Enum ModelCheck 

#[non_exhaustive]
pub enum ModelCheck {
    ModelId,
    ParamIds,
    StatLabels,
    ActionIds,
    Palette,
    Metadata,
    DefaultSetup,
    DefaultsFit,
    ApplyModes,
    Views,
    ParallelJobs,
    Actions,
    StatCount,
    ThreadCount,
    SameSeed,
    SeedSensitivity,
    SamplingCadence,
    BaselineBuild,
    FullSubmission,
    SampledSlice,
}

Available on crate features testing only.

One contract of a model entry that the kit checks.

Variants (Non-exhaustive)

Non-exhaustive enums could have additional variants added in future. Therefore, when matching against variants of non-exhaustive enums, an extra wildcard arm must be added to account for any future variants.

ModelId

The id meets the grammar of model ids.

ParamIds

Parameter ids are unique, and none repeats an id the engine prepends. Each is used on the command line and in a design table, so it is not empty, holds no whitespace and no =, and does not start with action..

StatLabels

Stat labels are unique.

ActionIds

Action ids are unique. Each is used on the command line and in a design table, so it is not empty, and holds no whitespace and no =.

Palette

A declared palette has colours.

Metadata

The backend, the structure, the topology hint and the device demand agree.

DefaultSetup

The declared defaults pass henad_compute::simulation::RunSetup::from_parts, as the app’s Build checks them.

DefaultsFit

A GPU model’s declared defaults fit wgpu::Limits::default(), the WebGPU baseline a browser offers.

ApplyModes

Each declared apply mode matches what set_param accepts.

Views

The factory returns the declared backend, and its views or GPU layers match the topology hint.

ParallelJobs

Only a CPU model reports parallel_jobs, and never zero jobs.

Actions

Every declared action is accepted, and the index after the last action is rejected.

StatCount

Every declared stat gets a value. A model that returns more values than it declares passes, since the engine drops the extra values before the check sees them.

ThreadCount

A CPU model’s stats and exported state are the same at the low and the high thread count.

SameSeed

Two builds on one seed agree.

SeedSensitivity

Two seeds differ in some stat.

SamplingCadence

A run sampled every tick ends where a run sampled every seventh tick ends.

BaselineBuild

The model builds on the device at its declared defaults, and its declared demand fits the device.

FullSubmission

One submission of MAX_STEPS_PER_SUBMISSION steps executes every step. A model that replays exactly reads back the same stats as single steps, and any other model that declares stats reads back some stat that is not zero.

SampledSlice

A sampled slice reads back the same stats as a snapshot.

Implementations

impl ModelCheck

pub const ALL: &'static [Self]

Every check, in the order check_model runs them.

Trait Implementations

impl Clone for ModelCheck

fn clone(&self) -> Self

Returns a duplicate of the value.

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

Performs copy-assignment from source.

impl Copy for ModelCheck

impl Debug for ModelCheck

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

Formats the value using the given formatter.

impl Display for ModelCheck

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

Formats the value using the given formatter.

impl Eq for ModelCheck

impl Hash for ModelCheck

fn hash<__H: Hasher>(&self, state: &mut __H)

Feeds this value into the given Hasher.

fn hash_slice<H>(data: &[Self], state: &mut H)

where H: Hasher, Self: Sized,

Feeds a slice of this type into the given Hasher.

impl PartialEq for ModelCheck

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

Equality operator ==.

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

Inequality operator !=.

impl StructuralPartialEq for ModelCheck