ASI add-on for existing OptiScaler installs

Add an optional D4R_INSTALL_MODE=asi release (d4r-<version>-asi.zip)
holding only d4r.asi and a companion d4r/ folder, to drop into an
existing OptiScaler plugin directory.

- Shim: export InitializeASI. It identifies OptiScaler by its PE export
  name to find the active INI, then launches d4r-setup.exe out of
  process (no GPU or network work under the loader lock). The .asi
  build loads its runtime and config from the companion d4r/ folder.

- tools/d4r_asi_setup.cpp: native helper. It waits for the game to
  exit, installs the D3D12 wrapper and the private d4r12core.dll beside
  the EXE, and merges the NGX route into the loader's INI (including
  GE-Proton's prefix-local INI) with backups. Unchanged launches do
  nothing. It skips Proton helpers and background utilities and runs
  setup for one plugin install at a time.

- scripts/private_d3d12_core.py: points vkd3d's d3d12.dll forwarder at
  d4r12core.dll without moving code addresses.

- install.sh runs from inside the ASI's d4r/ as well as the proxy
  release root. d4r-check.sh checks ASI installs (loader, plugin path,
  LoadAsiPlugins, HookOriginalNvngxOnly, Wine drive paths).

- Packaging, build script, docs, ASI_README.txt and tests.

packaging/RELEASE_NOTES.md

@@ -1,6 +1,10 @@
 <!-- Modified in this fork for CUDA Ray Reconstruction support and validation (2026). -->
 # Unreleased
 
+- ASI setup ignores Proton accessibility helpers and other background utilities, and serializes setup per plugin installation. Its D3D12 configuration uses regular Windows library hooks and disables Agility SDK upgrades. Game-specific startup remains a separate runtime check.
+
+- ASI setup now finds the INI beside the active OptiScaler loader, including GE-Proton's `system32/umu` installation. Missing-file errors identify the actual missing path, and INI updates are staged beside their destination even when the game and prefix are on different filesystems.
+- Optional OptiScaler ASI add-on (`D4R_INSTALL_MODE=asi`): drop `d4r.asi` and its companion `d4r/` folder into the existing plugin directory, run its `install.sh` once for NVIDIA's DLLs, then launch. The ASI installs the supplied D3D12 wrapper/private core beside the EXE and configures NGX after the game closes, with backups. The runtime stays in the plugin folder; no Python is needed. Game rendering still requires validation.
 - Native engine, preset L (DLSS 4.5, render preset 12): the engine now runs L ([Engine] Enabled, models in `engine/l` and `engine/l-ldr`, built by `engine/compile_m.py` from a capture of preset L; `scripts/install_d4r_runtime.sh` stages `D4R_ENGINE_MODEL_L` / `D4R_ENGINE_MODEL_L_LDR`). Every stage, `d4r-m` on captured frames and the D3D12 harness are byte-identical to the CUDA backend running NVIDIA's translated kernels, apart from isolated single-frame knife edges (96-122 dB). 1280x720 -> 3840x2160 takes 3.8 ms on an RX 9070 XT. Display-resolution motion vectors keep the CUDA backend, as for M. Details: `engine/L_NOTES.md`.
 - Native engine fixes found with L, which also apply to M (rebuild M packages with the current `compile_m.py`): the encoders' patch merge is clamped to ±2π as NVIDIA's kernels do; the automatic exposure is measured on NGX's grid at Ultra Performance (half the render size, at least 256, when the output is more than twice the render size; half the output was used, which changed every value downstream); the downsample's filter weights are exact (glslang rounded them toward zero: 92% of the output values were identical, now all); and the network uses the summation order of NVIDIA's kernels as ZLUDA runs them. M is now byte-identical to that CUDA backend in the harness too, where it measured 36-41 dB.
 - `scripts/check_redistributable.py` refuses NVIDIA's binaries, NGX SDK headers, PTX and code objects, d4r's PTX translations and the engine's model data. `scripts/package_release.sh` runs it on every release before zipping, and `git config core.hooksPath scripts/git-hooks` runs it before each commit.