Module emit 

Re-exports

pub use coopmat::matmul_tc_wgsl;

pub use coopmat::matmul_tc_wgsl_tiled;

pub use coopmat::MATMUL_TC_TILED_ENTRY;

pub use cuda_c::emit_cuda_c;

pub use cuda_graph::emit_graph_cuda_c;

pub use cuda_graph::graph_cuda_entry;

pub use cuda_graph::CudaCLowerer;

pub use df64::GEMM_DF64_ENTRY;

pub use df64::GEMM_DF64_WGSL;

pub use dxc::compile_hlsl_to_spirv;

pub use dxc_cache::clear_dxc_cache;

pub use dxc_cache::compile_hlsl_to_spirv_cached;

pub use dxc_cache::dxc_cache_len;

pub use graph_hlsl::conv2d_hlsl;

pub use graph_hlsl::emit_graph_hlsl;

pub use graph_hlsl::HlslLowerer;

pub use graph_msl::conv2d_msl;

pub use graph_msl::emit_graph_msl;

pub use graph_msl::MslLowerer;

pub use hlsl::emit_hlsl;

pub use msl::emit_msl;

pub use ptx::emit_ptx;

pub use spirv::decode_spirv_words;

pub use spirv::emit_spirv;

pub use spirv::emit_spirv_patched;

pub use spirv::patch_spirv_workgroup_size;

pub use wgsl::emit_graph_wgsl;

pub use wgsl::emit_wgsl;

Modules

coopmat

Cooperative-matrix (tensor-core) WGSL emission (plan §18).

cuda_c

CUDA-C emission, compiled to PTX by NVRTC at runtime (mirrors the HLSL -> DXC -> DXIL path). Storage buffers become pointer parameters in binding order; a uniform block is passed by value as the last parameter. This is what the native CUDA backend executes for the differential oracle.

cuda_c_fused

Fused CUDA-C kernels for the mega-pass — combine multiple kernel launches into one to reduce launch overhead and intermediate global memory traffic.

cuda_graph

CUDA-C lowering of the compute-graph IR — the second backend in one pass (plan §4, Phase 5). The same ComputeGraph that emit_graph_wgsl lowers to WGSL is lowered here to CUDA-C by a CudaCLowerer, with no per-kernel.id branches — both backends are just Lowerer impls walked by the one lower_graph driver.

df64

Double-single (df64) emulated double-precision GEMM in raw WGSL.

dxc

dxc_cache

DXC SPIR-V binary cache — avoids re-running DXC on identical HLSL source.

graph_hlsl

HLSL lowering of the portable compute-graph nodes (plan §4 / P7) — the fourth backend the one lower_graph driver walks. HlslLowerer implements the same Lowerer trait, emitting HLSL compute shaders for the portable native kit (Elementwise/Reduce/Broadcast) with the same binding ABI + math as the WGSL graph_ops kernels.

graph_msl

MSL (Metal) lowering of the portable compute-graph nodes (plan §4 / P7) — the third backend the one lower_graph driver walks, after WGSL and CUDA-C. MslLowerer implements the same Lowerer trait, emitting Metal for the portable native kit (Elementwise/Reduce/Broadcast) — same binding ABI + math as the WGSL graph_ops kernels, so a portable graph lowers to Metal with no per-id branch.

hlsl

hlsl_wave

HLSL wave-intrinsic decode kernel emitters.

msl

ptx

spirv

SPIR-V emission target.

wgsl

Structs

GeneratedShader

Enums

TargetBackend

Functions

emit_shader