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.
This commit is contained in:
ayo committed
1 parent
17c1828f97
commit
34e6b0fe71
83 files changed
+11300
-195
No files matched your search
@@ -0,0 +1,88 @@
|
||||
"""Opt-in GPU checks for wire color, history eligibility, padding and immutable inputs."""
|
||||
import os
|
||||
from pathlib import Path
|
||||
import struct
|
||||
import subprocess
|
||||
import tempfile
|
||||
import unittest
|
||||
|
||||
ROOT = Path(__file__).resolve().parents[1]
|
||||
|
||||
|
||||
@unittest.skipUnless(os.environ.get('D4R_RUN_GPU_TESTS') == '1', 'opt-in GPU test')
|
||||
class WireHistoryGpuTests(unittest.TestCase):
|
||||
def test_wire_source_history_reset_and_expiry(self):
|
||||
runner = os.environ['D4R_DUMP_RUNNER']
|
||||
kernel = Path(os.environ.get('D4R_WIRE_TEST_KERNEL', str(ROOT / 'build/wine-nvcuda/generated' /
|
||||
(os.environ.get('D4R_GPU_ARCH', 'gfx1101') + '.hsaco'))))
|
||||
w, h = 65, 67 # partial workgroups on both axes
|
||||
pitch, dp, mp = w*8+16, w*4+12, w*4+20
|
||||
pixels = w*h
|
||||
rgba = struct.pack('<4e', .5, .5, .5, .5)
|
||||
observed = struct.pack('<4e', .125, .125, .125, .75)
|
||||
repaired = struct.pack('<4e', .125, .125, .125, .5)
|
||||
current = b''.join(rgba*w + b'\xa5'*16 for _ in range(h))
|
||||
color = b''.join(observed*w + b'\xa5'*16 for _ in range(h))
|
||||
native = b''.join(struct.pack('<'+'f'*w, *[.5 if x == 32 else 1. for x in range(w)]) +
|
||||
b'\xa5'*12 for _ in range(h))
|
||||
motion = b''.join(b'\0'*(w*4) + b'\xa5'*20 for _ in range(h))
|
||||
for sky_only in (False, True):
|
||||
for conservative in (False, True):
|
||||
for sparse in (False, True):
|
||||
for mode in ('source', 'history', 'reset', 'expired', 'disoccluded', 'missing', 'neighbor', 'far', 'backdrop'):
|
||||
for jitter in (0, -1, 1):
|
||||
with self.subTest(mode=mode, jitter=jitter, conservative=conservative, sparse=sparse, sky_only=sky_only), tempfile.TemporaryDirectory() as tmp:
|
||||
directory = Path(tmp); output = directory/'out'; output.mkdir()
|
||||
source = mode in ('source', 'backdrop')
|
||||
history = mode in ('history', 'expired', 'disoccluded', 'neighbor')
|
||||
z = .999988 if mode == 'far' else .5
|
||||
near = b''.join(struct.pack('<'+'f'*w, *[z if x == 32 else .9 if mode == 'backdrop' else 1. for x in range(w)]) + b'\xa5'*12 for _ in range(h))
|
||||
previous = color
|
||||
if mode == 'neighbor':
|
||||
previous = b''.join(b''.join(observed if x == 32-jitter else struct.pack('<4e', .01, .01, .01, .75) for x in range(w)) + b'\xa5'*16 for _ in range(h))
|
||||
input_depth = b''.join(struct.pack('<f', z+1e-6 if mode == 'far' else .5 if source else 1.)*w + b'\xa5'*12 for _ in range(h))
|
||||
arrays = [current, color, near, motion, previous,
|
||||
bytes([(4 if conservative else 16) if mode == 'expired' else (3 if conservative else 15)])*pixels,
|
||||
struct.pack('<f', 0. if mode == 'disoccluded' else .5)*pixels,
|
||||
b'\xa5'*(pitch*h), b'\xa5'*(pixels+64), b'\xa5'*(pixels*4+64),
|
||||
input_depth, b'\xa5'*(pitch*h)]
|
||||
if sparse: arrays.append(b'\xa5'*(pixels+64))
|
||||
addresses = [(i+1)*0x100000000 for i in range(len(arrays))]
|
||||
args = struct.pack('<17Q5I2fI', *addresses[:12], pitch, dp, mp, pitch, dp,
|
||||
w, h, w, h, int(history), jitter, -jitter, int(conservative) | (2 if sky_only else 0))
|
||||
if sparse: args += struct.pack('<Q', addresses[12])
|
||||
(directory/'args.bin').write_bytes(args)
|
||||
manifest = ['kernel '+('d4r_wire_history_sparse' if sparse else 'd4r_wire_history'), 'launch 5 5 1 256 1 1 0', f'args {len(args)}']
|
||||
for i, data in enumerate(arrays):
|
||||
(directory/f'alloc-{i}.bin').write_bytes(data)
|
||||
manifest += [f'alloc {i} {hex(addresses[i])} {len(data)}', f'pointer {168 if i == 12 else i*8} {i} 0']
|
||||
(directory/'manifest.txt').write_text('\n'.join(manifest)+'\n')
|
||||
if sparse:
|
||||
subprocess.run([runner, str(kernel), 'd4r_wire_history_sparse', str(directory), str(output), '1'], check=True, capture_output=True)
|
||||
for i in range(len(arrays)):
|
||||
(directory/f'alloc-{i}.bin').write_bytes((output/f'alloc-{i}.bin').read_bytes())
|
||||
manifest[0]='kernel d4r_wire_history_apply'
|
||||
(directory/'manifest.txt').write_text('\n'.join(manifest)+'\n')
|
||||
subprocess.run([runner, str(kernel), 'd4r_wire_history_apply', str(directory), str(output), '1'], check=True, capture_output=True)
|
||||
else:
|
||||
subprocess.run([runner, str(kernel), 'd4r_wire_history', str(directory), str(output), '1'], check=True, capture_output=True)
|
||||
for i in ((1, 2, 3, 4, 5, 6, 10) if sparse else (0, 1, 2, 3, 4, 5, 6, 10)):
|
||||
self.assertEqual((output/f'alloc-{i}.bin').read_bytes(), arrays[i])
|
||||
expected = bytearray(current)
|
||||
cache = bytearray(current)
|
||||
age = bytearray(b'\xff'*pixels + b'\xa5'*64)
|
||||
depth = bytearray(struct.pack('<f', 1.)*pixels + b'\xa5'*64)
|
||||
for y in range(16, h-16):
|
||||
if mode == 'backdrop' and sky_only: continue
|
||||
x = 32-jitter
|
||||
offset, index = y*pitch+x*8, y*w+x
|
||||
depth[index*4:index*4+4] = struct.pack('<f', z)
|
||||
if mode in ('source', 'history', 'neighbor', 'backdrop') or (mode == 'far' and not conservative):
|
||||
value = .01 if mode == 'neighbor' and not conservative else .125
|
||||
expected[offset:offset+8] = struct.pack('<4e', value, value, value, .5)
|
||||
cache[offset:offset+8] = struct.pack('<4e', value, value, value, .75)
|
||||
age[index] = 0 if source or mode == 'far' else 4 if conservative else 16
|
||||
self.assertEqual((output/f'alloc-{0 if sparse else 7}.bin').read_bytes(), expected)
|
||||
self.assertEqual((output/'alloc-8.bin').read_bytes(), age)
|
||||
self.assertEqual((output/'alloc-9.bin').read_bytes(), depth)
|
||||
self.assertEqual((output/'alloc-11.bin').read_bytes(), cache)
|
||||
Reference in new issue
Block a user