mirror of
https://github.com/tangerineturtle1/NX2navi.git
synced 2026-11-15 17:09:57 -04:00
df2427d00a
This pull request fixed lighting that added up frame after frame under shadow storage on Linux: a GPU write noted while a part was recorded kept that part's CPU inputs (the game's resets of its lighting accumulators) out of the shadow. Its fix marked noted ranges as submitted in vk_memory.c; the base copies that came with the black-world fix already move every CPU-changed byte in before the part runs, so that part is not taken and no SHADOW_PENDING_INPUTS setting is added. Taken from it: - tex_probe.c (--probe-vkmem): a range noted as a future GPU write still gets the CPU's bytes before the GPU reads it. The pull request's other check, that a CPU write made after a GPU result loses to it until the write-back, is left out: with base copies the later CPU write wins, as it does on the console. - vk_dump.c: NX2NAVI_DUMP_SHADERS=<hash>,... copies images only for passes that use those shaders, and NX2NAVI_DUMP_RAW=1 also saves each copied image's texel bytes. - tools/fix_linux_ambient_power.py, as it is: patches one translated shader of a player's own cache so a zero exponent no longer gives NaN. It holds two hashes and no shader data. Pull request #264 by intellq. The Windows files compile; the probe was not run, and nothing was run on Linux here. Co-Authored-By: Claude Fable 5.2 <noreply@anthropic.com>
90 lines
3.8 KiB
Python
90 lines
3.8 KiB
Python
#!/usr/bin/env python3
|
|
"""Apply the verified Ocarina 1.1.1 ambient-power workaround to a Linux shader cache file.
|
|
|
|
The affected shader evaluates exp2(log2(height) * exponent). A zero height
|
|
and zero exponent produce a NaN, which then contaminates ambient lighting
|
|
and volumetric history. Select a zero product when the exponent is zero.
|
|
This is a narrowly scoped compatibility workaround, not a claim of exact
|
|
NVIDIA instruction equivalence. It does not address exposure or memory use.
|
|
No game shader data is embedded in this tool.
|
|
"""
|
|
import argparse
|
|
from pathlib import Path
|
|
import re
|
|
import shutil
|
|
import subprocess
|
|
import tempfile
|
|
|
|
ORIGINAL_HASH = 0xA806B657C39684BD
|
|
PATCHED_HASH = 0x0A0281063D5EF5BD
|
|
|
|
|
|
def shader_hash(data):
|
|
value = 0xCBF29CE484222325
|
|
for byte in data:
|
|
value = ((value ^ byte) * 0x100000001B3) & 0xFFFFFFFFFFFFFFFF
|
|
return value
|
|
|
|
|
|
def patch_text(source):
|
|
pattern = re.compile(
|
|
r'(?P<ptr>%\w+) = OpAccessChain %\w+ %cp_c3 %int_0 %int_2 %int_1\s+'
|
|
r'(?P<exponent>%\w+) = OpLoad %float (?P=ptr)\s+'
|
|
r'(?P<product>%\w+) = OpFMul %float (?P<logarithm>%\w+) (?P=exponent)\s+'
|
|
r'(?P<result>%\w+) = OpExtInst %float %\w+ Exp2 (?P=product)'
|
|
)
|
|
matches = list(pattern.finditer(source))
|
|
if len(matches) != 1:
|
|
raise ValueError('Expected exactly one supported ambient-power expression')
|
|
m = matches[0]
|
|
old = f"{m['product']} = OpFMul %float {m['logarithm']} {m['exponent']}"
|
|
replacement = (
|
|
f"%linux_pow_product = OpFMul %float {m['logarithm']} {m['exponent']}\n"
|
|
f"%linux_pow_zero_exponent = OpFOrdEqual %bool {m['exponent']} %float_0\n"
|
|
f"{m['product']} = OpSelect %float %linux_pow_zero_exponent %float_0 %linux_pow_product"
|
|
)
|
|
if source.count(old) != 1:
|
|
raise ValueError('The supported multiply instruction is not unique')
|
|
return source.replace(old, replacement)
|
|
|
|
|
|
def main():
|
|
parser = argparse.ArgumentParser(description=__doc__)
|
|
parser.add_argument('shader', type=Path, help='Supported .spv shader cache file')
|
|
destination = parser.add_mutually_exclusive_group(required=True)
|
|
destination.add_argument('--output', type=Path)
|
|
destination.add_argument('--in-place', action='store_true')
|
|
args = parser.parse_args()
|
|
original = args.shader.read_bytes()
|
|
current = shader_hash(original)
|
|
if current == PATCHED_HASH:
|
|
if args.output:
|
|
args.output.write_bytes(original)
|
|
print(f'Already patched: {current:016x}')
|
|
return
|
|
if current != ORIGINAL_HASH:
|
|
parser.error(f'Unsupported shader hash {current:016x}; file left unchanged')
|
|
for command in ('spirv-dis', 'spirv-as', 'spirv-val'):
|
|
if not shutil.which(command):
|
|
parser.error(f'Required command not found: {command}')
|
|
target = args.shader if args.in_place else args.output
|
|
with tempfile.TemporaryDirectory(prefix='.ambient-power-', dir=target.parent) as work:
|
|
work = Path(work)
|
|
assembly = work / 'shader.spvasm'
|
|
binary = work / 'shader.spv'
|
|
subprocess.run(['spirv-dis', str(args.shader), '-o', str(assembly)], check=True)
|
|
assembly.write_text(patch_text(assembly.read_text()))
|
|
subprocess.run(['spirv-as', '--target-env', 'vulkan1.1', str(assembly), '-o', str(binary)], check=True)
|
|
subprocess.run(['spirv-val', '--target-env', 'vulkan1.1', str(binary)], check=True)
|
|
patched = binary.read_bytes()
|
|
result = shader_hash(patched)
|
|
if result != PATCHED_HASH:
|
|
raise ValueError(f'Unexpected assembled shader hash {result:016x}; file left unchanged')
|
|
binary.chmod(args.shader.stat().st_mode & 0o777)
|
|
binary.replace(target)
|
|
print(f'Patched {target}: {current:016x} -> {result:016x}; Vulkan 1.1 validation passed')
|
|
|
|
|
|
if __name__ == '__main__':
|
|
main()
|