image

Documentation

Index

• type CUDA
  • func NewCUDAImageFromEnv(env []string) (CUDA, error)
  • func NewCUDAImageFromSpec(spec *specs.Spec) (CUDA, error)
  • func (i CUDA) GetRequirements() ([]string, error)
  • func (i CUDA) HasDisableRequire() bool
  • func (i CUDA) IsLegacy() bool

Types

type CUDA

type CUDA map[string]string

CUDA represents a CUDA image that can be used for GPU computing. This wraps a map of environment variable to values that can be used to perform lookups such as requirements.

func NewCUDAImageFromEnv

func NewCUDAImageFromEnv(env []string) (CUDA, error)

NewCUDAImageFromEnv creates a CUDA image from the input environment. The environment is a list of strings of the form ENVAR=VALUE.

func NewCUDAImageFromSpec

func NewCUDAImageFromSpec(spec *specs.Spec) (CUDA, error)

NewCUDAImageFromSpec creates a CUDA image from the input OCI runtime spec. The process environment is read (if present) to construc the CUDA Image.

func (CUDA) GetRequirements

func (i CUDA) GetRequirements() ([]string, error)

GetRequirements returns the requirements from all NVIDIA_REQUIRE_ environment variables.

func (CUDA) HasDisableRequire

func (i CUDA) HasDisableRequire() bool

HasDisableRequire checks for the value of the NVIDIA_DISABLE_REQUIRE. If set to a valid (true) boolean value this can be used to disable the requirement checks

func (CUDA) IsLegacy

func (i CUDA) IsLegacy() bool

IsLegacy returns whether the associated CUDA image is a "legacy" image. An image is considered legacy if it has a CUDA_VERSION environment variable defined and no NVIDIA_REQUIRE_CUDA environment variable defined.