Struct FmaAdd 

pub struct FmaAdd;

Fused-multiply-add elementwise operation: out[i] = a[i] * b[i] + a[i] (binary form).

As an ElementOp, this interprets the two operand vectors as a and b, and computes fmadd(a, b, zero) — i.e. a * b with hardware FMA precision, accumulating into zero. To use as a ternary a*b + c accumulation, call Arch::fmadd directly.

Zero-Cost Guarantee

size_of::<FmaAdd>() == 0. Monomorphization over Arch eliminates the ZST entirely; the call site reduces to a direct Arch::fmadd instruction.

Trait Implementations

impl Clone for FmaAdd

fn clone(&self) -> FmaAdd

Returns a duplicate of the value.

fn clone_from(&mut self, source: &Self)

Performs copy-assignment from source.

impl Copy for FmaAdd

impl Debug for FmaAdd

fn fmt(&self, f: &mut Formatter<'_>) -> Result<(), Error>

Formats the value using the given formatter.

impl<T> ElementOp<T> for FmaAdd

where T: NumericElement,

unsafe fn apply<Arch>( self, a: <Arch as SimdKernel<T>>::Vector, b: <Arch as SimdKernel<T>>::Vector, ) -> <Arch as SimdKernel<T>>::Vector

where Arch: SimdKernel<T>,

fmadd(a, b, zero) — uses hardware FMA where available.

Safety

Processor must support the target feature of Arch.

fn apply_scalar(self, a: T, b: T) -> T

Scalar tail: a * b (scalar mul; the addend is the implicit zero).

impl Eq for FmaAdd

impl PartialEq for FmaAdd

fn eq(&self, other: &FmaAdd) -> bool

Equality operator ==.

fn ne(&self, other: &Rhs) -> bool

Inequality operator !=.

impl StructuralPartialEq for FmaAdd