calculate_dispatch_size()

std::array< uint32_t, 3 > MayaFlux::Buffers::VertexFieldProcessor::calculate_dispatch_size ( const std::shared_ptr< VKBuffer > &  buffer )

overrideprotectedvirtual

One thread per record in the resolved range.

The generated kernel guards its own tail, so rounding up is safe.

Reimplemented from MayaFlux::Buffers::ComputeProcessor.

Definition at line 212 of file VertexFieldProcessor.cpp.

{
    const auto& dispatch = get_dispatch_config();
    const uint32_t local_x = std::max(1U, dispatch.workgroup_x);
    const uint32_t groups = (m_params.vertex_count + local_x - 1) / local_x;
 
    return { std::max(1U, groups), 1U, 1U };
}

References MayaFlux::Buffers::ComputeProcessor::get_dispatch_config(), m_params, and MayaFlux::Buffers::VertexFieldProcessor::RangeParams::vertex_count.

Here is the call graph for this function: