rr: add FSR Ray Regeneration, exact RDNA3 inference and motion experiments

Add an experimental FSR Ray Regeneration backend for D3D12 RR requests:
- Convert depth, motion, normals and albedos into AMD denoiser inputs.
- Split and recompose diffuse/specular signals, then run Super Sampling.
- Add stability/detail controls, demodulation guards and halo limiting.
- Handle AMD provider discovery under NVIDIA GPU spoofing.
- Add runtime probes, synthetic reference scenes and per-dispatch profiling.

Add exact gfx11 DLSS-RR inference using f16 cooperative matrices:
- Reproduce FP8 conversions, accumulation and camera reprojection arithmetic.
- Widen model weights and avoid requiring VK_EXT_shader_float8.
- Extend model building, replay tools and descriptor metadata.

Expand motion preprocessing and experimental AW2 native-motion support:
- Add tiled LDS dilation, padded allocation handling and thin-only sky fill.
- Order dilation with input waits and NGX on the selected stream.
- Add jitter-aligned native composition, hybrid routing and wire history repair.
- Add prepared masks, sparse repair and independent motion/repair controls.

Add optional fused R11G11B10 input/output conversion, sampled stage timings
and profile summarization. Extend configuration, packaging, documentation,
FidelityFX headers and CPU/GPU regression coverage.

tests/test_wire_history.py

@@ -0,0 +1,54 @@
+"""Sparse wire publication rejects aliased scratch, inputs and overflowing ranges."""
+from pathlib import Path
+import subprocess
+import tempfile
+import unittest
+
+ROOT = Path(__file__).resolve().parents[1]
+
+
+class WireHistoryTests(unittest.TestCase):
+    def test_sparse_buffer_contract(self):
+        source = r'''
+#include "d4r_wire_history.h"
+#include <assert.h>
+#include <string.h>
+int main(void) {
+    D4rWireHistorySparseParams p; memset(&p,0,sizeof(p));
+    D4rWireHistoryParams* b=&p.base;
+    b->width=65;b->height=67;b->render_width=21;b->render_height=22;
+    b->current=0x10000000;b->color=0x20000000;b->native_depth=0x30000000;
+    b->motion=0x40000000;b->previous=0x50000000;b->previous_age=0x60000000;
+    b->previous_depth=0x70000000;b->output=0x80000000;b->output_age=0x90000000;
+    b->output_depth=0xa0000000;b->input_depth=0xb0000000;b->output_history=0xc0000000;
+    b->color_pitch=21*8;b->depth_pitch=b->motion_pitch=65*4;
+    b->output_pitch=65*8;b->input_depth_pitch=21*4;p.patch_mask=0xd0000000;
+    assert(sizeof(p)==176 && d4r_wire_history_sparse_parameters_valid(&p));
+    assert(!d4r_wire_history_sparse_parameters_valid(0));
+    D4rWireHistorySparseParams q=p;q.patch_mask=0;assert(!d4r_wire_history_sparse_parameters_valid(&q));
+    q=p;q.patch_mask=UINT64_MAX-16;assert(!d4r_wire_history_sparse_parameters_valid(&q));
+    q=p;q.patch_mask=q.base.current+8;assert(!d4r_wire_history_sparse_parameters_valid(&q));
+    q=p;q.patch_mask=q.base.previous_age+1;assert(!d4r_wire_history_sparse_parameters_valid(&q));
+    q=p;q.base.current=q.base.color;assert(!d4r_wire_history_sparse_parameters_valid(&q));
+    q=p;q.base.output=q.base.native_depth+4;assert(!d4r_wire_history_sparse_parameters_valid(&q));
+    q=p;q.base.current=q.base.output_history+8;assert(!d4r_wire_history_sparse_parameters_valid(&q));
+    q=p;q.base.output_pitch=UINT64_MAX-7;assert(!d4r_wire_history_sparse_parameters_valid(&q));
+    q=p;q.base.current=UINT64_MAX-1024;assert(!d4r_wire_history_sparse_parameters_valid(&q));
+    q=p;q.patch_mask=q.base.current+66*q.base.output_pitch+65*8;
+    assert(d4r_wire_history_sparse_parameters_valid(&q));
+    D4rWireHistoryPreparedParams prepared={p,0xe0000000};
+    assert(sizeof(prepared)==184 && d4r_wire_history_prepared_parameters_valid(&prepared));
+    D4rWireHistoryPreparedParams no=prepared;no.axes=0;assert(!d4r_wire_history_prepared_parameters_valid(&no));
+    no=prepared;no.axes=no.sparse.base.current;assert(!d4r_wire_history_prepared_parameters_valid(&no));
+    no=prepared;no.axes=no.sparse.patch_mask;assert(!d4r_wire_history_prepared_parameters_valid(&no));
+    no=prepared;no.axes=UINT64_MAX-16;assert(!d4r_wire_history_prepared_parameters_valid(&no));
+    return 0;
+}
+'''
+        with tempfile.TemporaryDirectory() as tmp:
+            runner = Path(tmp)/'wire-contract'
+            for compiler, language, standard in (('cc', 'c', 'c11'), ('c++', 'c++', 'c++17')):
+                subprocess.run([compiler, '-x', language, '-std='+standard, '-Wall', '-Wextra', '-Werror',
+                                '-I', str(ROOT/'tools'), '-o', str(runner), '-'],
+                               input=source, text=True, check=True, capture_output=True)
+                subprocess.run([str(runner)], check=True)