Struct SyncFullOrderedLoading

pub struct SyncFullOrderedLoading {}

Similar to sync_full_tilewise, but includes additional validation checks.

This function operates only on the LHS (left-hand side).

• In the single-row case, behavior is similar to tilewise with row-major tiling order. However, it will explicitly fail if any plane does not load its entire row.
• In the multi-row case, it too will fail if a plane does not load all its rows. Within each plane, the local tiling order is column-major.

Trait Implementations

impl Clone for SyncFullOrderedLoading

fn clone(&self) -> SyncFullOrderedLoading

Returns a duplicate of the value.

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

Performs copy-assignment from source.

impl CubeType for SyncFullOrderedLoading

type ExpandType = SyncFullOrderedLoadingExpand

fn into_mut(scope: &mut Scope, expand: Self::ExpandType) -> Self::ExpandType

Wrapper around the init method, necessary to type inference.

impl LoadMaxRoundPlaneCount for SyncFullOrderedLoading

fn max_round_plane_count( tiling_scheme: &TilingScheme, ident: InputIdent, _line_size: u8, _plane_dim: u32, ) -> u32

Returns the largest number of planes that evenly divides the tasks.

impl LoadingValidation for SyncFullOrderedLoading

fn check<C: GlobalConfig>( config: &C, ident: Ident, ) -> Result<(), InvalidConfigError>

Verify that configs are valid for a loader, otherwise return an error stating why

impl SyncFullLoadingStrategy for SyncFullOrderedLoading

type TilingLayout = ContiguousTilingLayout<OrderedTilingOrder>

The layout describing how data is tiled across the stage.

type Job<MP: MatmulPrecision> = SyncFullTilewiseJob

The LoadingJob for this strategy.

fn new_job<MP: MatmulPrecision, G: GlobalConfig>( input_ident: InputIdent, config: G, ) -> Self::Job<MP>

Returns the job with preliminary calculations done.

fn __expand_new_job<MP: MatmulPrecision, G: GlobalConfig>( scope: &mut Scope, input_ident: InputIdent, config: G, ) -> <Self::Job<MP> as CubeType>::ExpandType

impl Copy for SyncFullOrderedLoading