Expand description
Auto-generated module
🤖 Generated with SplitRS
Structs§
- Buffer
Binding - A binding entry associating a buffer with a binding index and usage.
- Buffer
Handle - Opaque handle to a GPU/CPU buffer (usize-indexed, used by ComputeBackend).
- Buffer
Id - Newtype handle for GPU buffers in the dispatcher model.
- Compute
Dispatcher - Manages GPU buffers and dispatches parallel map/reduce operations.
- Compute
Pass - Records a sequence of kernel dispatches for batched execution.
- CpuBackend
- CPU fallback compute backend.
- GpuBuffer
- A CPU-resident buffer that mimics a GPU storage buffer.
- GpuCommand
Encoder - A recorded sequence of GPU commands (mock encoder).
- Kernel
Spec - Specification for a GPU compute kernel dispatch.
- Memory
Bandwidth Model - GPU memory bandwidth model.
- Occupancy
Model - GPU occupancy model (simplified).
- Resource
Lifecycle - Tracks buffer lifecycle (creation, writes, reads) for debugging.
- Timeline
Semaphore - A mock GPU timeline semaphore for synchronising multi-pass GPU work.
- Warp
Divergence Record - A record of divergent branches observed in a kernel.
Enums§
- Buffer
Usage - Describes how a buffer is used in a compute pass.
- GpuCommand
- A single GPU command entry.
- GpuError
- Errors that can occur during GPU (or CPU-fallback) dispatch.
- Pipeline
Barrier - Specifies the type of pipeline barrier needed between passes.
- Resource
Event - A single resource event for lifecycle tracking.