Enum CudaKernelCapabilityFailure 

pub enum CudaKernelCapabilityFailure {
    SmVersion {
        required_major: u16,
        required_minor: u16,
        actual_major: u16,
        actual_minor: u16,
    },
    ThreadsPerBlock {
        requested: u32,
        maximum: u32,
    },
    SharedMemory {
        requested: u64,
        maximum: u64,
    },
    CooperativeLaunch,
    TensorCores,
}

Capability failure reason for one CUDA launch.

Variants

SmVersion

Device SM version is below the kernel requirement.

Fields

required_major: u16

Required major version.

required_minor: u16

Required minor version.

actual_major: u16

Actual major version.

actual_minor: u16

Actual minor version.

ThreadsPerBlock

Requested block size exceeds the device limit.

Fields

requested: u32

Requested threads per block.

maximum: u32

Device maximum.

SharedMemory

Requested shared memory exceeds the device limit.

Fields

requested: u64

Requested shared memory bytes.

maximum: u64

Device maximum.

CooperativeLaunch

Cooperative launch is required but unsupported.

TensorCores

Tensor cores are required but unsupported.

Trait Implementations

impl Clone for CudaKernelCapabilityFailure

fn clone(&self) -> CudaKernelCapabilityFailure

Returns a duplicate of the value.

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

Performs copy-assignment from source.

impl Debug for CudaKernelCapabilityFailure

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

Formats the value using the given formatter.

impl PartialEq for CudaKernelCapabilityFailure

fn eq(&self, other: &CudaKernelCapabilityFailure) -> 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 Eq for CudaKernelCapabilityFailure

impl StructuralPartialEq for CudaKernelCapabilityFailure