Table of Contents

Enum CudaRingKernelCompiler.CompilationStage

Namespace
DotCompute.Backends.CUDA.RingKernels
Assembly
DotCompute.Backends.CUDA.dll

Compilation stages for Ring Kernel PTX generation.

public enum CudaRingKernelCompiler.CompilationStage

Fields

Analysis = 2

Stage 2: Extract method signature, parameters, and types.

CudaGeneration = 3

Stage 3: Generate CUDA C++ kernel code.

Discovery = 1

Stage 1: Find [RingKernel] method via reflection.

ModuleLoad = 5

Stage 5: Load PTX module into CUDA context.

None = 0

No stage (default value).

PTXCompilation = 4

Stage 4: Compile CUDA → PTX with NVRTC.

Verification = 6

Stage 6: Verify function pointer retrieval.