_64::_mm_insert_ps

pub unsafe fn _mm_insert_ps(a: __m128, b: __m128, imm8: i32) -> __m128

This is supported on x86-64 and target feature sse4.1 only.

Select a single value in a to store at some position in b, Then zero elements according to imm8.

imm8 specifies which bits from operand a will be copied, which bits in the result they will be copied to, and which bits in the result will be cleared. The following assignments are made:

• Bits [7:6] specify the bits to copy from operand a:

  • 00: Selects bits [31:0] from operand a.
  • 01: Selects bits [63:32] from operand a.
  • 10: Selects bits [95:64] from operand a.
  • 11: Selects bits [127:96] from operand a.

• Bits [5:4] specify the bits in the result to which the selected bits from operand a are copied:

  • 00: Copies the selected bits from a to result bits [31:0].
  • 01: Copies the selected bits from a to result bits [63:32].
  • 10: Copies the selected bits from a to result bits [95:64].
  • 11: Copies the selected bits from a to result bits [127:96].

• Bits [3:0]: If any of these bits are set, the corresponding result element is cleared.

Intel's documentation