Enum GpuEligibility 

pub enum GpuEligibility {
    BackendNotCompiled,
    WorkloadBelowThreshold,
    Eligible,
}

Joint eligibility state for a GPU kernel at the call site.

Callers construct exactly one variant, which encodes both the compile-time backend presence and the runtime workload threshold check. Replacing the former (supported: bool, large_enough: bool) pair removes the possibility of silently swapping the two flags at a call site: each meaningful state has exactly one constructor and the match inside decide is total.

Variants

BackendNotCompiled

Vendor backend is not compiled into this build for this kernel.

WorkloadBelowThreshold

Backend is compiled in, but the workload (n, m, …) is below the runtime threshold for this kernel.

Eligible

Backend is compiled in and the workload is large enough; the only remaining gates are policy and runtime probe.

Implementations

impl GpuEligibility

pub const fn from_flags(supported: bool, large_enough: bool) -> Self

Combine the compile-time backend flag with the workload predicate into the canonical joint state. Use this only when you genuinely have two independent booleans; otherwise prefer constructing a variant directly.

Trait Implementations

impl Clone for GpuEligibility

fn clone(&self) -> GpuEligibility

Returns a duplicate of the value.

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

Performs copy-assignment from source.

impl Copy for GpuEligibility

impl Debug for GpuEligibility

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

Formats the value using the given formatter.

impl Eq for GpuEligibility

impl PartialEq for GpuEligibility

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

Tests for self and other values to be equal, and is used by ==.

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

Tests for !=. The default implementation is almost always sufficient, and should not be overridden without very good reason.

impl StructuralPartialEq for GpuEligibility