Module indirect_dispatch 

Indirect dispatch support for GPU compute kernels.

Indirect dispatch allows the GPU to determine workgroup counts from a buffer rather than from CPU-side values. This is essential for data-dependent workloads where the number of items is computed by a previous GPU pass (e.g., compaction, prefix-sum driven dispatch).

Structs

IndirectBuffer

A buffer that holds indirect dispatch arguments.

IndirectDispatchArgs

Arguments for an indirect compute dispatch, matching the layout expected by wgpu::RenderPass::dispatch_workgroups_indirect.

Enums

DispatchStrategy

Strategy for computing indirect dispatch arguments from an element count and a workgroup size.

Functions

compute_dispatch

Compute dispatch arguments from an element count and strategy.

validate_dispatch_limits

Validates that dispatch arguments do not exceed device limits.