Macro safe_arch::mul_i64_carryless_m128i

macro_rules! mul_i64_carryless_m128i {
    ($a:expr, lane 0, $b:expr, lane 0) => { ... };
    ($a:expr, lane 1, $b:expr, lane 0) => { ... };
    ($a:expr, lane 0, $b:expr, lane 1) => { ... };
    ($a:expr, lane 1, $b:expr, lane 1) => { ... };
}

This is supported with target feature pclmulqdq only.

Performs a "carryless" multiplication of two i64 values.

You specify m128i expressions and then , lane 0 or , lane 1 for each one to select which of the two i64 lanes is used in the multiplication.

let x = m128i::from([2_i64, 3]);
let y = m128i::from([4_i64, 500]);
//
let c: [i64; 2] = mul_i64_carryless_m128i!(x, lane 0, y, lane 0).into();
assert_eq!(c, [8_i64, 0]);
let c: [i64; 2] = mul_i64_carryless_m128i!(x, lane 1, y, lane 0).into();
assert_eq!(c, [12_i64, 0]);
let c: [i64; 2] = mul_i64_carryless_m128i!(x, lane 0, y, lane 1).into();
assert_eq!(c, [1000_i64, 0]);
let c: [i64; 2] = mul_i64_carryless_m128i!(x, lane 1, y, lane 1).into();
assert_eq!(c, [540_i64, 0]); // not 1500 like a normal mul would be!