Gaussian blur (three box passes each way, running sums per 16-pixel chunk), the
separable resample with the CPU's exact taps (uploaded tables), the guided filter
and the fast guided filter (coefficients on the subsampled grid), log luminance,
dark channel + airlight (percentile of every 7th value read back), white balance
and luminance/colour noise reduction now run as WGSL kernels. A render records
everything into one command encoder and reads back only the output (and the
airlight samples); defringe and spot removal still run on the CPU when used.
Equivalence unchanged: max 1 LSB, mean < 0.001 LSB over all test cases.
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
New L3 crate `lightcraft-gpu`: a wgpu device (Metal/Vulkan/DX12; none on wasm32),
WGSL kernels compiled once, buffer/dispatch/readback helpers, and a renderer that
runs the per-pixel stage (tone map, highlights/shadows, clarity/texture/sharpen,
dehaze, local mask terms, OkLCh colour ops, grading, vignette, gamut map, sRGB
encode, curves, grain) on the device. Earlier stages still run on the CPU and are
uploaded (per-stage hybrid). Stage results are cached per view next to the CPU
StageCache (new `StageCache::extension`).
The CPU pipeline stays the oracle and shares its resolved parameters with the
kernels: `finish::FinishParams` / `mask_terms`, `pipeline::plan` (stage-cache keys),
`local::plane_sigmas` / `nr_params` / `wb_matrix_for`, exact LUTs and OkLab
matrices exported for the shader prelude. Equivalence tests (skip without an
adapter): mean |d| < 0.5 LSB, max <= 3 LSB; measured max 1 LSB on Apple M4 Pro.
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
