mirror/RetroArch
1
0
Fork 0
mirror of https://github.com/libretro/RetroArch synced 2025-01-29 00:32:49 +00:00
Code Issues Packages Projects Releases Wiki Activity

Squashed 'deps/SPIRV-Cross/' changes from c9516fa917..e59cc24495

Browse Source 
e59cc24495 Merge pull request #607 from billhollings/master
ac57a30ad6 Regenerate MSL shaders to fix conflicts from merge.
ab2ea93e35 Merge branch 'master' of https://github.com/KhronosGroup/SPIRV-Cross
9b4defe202 CompilerMSL support matrices & arrays in stage-in & stage-out.
285f214071 Merge pull request #605 from KhronosGroup/unpack-regular-operations-msl
58fab58e5e Do not unpack transposed matrices.
192a882df3 Also unpack regular unary/binary operations on MSL.
dc565136fb Merge pull request #604 from KhronosGroup/fix-603
b86bd0a265 Unpack expressions when used in functions on MSL.
c5b6ba239c Merge pull request #599 from KhronosGroup/fix-295
f1e8555801 Emit matrix layouts in structs directly.
b1196f2ace Merge pull request #598 from KhronosGroup/fix-594
6bcc890e63 Sanitize underscores in general, not just for members.
6fea07f2fd Merge pull request #597 from KhronosGroup/fix-595-596-592
3a9b045dc3 Various maintenance fixes.
0a83bacf3e Merge pull request #593 from KhronosGroup/old-clang-compile
2a1ab4108b Fix compile on older clang.
6ef1c49ec0 Merge pull request #591 from KhronosGroup/fix-578
04b149feb0 Fix image load/store on cube arrays in MSL.
8bac5c09f3 Merge pull request #590 from KhronosGroup/fix-581
f65120c147 Deal with packed expressions in more scenarios.
db1ed375b0 Merge pull request #589 from KhronosGroup/fix-580
280fb93204 Add test for reading SSBO from fragment shader on MSL.
a1b3964cbd Run format_all.sh.
6b3da831be Declare read-only SSBOs as const device in MSL.
08336e7bbb Merge pull request #588 from KhronosGroup/pr-583
46bf17c5d3 Add SREM tests for HLSL/MSL.
7cba89b4a5 Add reference files for SREM.
6b144cc609 handle OpSRem
e792cd6160 no expression type for OpSRem
f1eacba244 Merge pull request #587 from zeux/master
91fd41816f Fix textureGrad compilation for legacy targets
8cc1fdbb30 Merge pull request #586 from KhronosGroup/fix-584
ba15daee33 Add support for inheriting bindings for combined image samplers.
f929c361c5 Merge pull request #577 from KhronosGroup/fix-575
bcaae84c76 Deal with scoping for Private variables.
26b887ec99 Fix atomic_compare_exchange_weak_explicit.
62c6d2d498 Merge pull request #576 from KhronosGroup/fix-574
fb7181bff1 Run format_all.sh.
c643addacd Only reflect spec constant if it actually has a constant ID.
991b655c72 Declare OpSpecConstantOp up-front on relevant targets.
0617b98613 Run format_all.sh.
3951b9456f Fix SpecConstantComposite if input is SpecConstantOp.
c74dc4578a Merge pull request #572 from KhronosGroup/vulkan-glsl-sampler-mediump
01080365fa Use mediump on images in --vulkan-semantics as well.
9d370aca58 Merge pull request #571 from KhronosGroup/fix-570
7eba247864 Handle inout properly with split access chains.
97e38bcd38 Merge pull request #569 from KhronosGroup/fix-566
b71f5dfc0c Fix split access chains for builtin arrays.
e4694a8403 Merge pull request #568 from KhronosGroup/fix-567
903b798da7 Fix GCC 8.1 build.
d3b966322d Merge pull request #565 from pmuetschard/master
aced6058b4 Don't limit GLSL identifiers with HLSL keywords.
2792f8f3f2 Merge pull request #564 from KhronosGroup/fix-563
85a8f066f4 Do not use RMW rewrite for matrices.
04f0a08cfc Merge pull request #562 from KhronosGroup/fix-561
d2df067dd4 Force recompile if we add row-major transpose functions in MSL.
7b95168c3d Do not clear spv_function_implementations on MSL.
9279750a1d Merge pull request #560 from KhronosGroup/fix-557
17be3c652f Deal with fake overloads when using combined image samplers.
e80d6e0142 Merge pull request #559 from KhronosGroup/fix-558
3187d89c31 Reduce test case ... MSL can't deal with certain unusual cases yet.
f3e810b8b3 Fix OpCompositeConstruct with arrays in MSL.
9d43e9c02a Merge pull request #556 from billhollings/master
ba1026b899 Merge branch 'master' of https://github.com/KhronosGroup/SPIRV-Cross
57213cb7ca Compiler MSL default gather offset when component specified.
1ebccb4466 Mention Wiki in README.
8f07df016b Merge pull request #555 from KhronosGroup/fix-553
7e23e69f2a Run format_all.sh.
47081f810a Fix GatherDref on GLSL.
6b478c928d Merge pull request #554 from KhronosGroup/query-fixes
40bbf6be7a Build combined dummy samplers for Query functions without sampler as well.
d93807a625 Deal with OpImageFetch without explicit LOD.
454691f434 Merge pull request #547 from KhronosGroup/op-image-usage-tracking
aaf397cd1f Fix usage tracking issue for OpImage.
bbf6b60350 Merge pull request #546 from KhronosGroup/fix-545
893ed379d5 Remove some stale test files.
cd442b9c2f Clean and fix some issues with test_shaders.py.
18958ecccb Add test for combined-sampler-reuse on GLSL.
0280800a8f Fix case where SampledImage would get flushed to temporary.
a090907607 Merge pull request #543 from KhronosGroup/fix-542
8b75e46433 Fix some formatting issues.
f56433b802 Add new tests for combined arrays of sampler + image.
a39eb4826b Combined array of images is starting to work ...
1478847122 Change AppVeyor badge link to the correct one.
d559a8df39 Add AppVeyor status badge.
dd39d36726 Merge pull request #540 from KhronosGroup/appveyor
cc82d14ee8 Add appveyor script.
d0c829b580 Add no-opt to CTest, make test_shaders.py faster when FXC is missing.
2804d9acd7 Merge pull request #539 from KhronosGroup/fix-533
e351e5c565 Use convert_to_string for lod clamp.
e30a94225f Complete MSL constexpr samplers.
64f9461d72 Check for array of samplers.
df58debf7a Add support for constexpr samplers in MSL.
3a8335eee0 Merge pull request #537 from KhronosGroup/fix-535
b9cd3dcd7f Run format_all.sh.
e930f79e2e Be a bit smarter about uint on legacy targets.
2684054bbd Merge pull request #538 from KhronosGroup/fix-io-block-location
b06c1af9b3 Distinguish between enhanced_layouts and SSO requirements.
28c9be8a23 Unsigned integers are disallowed on legacy targets.
7796a9f3ec Merge pull request #531 from KhronosGroup/subgroup
c266429be9 Partially implement subgroup ops for HLSL SM 6.0.
146ea76f52 Add test shader for subgroup.
e1ccfd5dbb Implement all of subgroup.
f6c0e53f58 Start adding Vulkan 1.1 subgroup support to GLSL.
489e04e09e Merge pull request #530 from KhronosGroup/fix-529
9c2761f69a Run format_all.sh.
8175e2e200 Fix depth compare textures when used in functions without argument.
edac731360 Merge pull request #528 from msiglreith/fix_root_constants_name
6bdd775936 Merge pull request #526 from taisei-project/strip-uniform-location
f260c452f7 Fix naming of root constants
4a43024dba Strip uniform locations for ESSL < 3.00 and GLSL < 430
9b8c2c4654 Merge pull request #524 from KhronosGroup/fix-512-523
694b314f87 Support empty structs.
31a3fdf4ee Decouple public require_extension and the internal use of the function.
d0133c5dd9 Merge pull request #509 from KhronosGroup/hlsl-semantic-counter-buffer
d8cdab2f57 Add test shader for semantic decoration.
05c66f9cac Add test shader which uses counter buffer extension.
215d3ca0a4 Add support for new HLSL semantic/counter buffer decorations.
7f84537350 Merge pull request #519 from KhronosGroup/fix-507
ac81a0ce68 Use declared binding in SPIR-V as a fallback for explicit MSL binds.
e8ca39b7b5 Add test for sampler image arrays.
382101bd05 Run format_all.sh.
1d082e9b79 Fix passing arrays of combined image samplers in HLSL.
5827dd54ea Support array of images and samplers in MSL.
81eb72a9a0 Ignore LOD when sampling 1D textures in MSL.
69ffd918a6 Merge pull request #522 from KhronosGroup/fix-515
938040be0b Only disable binding layout for UBOs.
fe697a80f5 Emit classic uniform for UBO in GLSL 1.30.
65be63fd04 Merge pull request #521 from KhronosGroup/fix-516
a6e211e00b Support dual-source blending on GLSL and MSL.
a1e30c8c2b Merge pull request #520 from KhronosGroup/fix-517
3229e6efb6 Add more illegal name replacement in MSL.
761b06bc73 Merge pull request #514 from KhronosGroup/fix-510
a901b4c814 Remove redundant temporary sort.
c1947aa447 Update glslang/SPIRV-Tools on Travis.
35f64d03bb Fix name aliasing for temporary variables.
5161d5ed3b Merge pull request #505 from KhronosGroup/msl-array-mrt
719cf9d42f Run format_all.sh.
8e90382675 Properly flatten MRT outputs in MSL.
6e6ca0b237 Attempt MRT-as-array in MSL.
fbfe13657a Merge pull request #502 from KhronosGroup/update-glslang-spirv-tools
4f348a9807 Update README and update_test_shaders.sh.
b46910e9f5 Access chain into packed vectors as arrays.
a6afda650f Add newly moved reference files.
cdbd376c42 Try building debug on Travis.
e7bf8d2f48 Refactor out noopt shaders to their own folders.
1a4a36b6cd Experiment a bit more with Travis.
5a746ba941 Split checkout and build on Travis CI.
4979d10b54 Implement packHalf2x16/unpackHalf2x16 on MSL.
ae2680c898 Make sure ballotARB inherits expression dependencies.
938c7debed Handle control-dependent temporaries.
012377f811 Refactor block IO emission to use bitset.
955eac522b Make use of the handy scripts for Travis as well.
9fbd8b789e Update tests for latest SPIRV-Tools and glslang.
e8e58844d4 Rewrite everything to use Bitset rather than uint64_t.
29315f3b3f Merge pull request #498 from JustSid/master
e66fd6c2a0 Added generation of the textureXYZGrad() legacy GLSL instruction
0f9cb0da0d Merge pull request #497 from KhronosGroup/fix-489
9a52713d77 Run format_all.sh.
a803e5ae38 Deprecate set_options()/get_options() interface, replace it.
236b2fa47c Merge pull request #496 from KhronosGroup/fix-494
e3b8e9455c Add test shader where a phi variable invalidates a temporary.
1de74fdabd Explicitly emit break/continues in selection branches.
e0efa737ca Expand the implementation of inherit_expression_dependencies.
01fbb3a342 Merge pull request #495 from KhronosGroup/fix-493
28cccc3dbb Emit complex continue blocks "properly".
54549a624f Add some test shaders for special merge methods.
8d557d4103 Handle cases where merge selects as also loop merge or continue blocks.
8c8a93f745 Merge pull request #492 from KhronosGroup/fix-488
922420e346 Disallow arrays and structs from becoming loop variables.
2e5d06d1ce Merge pull request #485 from KhronosGroup/fix-484
bdabd0c73a Disable double test in MSL FP16 tests.
723bec5d33 frexp/ldexp fails in debug builds, workaround temporarily.
0780820a7b Skip interpolateAt tests with FP16.
ac0e93f392 Run format_all.sh.
18ad1be3c3 Add FP16 test for MSL as well.
47d94ff8d9 Add FP16 to HLSL.
770ed25f29 Merge pull request #487 from KhronosGroup/fix-486
15a941cd3d Add builtin GLSL functions to blacklist.
d9da2db442 Some compat fixes for MSL and Half.
547278da12 Test denormal fp16 constants.
24dfe496c2 Run format_all.sh.
05348a66ca Add test shader for FP16 support.
bc0f698df3 Fix packing alignments for float16.
91f85d3412 Begin adding float16_t support to GLSL.
b0a2de5b63 Merge pull request #483 from KhronosGroup/fix-466
38d9d8af68 Add some more MSL test shaders.
5fe79eb59c Update tests.
294259e2f1 Fix type aliasing on MSL.

git-subtree-dir: deps/SPIRV-Cross
git-subtree-split: e59cc244958af6059f7bd1d16d833e17409dec37
This commit is contained in:
twinaphex 2018-06-20 09:32:45 +02:00
parent f48c1800c0
commit 0328298474
609 changed files with 15185 additions and 4607 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

23
.travis.yml
View File

@ -2,32 +2,17 @@ language: cpp
os:
- linux
- osx
osx_image: xcode8.3
osx_image: xcode9.3beta
# Use Ubuntu 14.04 LTS (Trusty) as the Linux testing environment.
sudo: required
dist: trusty
# We check out glslang and SPIRV-Tools at specific revisions to avoid test output mismatches
env:
- GLSLANG_REV=9c6f8cc29ba303b43ccf36deea6bb38a304f9b92 SPIRV_TOOLS_REV=e28edd458b729da7bbfd51e375feb33103709e6f
before_script:
- if [[ "$TRAVIS_OS_NAME" == "osx" ]]; then brew upgrade python3; fi
- git clone https://github.com/KhronosGroup/glslang.git glslang
- git clone https://github.com/KhronosGroup/SPIRV-Tools SPIRV-Tools
- git clone https://github.com/KhronosGroup/SPIRV-Headers.git SPIRV-Tools/external/spirv-headers
- ./checkout_glslang_spirv_tools.sh
script:
- git -C glslang checkout $GLSLANG_REV
- git -C SPIRV-Tools checkout $SPIRV_TOOLS_REV
- cd glslang && cmake . && make -j2 && cd ..
- cd SPIRV-Tools && cmake . && make -j2 && cd ..
- ./build_glslang_spirv_tools.sh Debug 2
- make -j2
- PATH=./glslang/StandAlone:./SPIRV-Tools/tools:$PATH
- ./test_shaders.py shaders
- ./test_shaders.py --msl shaders-msl
- ./test_shaders.py --hlsl shaders-hlsl
- ./test_shaders.py shaders --opt
- ./test_shaders.py --msl shaders-msl --opt
- ./test_shaders.py --hlsl shaders-hlsl --opt
- ./test_shaders.sh

9
CMakeLists.txt
View File

@ -128,12 +128,21 @@ if (${PYTHONINTERP_FOUND})
add_test(NAME spirv-cross-test
COMMAND ${PYTHON_EXECUTABLE} ${CMAKE_CURRENT_SOURCE_DIR}/test_shaders.py
${CMAKE_CURRENT_SOURCE_DIR}/shaders)
add_test(NAME spirv-cross-test-no-opt
COMMAND ${PYTHON_EXECUTABLE} ${CMAKE_CURRENT_SOURCE_DIR}/test_shaders.py
${CMAKE_CURRENT_SOURCE_DIR}/shaders-no-opt)
add_test(NAME spirv-cross-test-metal
COMMAND ${PYTHON_EXECUTABLE} ${CMAKE_CURRENT_SOURCE_DIR}/test_shaders.py --metal
${CMAKE_CURRENT_SOURCE_DIR}/shaders-msl)
add_test(NAME spirv-cross-test-metal-no-opt
COMMAND ${PYTHON_EXECUTABLE} ${CMAKE_CURRENT_SOURCE_DIR}/test_shaders.py --metal
${CMAKE_CURRENT_SOURCE_DIR}/shaders-msl-no-opt)
add_test(NAME spirv-cross-test-hlsl
COMMAND ${PYTHON_EXECUTABLE} ${CMAKE_CURRENT_SOURCE_DIR}/test_shaders.py --hlsl
${CMAKE_CURRENT_SOURCE_DIR}/shaders-hlsl)
add_test(NAME spirv-cross-test-hlsl-no-opt
COMMAND ${PYTHON_EXECUTABLE} ${CMAKE_CURRENT_SOURCE_DIR}/test_shaders.py --hlsl
${CMAKE_CURRENT_SOURCE_DIR}/shaders-hlsl-no-opt)
add_test(NAME spirv-cross-test-opt
COMMAND ${PYTHON_EXECUTABLE} ${CMAKE_CURRENT_SOURCE_DIR}/test_shaders.py --opt
${CMAKE_CURRENT_SOURCE_DIR}/shaders)

6
README.md
View File

@ -3,6 +3,7 @@
SPIRV-Cross is a tool designed for parsing and converting SPIR-V to other shader languages.
[![Build Status](https://travis-ci.org/KhronosGroup/SPIRV-Cross.svg?branch=master)](https://travis-ci.org/KhronosGroup/SPIRV-Cross)
[![Build Status](https://ci.appveyor.com/api/projects/status/github/KhronosGroup/SPIRV-Cross?svg=true&branch=master)](https://ci.appveyor.com/project/HansKristian-ARM/SPIRV-Cross)
## Features
@ -24,7 +25,7 @@ However, most missing features are expected to be "trivial" improvements at this
SPIRV-Cross has been tested on Linux, OSX and Windows.
The make and CMake build flavors offer the option to treat exceptions as assertions. To disable exceptions for make just append SPIRV_CROSS_EXCEPTIONS_TO_ASSERTIONS=1 to the command line. For CMake append -DSPIRV_CROSS_EXCEPTIONS_TO_ASSERTIONS=ON. By default exceptions are enabled.
The make and CMake build flavors offer the option to treat exceptions as assertions. To disable exceptions for make just append `SPIRV_CROSS_EXCEPTIONS_TO_ASSERTIONS=1` to the command line. For CMake append `-DSPIRV_CROSS_EXCEPTIONS_TO_ASSERTIONS=ON`. By default exceptions are enabled.
### Linux and macOS
@ -38,6 +39,8 @@ MinGW-w64 based compilation works with `make`, and an MSVC 2013 solution is also
### Using the C++ API
For more in-depth documentation than what's provided in this README, please have a look at the [Wiki](https://github.com/KhronosGroup/SPIRV-Cross/wiki).
To perform reflection and convert to other shader languages you can use the SPIRV-Cross API.
For example:
@ -279,6 +282,7 @@ although there are a couple of convenience script for doing this:
```
./checkout_glslang_spirv_tools.sh # Checks out glslang and SPIRV-Tools at a fixed revision which matches the reference output.
./build_glslang_spirv_tools.sh # Builds glslang and SPIRV-Tools.
./test_shaders.sh # Runs over all changes and makes sure that there are no deltas compared to reference files.
```

31
appveyor.yml Normal file
View File

@ -0,0 +1,31 @@
environment:
matrix:
- GENERATOR: "Visual Studio 12 2013 Win64"
CONFIG: Debug
- GENERATOR: "Visual Studio 12 2013 Win64"
CONFIG: Release
- GENERATOR: "Visual Studio 14 2015 Win64"
CONFIG: Debug
- GENERATOR: "Visual Studio 14 2015 Win64"
CONFIG: Release
- GENERATOR: "Visual Studio 12 2013"
CONFIG: Debug
- GENERATOR: "Visual Studio 12 2013"
CONFIG: Release
- GENERATOR: "Visual Studio 14 2015"
CONFIG: Debug
- GENERATOR: "Visual Studio 14 2015"
CONFIG: Release
build_script:
- git submodule update --init
- cmake "-G%GENERATOR%" -H. -B_builds
- cmake --build _builds --config "%CONFIG%"

26
build_glslang_spirv_tools.sh Executable file
View File

@ -0,0 +1,26 @@
#!/bin/bash
PROFILE=Release
if [ ! -z $1 ]; then
PROFILE=$1
fi
NPROC=$(nproc)
if [ ! -z $2 ]; then
NPROC=$2
fi
echo "Building glslang."
mkdir -p external/glslang-build
cd external/glslang-build
cmake ../glslang -DCMAKE_BUILD_TYPE=$PROFILE -G"Unix Makefiles"
make -j$NPROC
cd ../..
echo "Building SPIRV-Tools."
mkdir -p external/spirv-tools-build
cd external/spirv-tools-build
cmake ../spirv-tools -DCMAKE_BUILD_TYPE=$PROFILE -G"Unix Makefiles" -DSPIRV_WERROR=OFF
make -j$NPROC
cd ../..

35
checkout_glslang_spirv_tools.sh
View File

@ -1,7 +1,7 @@
#!/bin/bash
GLSLANG_REV=9c6f8cc29ba303b43ccf36deea6bb38a304f9b92
SPIRV_TOOLS_REV=e28edd458b729da7bbfd51e375feb33103709e6f
GLSLANG_REV=461ea09943e0e88ea854ab9e3b42d17d728af2ad
SPIRV_TOOLS_REV=53bc1623ecd3cc304d0d6feed8385e70c7ab30d3
if [ -d external/glslang ]; then
echo "Updating glslang to revision $GLSLANG_REV."
@ -18,13 +18,6 @@ else
fi
cd ../..
echo "Building glslang."
mkdir -p external/glslang-build
cd external/glslang-build
cmake ../glslang -DCMAKE_BUILD_TYPE=Release -G"Unix Makefiles"
make -j$(nproc)
cd ../..
if [ -d external/spirv-tools ]; then
echo "Updating SPIRV-Tools to revision $SPIRV_TOOLS_REV."
cd external/spirv-tools
@ -37,21 +30,15 @@ else
git clone git://github.com/KhronosGroup/SPIRV-Tools.git spirv-tools
cd spirv-tools
git checkout $SPIRV_TOOLS_REV
if [ -d external/spirv-headers ]; then
cd external/spirv-headers
git pull origin master
cd ../..
else
git clone git://github.com/KhronosGroup/SPIRV-Headers.git external/spirv-headers
fi
fi
cd ../..
echo "Building SPIRV-Tools."
mkdir -p external/spirv-tools-build
cd external/spirv-tools-build
cmake ../spirv-tools -DCMAKE_BUILD_TYPE=Release -G"Unix Makefiles"
make -j$(nproc)
if [ -d external/spirv-headers ]; then
cd external/spirv-headers
git pull origin master
cd ../..
else
git clone git://github.com/KhronosGroup/SPIRV-Headers.git external/spirv-headers
fi
cd ../..

3
format_all.sh
View File

@ -1,6 +1,7 @@
#!/bin/bash
for file in spirv_*.{cpp,hpp} include/spirv_cross/*.{hpp,h} samples/cpp/*.cpp main.cpp
#for file in spirv_*.{cpp,hpp} include/spirv_cross/*.{hpp,h} samples/cpp/*.cpp main.cpp
for file in spirv_*.{cpp,hpp} main.cpp
do
echo "Formatting file: $file ..."
clang-format -style=file -i $file

118
main.cpp
View File

@ -122,7 +122,7 @@ struct CLIParser
THROW("Tried to parse uint, but nothing left in arguments");
}
uint32_t val = stoul(*argv);
uint64_t val = stoul(*argv);
if (val > numeric_limits<uint32_t>::max())
{
THROW("next_uint() out of range");
@ -131,7 +131,7 @@ struct CLIParser
argc--;
argv++;
return val;
return uint32_t(val);
}
double next_double()
@ -212,7 +212,6 @@ static void print_resources(const Compiler &compiler, const char *tag, const vec
for (auto &res : resources)
{
auto &type = compiler.get_type(res.type_id);
auto mask = compiler.get_decoration_mask(res.id);
if (print_ssbo && compiler.buffer_is_hlsl_counter_buffer(res.id))
continue;
@ -221,8 +220,8 @@ static void print_resources(const Compiler &compiler, const char *tag, const vec
// for SSBOs and UBOs since those are the only meaningful names to use externally.
// Push constant blocks are still accessed by name and not block name, even though they are technically Blocks.
bool is_push_constant = compiler.get_storage_class(res.id) == StorageClassPushConstant;
bool is_block = (compiler.get_decoration_mask(type.self) &
((1ull << DecorationBlock) | (1ull << DecorationBufferBlock))) != 0;
bool is_block = compiler.get_decoration_bitset(type.self).get(DecorationBlock) ||
compiler.get_decoration_bitset(type.self).get(DecorationBufferBlock);
bool is_sized_block = is_block && (compiler.get_storage_class(res.id) == StorageClassUniform ||
compiler.get_storage_class(res.id) == StorageClassUniformConstant);
uint32_t fallback_id = !is_push_constant && is_block ? res.base_type_id : res.id;
@ -231,6 +230,12 @@ static void print_resources(const Compiler &compiler, const char *tag, const vec
if (is_sized_block)
block_size = uint32_t(compiler.get_declared_struct_size(compiler.get_type(res.base_type_id)));
Bitset mask;
if (print_ssbo)
mask = compiler.get_buffer_block_flags(res.id);
else
mask = compiler.get_decoration_bitset(res.id);
string array;
for (auto arr : type.array)
array = join("[", arr ? convert_to_string(arr) : "", "]") + array;
@ -238,17 +243,17 @@ static void print_resources(const Compiler &compiler, const char *tag, const vec
fprintf(stderr, " ID %03u : %s%s", res.id,
!res.name.empty() ? res.name.c_str() : compiler.get_fallback_name(fallback_id).c_str(), array.c_str());
if (mask & (1ull << DecorationLocation))
if (mask.get(DecorationLocation))
fprintf(stderr, " (Location : %u)", compiler.get_decoration(res.id, DecorationLocation));
if (mask & (1ull << DecorationDescriptorSet))
if (mask.get(DecorationDescriptorSet))
fprintf(stderr, " (Set : %u)", compiler.get_decoration(res.id, DecorationDescriptorSet));
if (mask & (1ull << DecorationBinding))
if (mask.get(DecorationBinding))
fprintf(stderr, " (Binding : %u)", compiler.get_decoration(res.id, DecorationBinding));
if (mask & (1ull << DecorationInputAttachmentIndex))
if (mask.get(DecorationInputAttachmentIndex))
fprintf(stderr, " (Attachment : %u)", compiler.get_decoration(res.id, DecorationInputAttachmentIndex));
if (mask & (1ull << DecorationNonReadable))
if (mask.get(DecorationNonReadable))
fprintf(stderr, " writeonly");
if (mask & (1ull << DecorationNonWritable))
if (mask.get(DecorationNonWritable))
fprintf(stderr, " readonly");
if (is_sized_block)
fprintf(stderr, " (BlockSize : %u bytes)", block_size);
@ -284,7 +289,7 @@ static const char *execution_model_to_str(spv::ExecutionModel model)
static void print_resources(const Compiler &compiler, const ShaderResources &res)
{
uint64_t modes = compiler.get_execution_mode_mask();
auto &modes = compiler.get_execution_mode_bitset();
fprintf(stderr, "Entry points:\n");
auto entry_points = compiler.get_entry_points_and_stages();
@ -293,11 +298,7 @@ static void print_resources(const Compiler &compiler, const ShaderResources &res
fprintf(stderr, "\n");
fprintf(stderr, "Execution modes:\n");
for (unsigned i = 0; i < 64; i++)
{
if (!(modes & (1ull << i)))
continue;
modes.for_each_bit([&](uint32_t i) {
auto mode = static_cast<ExecutionMode>(i);
uint32_t arg0 = compiler.get_execution_mode_argument(mode, 0);
uint32_t arg1 = compiler.get_execution_mode_argument(mode, 1);
@ -353,7 +354,7 @@ static void print_resources(const Compiler &compiler, const ShaderResources &res
default:
break;
}
}
});
fprintf(stderr, "\n");
print_resources(compiler, "subpass inputs", res.subpass_inputs);
@ -487,25 +488,47 @@ struct CLIArguments
bool flatten_multidimensional_arrays = false;
bool use_420pack_extension = true;
bool remove_unused = false;
bool combined_samplers_inherit_bindings = false;
};
static void print_help()
{
fprintf(stderr, "Usage: spirv-cross [--output <output path>] [SPIR-V file] [--es] [--no-es] "
"[--version <GLSL version>] [--dump-resources] [--help] [--force-temporary] "
"[--vulkan-semantics] [--flatten-ubo] [--fixup-clipspace] [--flip-vert-y] [--iterations iter] "
"[--cpp] [--cpp-interface-name <name>] "
"[--msl] [--msl-version <MMmmpp>]"
"[--hlsl] [--shader-model] [--hlsl-enable-compat] "
"[--separate-shader-objects]"
"[--pls-in format input-name] [--pls-out format output-name] [--remap source_name target_name "
"components] [--extension ext] [--entry name] [--stage <stage (vert, frag, geom, tesc, tese, "
"comp)>] [--remove-unused-variables] "
"[--flatten-multidimensional-arrays] [--no-420pack-extension] "
"[--remap-variable-type <variable_name> <new_variable_type>] "
"[--rename-interface-variable <in|out> <location> <new_variable_name>] "
"[--set-hlsl-vertex-input-semantic <location> <semantic>] "
"[--rename-entry-point <old> <new> <stage>] "
fprintf(stderr, "Usage: spirv-cross\n"
"\t[--output <output path>]\n"
"\t[SPIR-V file]\n"
"\t[--es]\n"
"\t[--no-es]\n"
"\t[--version <GLSL version>]\n"
"\t[--dump-resources]\n"
"\t[--help]\n"
"\t[--force-temporary]\n"
"\t[--vulkan-semantics]\n"
"\t[--flatten-ubo]\n"
"\t[--fixup-clipspace]\n"
"\t[--flip-vert-y]\n"
"\t[--iterations iter]\n"
"\t[--cpp]\n"
"\t[--cpp-interface-name <name>]\n"
"\t[--msl]\n"
"\t[--msl-version <MMmmpp>]\n"
"\t[--hlsl]\n"
"\t[--shader-model]\n"
"\t[--hlsl-enable-compat]\n"
"\t[--separate-shader-objects]\n"
"\t[--pls-in format input-name]\n"
"\t[--pls-out format output-name]\n"
"\t[--remap source_name target_name components]\n"
"\t[--extension ext]\n"
"\t[--entry name]\n"
"\t[--stage <stage (vert, frag, geom, tesc, tese comp)>]\n"
"\t[--remove-unused-variables]\n"
"\t[--flatten-multidimensional-arrays]\n"
"\t[--no-420pack-extension]\n"
"\t[--remap-variable-type <variable_name> <new_variable_type>]\n"
"\t[--rename-interface-variable <in|out> <location> <new_variable_name>]\n"
"\t[--set-hlsl-vertex-input-semantic <location> <semantic>]\n"
"\t[--rename-entry-point <old> <new> <stage>]\n"
"\t[--combined-samplers-inherit-bindings]"
"\n");
}
@ -711,6 +734,8 @@ static int main_inner(int argc, char *argv[])
});
cbs.add("--remove-unused-variables", [&args](CLIParser &) { args.remove_unused = true; });
cbs.add("--combined-samplers-inherit-bindings",
[&args](CLIParser &) { args.combined_samplers_inherit_bindings = true; });
cbs.default_handler = [&args](const char *value) { args.input = value; };
cbs.error_handler = [] { print_help(); };
@ -748,10 +773,10 @@ static int main_inner(int argc, char *argv[])
compiler = unique_ptr<CompilerMSL>(new CompilerMSL(read_spirv_file(args.input)));
auto *msl_comp = static_cast<CompilerMSL *>(compiler.get());
auto msl_opts = msl_comp->get_options();
auto msl_opts = msl_comp->get_msl_options();
if (args.set_msl_version)
msl_opts.msl_version = args.msl_version;
msl_comp->set_options(msl_opts);
msl_comp->set_msl_options(msl_opts);
}
else if (args.hlsl)
compiler = unique_ptr<CompilerHLSL>(new CompilerHLSL(read_spirv_file(args.input)));
@ -851,14 +876,14 @@ static int main_inner(int argc, char *argv[])
if (!entry_point.empty())
compiler->set_entry_point(entry_point, model);
if (!args.set_version && !compiler->get_options().version)
if (!args.set_version && !compiler->get_common_options().version)
{
fprintf(stderr, "Didn't specify GLSL version and SPIR-V did not specify language.\n");
print_help();
return EXIT_FAILURE;
}
CompilerGLSL::Options opts = compiler->get_options();
CompilerGLSL::Options opts = compiler->get_common_options();
if (args.set_version)
opts.version = args.version;
if (args.set_es)
@ -870,13 +895,13 @@ static int main_inner(int argc, char *argv[])
opts.vulkan_semantics = args.vulkan_semantics;
opts.vertex.fixup_clipspace = args.fixup;
opts.vertex.flip_vert_y = args.yflip;
compiler->set_options(opts);
compiler->set_common_options(opts);
// Set HLSL specific options.
if (args.hlsl)
{
auto *hlsl = static_cast<CompilerHLSL *>(compiler.get());
auto hlsl_opts = hlsl->get_options();
auto hlsl_opts = hlsl->get_hlsl_options();
if (args.set_shader_model)
{
if (args.shader_model < 30)
@ -894,11 +919,19 @@ static int main_inner(int argc, char *argv[])
hlsl_opts.point_size_compat = true;
hlsl_opts.point_coord_compat = true;
}
hlsl->set_options(hlsl_opts);
hlsl->set_hlsl_options(hlsl_opts);
}
if (build_dummy_sampler)
compiler->build_dummy_sampler_for_combined_images();
{
uint32_t sampler = compiler->build_dummy_sampler_for_combined_images();
if (sampler != 0)
{
// Set some defaults to make validation happy.
compiler->set_decoration(sampler, DecorationDescriptorSet, 0);
compiler->set_decoration(sampler, DecorationBinding, 0);
}
}
ShaderResources res;
if (args.remove_unused)
@ -961,6 +994,9 @@ static int main_inner(int argc, char *argv[])
if (combined_image_samplers)
{
compiler->build_combined_image_samplers();
if (args.combined_samplers_inherit_bindings)
spirv_cross_util::inherit_combined_sampler_bindings(*compiler);
// Give the remapped combined samplers new names.
for (auto &remap : compiler->get_combined_image_samplers())
{

0
reference/shaders-hlsl/asm/comp/storage-buffer-basic.nofxc.asm.comp → reference/opt/shaders-hlsl/asm/comp/storage-buffer-basic.invalid.nofxc.asm.comp
View File

30
reference/opt/shaders-hlsl/asm/frag/combined-sampler-reuse.asm.frag Normal file
View File

@ -0,0 +1,30 @@
Texture2D<float4> uTex : register(t1);
SamplerState uSampler : register(s0);
static float4 FragColor;
static float2 vUV;
struct SPIRV_Cross_Input
{
float2 vUV : TEXCOORD0;
};
struct SPIRV_Cross_Output
{
float4 FragColor : SV_Target0;
};
void frag_main()
{
FragColor = uTex.Sample(uSampler, vUV);
FragColor += uTex.Sample(uSampler, vUV, int2(1, 1));
}
SPIRV_Cross_Output main(SPIRV_Cross_Input stage_input)
{
vUV = stage_input.vUV;
frag_main();
SPIRV_Cross_Output stage_output;
stage_output.FragColor = FragColor;
return stage_output;
}

8
reference/opt/shaders-hlsl/asm/frag/empty-struct.asm.frag Normal file
View File

@ -0,0 +1,8 @@
void frag_main()
{
}
void main()
{
frag_main();
}

2
reference/opt/shaders-hlsl/asm/frag/function-overload-alias.asm.frag
View File

@ -7,7 +7,7 @@ struct SPIRV_Cross_Output
void frag_main()
{
FragColor = (((1.0f.xxxx + 1.0f.xxxx) + (1.0f.xxx.xyzz + 1.0f.xxxx)) + (1.0f.xxxx + 2.0f.xxxx)) + (1.0f.xx.xyxy + 2.0f.xxxx);
FragColor = 10.0f.xxxx;
}
SPIRV_Cross_Output main()

31
reference/opt/shaders-hlsl/asm/frag/image-extract-reuse.asm.frag Normal file
View File

@ -0,0 +1,31 @@
Texture2D<float4> uTexture : register(t0);
SamplerState _uTexture_sampler : register(s0);
static int2 Size;
struct SPIRV_Cross_Output
{
int2 Size : SV_Target0;
};
uint2 SPIRV_Cross_textureSize(Texture2D<float4> Tex, uint Level, out uint Param)
{
uint2 ret;
Tex.GetDimensions(Level, ret.x, ret.y, Param);
return ret;
}
void frag_main()
{
uint _19_dummy_parameter;
uint _20_dummy_parameter;
Size = int2(SPIRV_Cross_textureSize(uTexture, uint(0), _19_dummy_parameter)) + int2(SPIRV_Cross_textureSize(uTexture, uint(1), _20_dummy_parameter));
}
SPIRV_Cross_Output main()
{
frag_main();
SPIRV_Cross_Output stage_output;
stage_output.Size = Size;
return stage_output;
}

57
reference/opt/shaders-hlsl/asm/frag/implicit-read-dep-phi.asm.frag Normal file
View File

@ -0,0 +1,57 @@
Texture2D<float4> uImage : register(t0);
SamplerState _uImage_sampler : register(s0);
static float4 v0;
static float4 FragColor;
struct SPIRV_Cross_Input
{
float4 v0 : TEXCOORD0;
};
struct SPIRV_Cross_Output
{
float4 FragColor : SV_Target0;
};
void frag_main()
{
float phi;
float4 _36;
int _51;
_51 = 0;
phi = 1.0f;
_36 = float4(1.0f, 2.0f, 1.0f, 2.0f);
for (;;)
{
FragColor = _36;
if (_51 < 4)
{
if (v0[_51] > 0.0f)
{
float2 _48 = phi.xx;
_51++;
phi += 2.0f;
_36 = uImage.SampleLevel(_uImage_sampler, _48, 0.0f);
continue;
}
else
{
break;
}
}
else
{
break;
}
}
}
SPIRV_Cross_Output main(SPIRV_Cross_Input stage_input)
{
v0 = stage_input.v0;
frag_main();
SPIRV_Cross_Output stage_output;
stage_output.FragColor = FragColor;
return stage_output;
}

29
reference/opt/shaders-hlsl/asm/frag/srem.asm.frag Normal file
View File

@ -0,0 +1,29 @@
static float4 FragColor;
static int4 vA;
static int4 vB;
struct SPIRV_Cross_Input
{
nointerpolation int4 vA : TEXCOORD0;
nointerpolation int4 vB : TEXCOORD1;
};
struct SPIRV_Cross_Output
{
float4 FragColor : SV_Target0;
};
void frag_main()
{
FragColor = float4(vA - vB * (vA / vB));
}
SPIRV_Cross_Output main(SPIRV_Cross_Input stage_input)
{
vA = stage_input.vA;
vB = stage_input.vB;
frag_main();
SPIRV_Cross_Output stage_output;
stage_output.FragColor = FragColor;
return stage_output;
}

29
reference/opt/shaders-hlsl/asm/frag/texel-fetch-no-lod.asm.frag Normal file
View File

@ -0,0 +1,29 @@
Texture2D<float4> uTexture : register(t0);
SamplerState _uTexture_sampler : register(s0);
static float4 gl_FragCoord;
static float4 FragColor;
struct SPIRV_Cross_Input
{
float4 gl_FragCoord : SV_Position;
};
struct SPIRV_Cross_Output
{
float4 FragColor : SV_Target0;
};
void frag_main()
{
FragColor = uTexture.Load(int3(int2(gl_FragCoord.xy), 0));
}
SPIRV_Cross_Output main(SPIRV_Cross_Input stage_input)
{
gl_FragCoord = stage_input.gl_FragCoord;
frag_main();
SPIRV_Cross_Output stage_output;
stage_output.FragColor = FragColor;
return stage_output;
}

16
reference/opt/shaders-hlsl/asm/frag/unreachable.asm.frag
View File

@ -11,25 +11,23 @@ struct SPIRV_Cross_Output
float4 FragColor : SV_Target0;
};
float4 _21;
void frag_main()
{
float4 _33;
do
bool _29;
for (;;)
{
if (counter == 10)
_29 = counter == 10;
if (_29)
{
_33 = 10.0f.xxxx;
break;
}
else
{
_33 = 30.0f.xxxx;
break;
}
} while (false);
FragColor = _33;
}
bool4 _35 = _29.xxxx;
FragColor = float4(_35.x ? 10.0f.xxxx.x : 30.0f.xxxx.x, _35.y ? 10.0f.xxxx.y : 30.0f.xxxx.y, _35.z ? 10.0f.xxxx.z : 30.0f.xxxx.z, _35.w ? 10.0f.xxxx.w : 30.0f.xxxx.w);
}
SPIRV_Cross_Output main(SPIRV_Cross_Input stage_input)

8
reference/opt/shaders-hlsl/asm/vert/empty-struct-composite.asm.vert
View File

@ -1,8 +0,0 @@
void vert_main()
{
}
void main()
{
vert_main();
}

37
reference/opt/shaders-hlsl/asm/vert/spec-constant-op-composite.asm.vert Normal file
View File

@ -0,0 +1,37 @@
static const int _7 = -10;
static const uint _8 = 100u;
static const int _20 = (_7 + 2);
static const uint _25 = (_8 % 5u);
static const int4 _30 = int4(20, 30, _20, _20);
static const int2 _32 = int2(_30.y, _30.x);
static const int _33 = _30.y;
static float4 gl_Position;
static int _4;
struct SPIRV_Cross_Output
{
nointerpolation int _4 : TEXCOORD0;
float4 gl_Position : SV_Position;
};
void vert_main()
{
float4 _64 = 0.0f.xxxx;
_64.y = float(_20);
float4 _68 = _64;
_68.z = float(_25);
float4 _52 = _68 + float4(_30);
float2 _56 = _52.xy + float2(_32);
gl_Position = float4(_56.x, _56.y, _52.z, _52.w);
_4 = _33;
}
SPIRV_Cross_Output main()
{
vert_main();
SPIRV_Cross_Output stage_output;
stage_output.gl_Position = gl_Position;
stage_output._4 = _4;
return stage_output;
}

2
reference/opt/shaders-hlsl/comp/builtins.comp
View File

@ -1,3 +1,5 @@
static const uint3 gl_WorkGroupSize = uint3(8u, 4u, 2u);
void comp_main()
{
}

38
reference/opt/shaders-hlsl/comp/image.comp
View File

@ -34,24 +34,26 @@ struct SPIRV_Cross_Input
void comp_main()
{
uImageOutF[int2(gl_GlobalInvocationID.xy)] = uImageInF[int2(gl_GlobalInvocationID.xy)].x;
uImageOutI[int2(gl_GlobalInvocationID.xy)] = uImageInI[int2(gl_GlobalInvocationID.xy)].x;
uImageOutU[int2(gl_GlobalInvocationID.xy)] = uImageInU[int2(gl_GlobalInvocationID.xy)].x;
uImageOutBuffer[int(gl_GlobalInvocationID.x)] = uImageInBuffer[int(gl_GlobalInvocationID.x)].x;
uImageOutF2[int2(gl_GlobalInvocationID.xy)] = uImageInF2[int2(gl_GlobalInvocationID.xy)].xy;
uImageOutI2[int2(gl_GlobalInvocationID.xy)] = uImageInI2[int2(gl_GlobalInvocationID.xy)].xy;
uImageOutU2[int2(gl_GlobalInvocationID.xy)] = uImageInU2[int2(gl_GlobalInvocationID.xy)].xy;
float4 _135 = uImageInBuffer2[int(gl_GlobalInvocationID.x)].xyyy;
uImageOutBuffer2[int(gl_GlobalInvocationID.x)] = _135.xy;
uImageOutF4[int2(gl_GlobalInvocationID.xy)] = uImageInF4[int2(gl_GlobalInvocationID.xy)];
int4 _165 = uImageInI4[int2(gl_GlobalInvocationID.xy)];
uImageOutI4[int2(gl_GlobalInvocationID.xy)] = _165;
uint4 _180 = uImageInU4[int2(gl_GlobalInvocationID.xy)];
uImageOutU4[int2(gl_GlobalInvocationID.xy)] = _180;
uImageOutBuffer4[int(gl_GlobalInvocationID.x)] = uImageInBuffer4[int(gl_GlobalInvocationID.x)];
uImageNoFmtF[int2(gl_GlobalInvocationID.xy)] = _135;
uImageNoFmtU[int2(gl_GlobalInvocationID.xy)] = _180;
uImageNoFmtI[int2(gl_GlobalInvocationID.xy)] = _165;
int2 _23 = int2(gl_GlobalInvocationID.xy);
uImageOutF[_23] = uImageInF[_23].x;
uImageOutI[_23] = uImageInI[_23].x;
uImageOutU[_23] = uImageInU[_23].x;
int _74 = int(gl_GlobalInvocationID.x);
uImageOutBuffer[_74] = uImageInBuffer[_74].x;
uImageOutF2[_23] = uImageInF2[_23].xy;
uImageOutI2[_23] = uImageInI2[_23].xy;
uImageOutU2[_23] = uImageInU2[_23].xy;
float4 _135 = uImageInBuffer2[_74].xyyy;
uImageOutBuffer2[_74] = _135.xy;
uImageOutF4[_23] = uImageInF4[_23];
int4 _165 = uImageInI4[_23];
uImageOutI4[_23] = _165;
uint4 _180 = uImageInU4[_23];
uImageOutU4[_23] = _180;
uImageOutBuffer4[_74] = uImageInBuffer4[_74];
uImageNoFmtF[_23] = _135;
uImageNoFmtU[_23] = _180;
uImageNoFmtI[_23] = _165;
}
[numthreads(1, 1, 1)]

20
reference/opt/shaders-hlsl/comp/rmw-matrix.comp Normal file
View File

@ -0,0 +1,20 @@
RWByteAddressBuffer _11 : register(u0);
void comp_main()
{
_11.Store(0, asuint(asfloat(_11.Load(0)) * asfloat(_11.Load(96))));
_11.Store4(16, asuint(asfloat(_11.Load4(16)) * asfloat(_11.Load4(112))));
float4x4 _35 = asfloat(uint4x4(_11.Load4(128), _11.Load4(144), _11.Load4(160), _11.Load4(176)));
float4x4 _37 = asfloat(uint4x4(_11.Load4(32), _11.Load4(48), _11.Load4(64), _11.Load4(80)));
float4x4 _38 = mul(_35, _37);
_11.Store4(32, asuint(_38[0]));
_11.Store4(48, asuint(_38[1]));
_11.Store4(64, asuint(_38[2]));
_11.Store4(80, asuint(_38[3]));
}
[numthreads(1, 1, 1)]
void main()
{
comp_main();
}

2
reference/opt/shaders-hlsl/comp/shared.comp
View File

@ -17,7 +17,7 @@ void comp_main()
{
sShared[gl_LocalInvocationIndex] = asfloat(_22.Load(gl_GlobalInvocationID.x * 4 + 0));
GroupMemoryBarrierWithGroupSync();
_44.Store(gl_GlobalInvocationID.x * 4 + 0, asuint(sShared[(4u - gl_LocalInvocationIndex) - 1u]));
_44.Store(gl_GlobalInvocationID.x * 4 + 0, asuint(sShared[3u - gl_LocalInvocationIndex]));
}
[numthreads(4, 1, 1)]

67
reference/opt/shaders-hlsl/comp/subgroups.invalid.nofxc.sm60.comp Normal file
View File

@ -0,0 +1,67 @@
RWByteAddressBuffer _9 : register(u0, space0);
static uint4 gl_SubgroupEqMask;
static uint4 gl_SubgroupGeMask;
static uint4 gl_SubgroupGtMask;
static uint4 gl_SubgroupLeMask;
static uint4 gl_SubgroupLtMask;
void comp_main()
{
_9.Store(0, asuint(float(WaveGetLaneCount())));
_9.Store(0, asuint(float(WaveGetLaneIndex())));
_9.Store(0, asuint(float4(gl_SubgroupEqMask).x));
_9.Store(0, asuint(float4(gl_SubgroupGeMask).x));
_9.Store(0, asuint(float4(gl_SubgroupGtMask).x));
_9.Store(0, asuint(float4(gl_SubgroupLeMask).x));
_9.Store(0, asuint(float4(gl_SubgroupLtMask).x));
uint4 _75 = WaveActiveBallot(true);
float4 _88 = WaveActiveSum(20.0f.xxxx);
int4 _94 = WaveActiveSum(int4(20, 20, 20, 20));
float4 _96 = WaveActiveProduct(20.0f.xxxx);
int4 _98 = WaveActiveProduct(int4(20, 20, 20, 20));
float4 _127 = WavePrefixProduct(_96) * _96;
int4 _129 = WavePrefixProduct(_98) * _98;
}
[numthreads(1, 1, 1)]
void main()
{
gl_SubgroupEqMask = 1u << (WaveGetLaneIndex() - uint4(0, 32, 64, 96));
if (WaveGetLaneIndex() >= 32) gl_SubgroupEqMask.x = 0;
if (WaveGetLaneIndex() >= 64 || WaveGetLaneIndex() < 32) gl_SubgroupEqMask.y = 0;
if (WaveGetLaneIndex() >= 96 || WaveGetLaneIndex() < 64) gl_SubgroupEqMask.z = 0;
if (WaveGetLaneIndex() < 96) gl_SubgroupEqMask.w = 0;
gl_SubgroupGeMask = ~((1u << (WaveGetLaneIndex() - uint4(0, 32, 64, 96))) - 1u);
if (WaveGetLaneIndex() >= 32) gl_SubgroupGeMask.x = 0u;
if (WaveGetLaneIndex() >= 64) gl_SubgroupGeMask.y = 0u;
if (WaveGetLaneIndex() >= 96) gl_SubgroupGeMask.z = 0u;
if (WaveGetLaneIndex() < 32) gl_SubgroupGeMask.y = ~0u;
if (WaveGetLaneIndex() < 64) gl_SubgroupGeMask.z = ~0u;
if (WaveGetLaneIndex() < 96) gl_SubgroupGeMask.w = ~0u;
uint gt_lane_index = WaveGetLaneIndex() + 1;
gl_SubgroupGtMask = ~((1u << (gt_lane_index - uint4(0, 32, 64, 96))) - 1u);
if (gt_lane_index >= 32) gl_SubgroupGtMask.x = 0u;
if (gt_lane_index >= 64) gl_SubgroupGtMask.y = 0u;
if (gt_lane_index >= 96) gl_SubgroupGtMask.z = 0u;
if (gt_lane_index >= 128) gl_SubgroupGtMask.w = 0u;
if (gt_lane_index < 32) gl_SubgroupGtMask.y = ~0u;
if (gt_lane_index < 64) gl_SubgroupGtMask.z = ~0u;
if (gt_lane_index < 96) gl_SubgroupGtMask.w = ~0u;
uint le_lane_index = WaveGetLaneIndex() + 1;
gl_SubgroupLeMask = (1u << (le_lane_index - uint4(0, 32, 64, 96))) - 1u;
if (le_lane_index >= 32) gl_SubgroupLeMask.x = ~0u;
if (le_lane_index >= 64) gl_SubgroupLeMask.y = ~0u;
if (le_lane_index >= 96) gl_SubgroupLeMask.z = ~0u;
if (le_lane_index >= 128) gl_SubgroupLeMask.w = ~0u;
if (le_lane_index < 32) gl_SubgroupLeMask.y = 0u;
if (le_lane_index < 64) gl_SubgroupLeMask.z = 0u;
if (le_lane_index < 96) gl_SubgroupLeMask.w = 0u;
gl_SubgroupLtMask = (1u << (WaveGetLaneIndex() - uint4(0, 32, 64, 96))) - 1u;
if (WaveGetLaneIndex() >= 32) gl_SubgroupLtMask.x = ~0u;
if (WaveGetLaneIndex() >= 64) gl_SubgroupLtMask.y = ~0u;
if (WaveGetLaneIndex() >= 96) gl_SubgroupLtMask.z = ~0u;
if (WaveGetLaneIndex() < 32) gl_SubgroupLtMask.y = 0u;
if (WaveGetLaneIndex() < 64) gl_SubgroupLtMask.z = 0u;
if (WaveGetLaneIndex() < 96) gl_SubgroupLtMask.w = 0u;
comp_main();
}

35
reference/opt/shaders-hlsl/frag/array-lut-no-loop-variable.frag Normal file
View File

@ -0,0 +1,35 @@
static const float _17[5] = { 1.0f, 2.0f, 3.0f, 4.0f, 5.0f };
static float4 FragColor;
static float4 v0;
struct SPIRV_Cross_Input
{
float4 v0 : TEXCOORD0;
};
struct SPIRV_Cross_Output
{
float4 FragColor : SV_Target0;
};
void frag_main()
{
float lut[5] = _17;
for (int _46 = 0; _46 < 4; )
{
int _33 = _46 + 1;
FragColor += lut[_33].xxxx;
_46 = _33;
continue;
}
}
SPIRV_Cross_Output main(SPIRV_Cross_Input stage_input)
{
v0 = stage_input.v0;
frag_main();
SPIRV_Cross_Output stage_output;
stage_output.FragColor = FragColor;
return stage_output;
}

54
reference/opt/shaders-hlsl/frag/control-dependent-in-branch.desktop.frag Normal file
View File

@ -0,0 +1,54 @@
Texture2D<float4> uSampler : register(t0);
SamplerState _uSampler_sampler : register(s0);
static float4 FragColor;
static float4 vInput;
struct SPIRV_Cross_Input
{
float4 vInput : TEXCOORD0;
};
struct SPIRV_Cross_Output
{
float4 FragColor : SV_Target0;
};
void frag_main()
{
FragColor = vInput;
float4 _23 = uSampler.Sample(_uSampler_sampler, vInput.xy);
float4 _26 = ddx(vInput);
float4 _29 = ddy(vInput);
float4 _32 = fwidth(vInput);
float4 _35 = ddx_coarse(vInput);
float4 _38 = ddy_coarse(vInput);
float4 _41 = fwidth(vInput);
float4 _44 = ddx_fine(vInput);
float4 _47 = ddy_fine(vInput);
float4 _50 = fwidth(vInput);
float _56_tmp = uSampler.CalculateLevelOfDetail(_uSampler_sampler, vInput.zw);
if (vInput.y > 10.0f)
{
FragColor += _23;
FragColor += _26;
FragColor += _29;
FragColor += _32;
FragColor += _35;
FragColor += _38;
FragColor += _41;
FragColor += _44;
FragColor += _47;
FragColor += _50;
FragColor += float2(_56_tmp, _56_tmp).xyxy;
}
}
SPIRV_Cross_Output main(SPIRV_Cross_Input stage_input)
{
vInput = stage_input.vInput;
frag_main();
SPIRV_Cross_Output stage_output;
stage_output.FragColor = FragColor;
return stage_output;
}

45
reference/opt/shaders-hlsl/frag/fp16.desktop.frag Normal file
View File

@ -0,0 +1,45 @@
static min16float4 v4;
static min16float3 v3;
static min16float v1;
static min16float2 v2;
static float o1;
static float2 o2;
static float3 o3;
static float4 o4;
struct SPIRV_Cross_Input
{
min16float v1 : TEXCOORD0;
min16float2 v2 : TEXCOORD1;
min16float3 v3 : TEXCOORD2;
min16float4 v4 : TEXCOORD3;
};
struct SPIRV_Cross_Output
{
float o1 : SV_Target0;
float2 o2 : SV_Target1;
float3 o3 : SV_Target2;
float4 o4 : SV_Target3;
};
void frag_main()
{
min16float4 _324;
min16float4 _387 = modf(v4, _324);
}
SPIRV_Cross_Output main(SPIRV_Cross_Input stage_input)
{
v4 = stage_input.v4;
v3 = stage_input.v3;
v1 = stage_input.v1;
v2 = stage_input.v2;
frag_main();
SPIRV_Cross_Output stage_output;
stage_output.o1 = o1;
stage_output.o2 = o2;
stage_output.o3 = o3;
stage_output.o4 = o4;
return stage_output;
}

101
reference/opt/shaders-hlsl/frag/image-query-selective.frag
View File

@ -2,6 +2,30 @@ Texture1D<uint4> uSampler1DUint : register(t0);
SamplerState _uSampler1DUint_sampler : register(s0);
Texture1D<int4> uSampler1DInt : register(t0);
SamplerState _uSampler1DInt_sampler : register(s0);
Texture1D<float4> uSampler1DFloat : register(t0);
SamplerState _uSampler1DFloat_sampler : register(s0);
Texture2DArray<int4> uSampler2DArray : register(t2);
SamplerState _uSampler2DArray_sampler : register(s2);
Texture3D<float4> uSampler3D : register(t3);
SamplerState _uSampler3D_sampler : register(s3);
TextureCube<float4> uSamplerCube : register(t4);
SamplerState _uSamplerCube_sampler : register(s4);
TextureCubeArray<uint4> uSamplerCubeArray : register(t5);
SamplerState _uSamplerCubeArray_sampler : register(s5);
Buffer<float4> uSamplerBuffer : register(t6);
Texture2DMS<int4> uSamplerMS : register(t7);
SamplerState _uSamplerMS_sampler : register(s7);
Texture2DMSArray<float4> uSamplerMSArray : register(t8);
SamplerState _uSamplerMSArray_sampler : register(s8);
Texture2D<float4> uSampler2D : register(t1);
SamplerState _uSampler2D_sampler : register(s1);
uint SPIRV_Cross_textureSize(Texture1D<float4> Tex, uint Level, out uint Param)
{
uint ret;
Tex.GetDimensions(Level, ret.x, Param);
return ret;
}
uint SPIRV_Cross_textureSize(Texture1D<int4> Tex, uint Level, out uint Param)
{
@ -17,10 +41,87 @@ uint SPIRV_Cross_textureSize(Texture1D<uint4> Tex, uint Level, out uint Param)
return ret;
}
uint2 SPIRV_Cross_textureSize(Texture2D<float4> Tex, uint Level, out uint Param)
{
uint2 ret;
Tex.GetDimensions(Level, ret.x, ret.y, Param);
return ret;
}
uint3 SPIRV_Cross_textureSize(Texture2DArray<int4> Tex, uint Level, out uint Param)
{
uint3 ret;
Tex.GetDimensions(Level, ret.x, ret.y, ret.z, Param);
return ret;
}
uint3 SPIRV_Cross_textureSize(Texture3D<float4> Tex, uint Level, out uint Param)
{
uint3 ret;
Tex.GetDimensions(Level, ret.x, ret.y, ret.z, Param);
return ret;
}
uint SPIRV_Cross_textureSize(Buffer<float4> Tex, uint Level, out uint Param)
{
uint ret;
Tex.GetDimensions(ret.x);
Param = 0u;
return ret;
}
uint2 SPIRV_Cross_textureSize(TextureCube<float4> Tex, uint Level, out uint Param)
{
uint2 ret;
Tex.GetDimensions(Level, ret.x, ret.y, Param);
return ret;
}
uint3 SPIRV_Cross_textureSize(TextureCubeArray<uint4> Tex, uint Level, out uint Param)
{
uint3 ret;
Tex.GetDimensions(Level, ret.x, ret.y, ret.z, Param);
return ret;
}
uint2 SPIRV_Cross_textureSize(Texture2DMS<int4> Tex, uint Level, out uint Param)
{
uint2 ret;
Tex.GetDimensions(ret.x, ret.y, Param);
return ret;
}
uint3 SPIRV_Cross_textureSize(Texture2DMSArray<float4> Tex, uint Level, out uint Param)
{
uint3 ret;
Tex.GetDimensions(ret.x, ret.y, ret.z, Param);
return ret;
}
void frag_main()
{
uint _17_dummy_parameter;
uint _24_dummy_parameter;
uint _32_dummy_parameter;
uint _42_dummy_parameter;
uint _50_dummy_parameter;
uint _60_dummy_parameter;
uint _68_dummy_parameter;
uint _76_dummy_parameter;
uint _84_dummy_parameter;
uint _92_dummy_parameter;
int _100;
SPIRV_Cross_textureSize(uSampler2D, 0u, _100);
int _104;
SPIRV_Cross_textureSize(uSampler2DArray, 0u, _104);
int _108;
SPIRV_Cross_textureSize(uSampler3D, 0u, _108);
int _112;
SPIRV_Cross_textureSize(uSamplerCube, 0u, _112);
int _116;
SPIRV_Cross_textureSize(uSamplerMS, 0u, _116);
int _120;
SPIRV_Cross_textureSize(uSamplerMSArray, 0u, _120);
}
void main()

107
reference/opt/shaders-hlsl/frag/image-query.frag
View File

@ -1,5 +1,112 @@
Texture1D<float4> uSampler1D : register(t0);
SamplerState _uSampler1D_sampler : register(s0);
Texture2D<float4> uSampler2D : register(t1);
SamplerState _uSampler2D_sampler : register(s1);
Texture2DArray<float4> uSampler2DArray : register(t2);
SamplerState _uSampler2DArray_sampler : register(s2);
Texture3D<float4> uSampler3D : register(t3);
SamplerState _uSampler3D_sampler : register(s3);
TextureCube<float4> uSamplerCube : register(t4);
SamplerState _uSamplerCube_sampler : register(s4);
TextureCubeArray<float4> uSamplerCubeArray : register(t5);
SamplerState _uSamplerCubeArray_sampler : register(s5);
Buffer<float4> uSamplerBuffer : register(t6);
Texture2DMS<float4> uSamplerMS : register(t7);
SamplerState _uSamplerMS_sampler : register(s7);
Texture2DMSArray<float4> uSamplerMSArray : register(t8);
SamplerState _uSamplerMSArray_sampler : register(s8);
uint SPIRV_Cross_textureSize(Texture1D<float4> Tex, uint Level, out uint Param)
{
uint ret;
Tex.GetDimensions(Level, ret.x, Param);
return ret;
}
uint2 SPIRV_Cross_textureSize(Texture2D<float4> Tex, uint Level, out uint Param)
{
uint2 ret;
Tex.GetDimensions(Level, ret.x, ret.y, Param);
return ret;
}
uint3 SPIRV_Cross_textureSize(Texture2DArray<float4> Tex, uint Level, out uint Param)
{
uint3 ret;
Tex.GetDimensions(Level, ret.x, ret.y, ret.z, Param);
return ret;
}
uint3 SPIRV_Cross_textureSize(Texture3D<float4> Tex, uint Level, out uint Param)
{
uint3 ret;
Tex.GetDimensions(Level, ret.x, ret.y, ret.z, Param);
return ret;
}
uint SPIRV_Cross_textureSize(Buffer<float4> Tex, uint Level, out uint Param)
{
uint ret;
Tex.GetDimensions(ret.x);
Param = 0u;
return ret;
}
uint2 SPIRV_Cross_textureSize(TextureCube<float4> Tex, uint Level, out uint Param)
{
uint2 ret;
Tex.GetDimensions(Level, ret.x, ret.y, Param);
return ret;
}
uint3 SPIRV_Cross_textureSize(TextureCubeArray<float4> Tex, uint Level, out uint Param)
{
uint3 ret;
Tex.GetDimensions(Level, ret.x, ret.y, ret.z, Param);
return ret;
}
uint2 SPIRV_Cross_textureSize(Texture2DMS<float4> Tex, uint Level, out uint Param)
{
uint2 ret;
Tex.GetDimensions(ret.x, ret.y, Param);
return ret;
}
uint3 SPIRV_Cross_textureSize(Texture2DMSArray<float4> Tex, uint Level, out uint Param)
{
uint3 ret;
Tex.GetDimensions(ret.x, ret.y, ret.z, Param);
return ret;
}
void frag_main()
{
uint _17_dummy_parameter;
uint _27_dummy_parameter;
uint _37_dummy_parameter;
uint _45_dummy_parameter;
uint _53_dummy_parameter;
uint _61_dummy_parameter;
uint _69_dummy_parameter;
uint _77_dummy_parameter;
uint _85_dummy_parameter;
int _89;
SPIRV_Cross_textureSize(uSampler1D, 0u, _89);
int _93;
SPIRV_Cross_textureSize(uSampler2D, 0u, _93);
int _97;
SPIRV_Cross_textureSize(uSampler2DArray, 0u, _97);
int _101;
SPIRV_Cross_textureSize(uSampler3D, 0u, _101);
int _105;
SPIRV_Cross_textureSize(uSamplerCube, 0u, _105);
int _109;
SPIRV_Cross_textureSize(uSamplerCubeArray, 0u, _109);
int _113;
SPIRV_Cross_textureSize(uSamplerMS, 0u, _113);
int _117;
SPIRV_Cross_textureSize(uSamplerMSArray, 0u, _117);
}
void main()

6
reference/opt/shaders-hlsl/frag/partial-write-preserve.frag
View File

@ -1,9 +1,3 @@
struct B
{
float a;
float b;
};
void frag_main()
{
}

4
reference/opt/shaders-hlsl/frag/resources.frag
View File

@ -4,7 +4,7 @@ cbuffer cbuf : register(b3)
};
cbuffer registers
{
float4 registers_d : packoffset(c0);
float4 registers_a : packoffset(c0);
};
Texture2D<float4> uSampledImage : register(t4);
SamplerState _uSampledImage_sampler : register(s4);
@ -26,7 +26,7 @@ struct SPIRV_Cross_Output
void frag_main()
{
FragColor = (uSampledImage.Sample(_uSampledImage_sampler, vTex) + uTexture.Sample(uSampler, vTex)) + (cbuf_a + registers_d);
FragColor = (uSampledImage.Sample(_uSampledImage_sampler, vTex) + uTexture.Sample(uSampler, vTex)) + (cbuf_a + registers_a);
}
SPIRV_Cross_Output main(SPIRV_Cross_Input stage_input)

37
reference/opt/shaders-hlsl/frag/row-major-layout-in-struct.frag Normal file
View File

@ -0,0 +1,37 @@
struct Foo
{
row_major float4x4 v;
row_major float4x4 w;
};
cbuffer _17 : register(b0)
{
Foo _17_foo : packoffset(c0);
};
static float4 FragColor;
static float4 vUV;
struct SPIRV_Cross_Input
{
float4 vUV : TEXCOORD0;
};
struct SPIRV_Cross_Output
{
float4 FragColor : SV_Target0;
};
void frag_main()
{
FragColor = mul(mul(vUV, _17_foo.w), _17_foo.v);
}
SPIRV_Cross_Output main(SPIRV_Cross_Input stage_input)
{
vUV = stage_input.vUV;
frag_main();
SPIRV_Cross_Output stage_output;
stage_output.FragColor = FragColor;
return stage_output;
}

4
reference/opt/shaders-hlsl/frag/sample-cmp-level-zero.frag
View File

@ -41,9 +41,7 @@ void frag_main()
{
float4 _80 = vDirRef;
_80.z = vDirRef.w;
float4 _87 = vDirRef;
_87.z = vDirRef.w;
FragColor = (((((((uSampler2D.SampleCmp(_uSampler2D_sampler, vUVRef.xy, vUVRef.z, int2(-1, -1)) + uSampler2DArray.SampleCmp(_uSampler2DArray_sampler, vDirRef.xyz, vDirRef.w, int2(-1, -1))) + uSamplerCube.SampleCmp(_uSamplerCube_sampler, vDirRef.xyz, vDirRef.w)) + uSamplerCubeArray.SampleCmp(_uSamplerCubeArray_sampler, vDirRef, 0.5f)) + uSampler2D.SampleCmpLevelZero(_uSampler2D_sampler, vUVRef.xy, vUVRef.z, int2(-1, -1))) + uSampler2DArray.SampleCmpLevelZero(_uSampler2DArray_sampler, vDirRef.xyz, vDirRef.w, int2(-1, -1))) + uSamplerCube.SampleCmpLevelZero(_uSamplerCube_sampler, vDirRef.xyz, vDirRef.w)) + uSampler2D.SampleCmp(_uSampler2D_sampler, SPIRV_Cross_projectTextureCoordinate(_80.xyz), vDirRef.z, int2(1, 1))) + uSampler2D.SampleCmpLevelZero(_uSampler2D_sampler, SPIRV_Cross_projectTextureCoordinate(_87.xyz), vDirRef.z, int2(1, 1));
FragColor = (((((((uSampler2D.SampleCmp(_uSampler2D_sampler, vUVRef.xy, vUVRef.z, int2(-1, -1)) + uSampler2DArray.SampleCmp(_uSampler2DArray_sampler, vDirRef.xyz, vDirRef.w, int2(-1, -1))) + uSamplerCube.SampleCmp(_uSamplerCube_sampler, vDirRef.xyz, vDirRef.w)) + uSamplerCubeArray.SampleCmp(_uSamplerCubeArray_sampler, vDirRef, 0.5f)) + uSampler2D.SampleCmpLevelZero(_uSampler2D_sampler, vUVRef.xy, vUVRef.z, int2(-1, -1))) + uSampler2DArray.SampleCmpLevelZero(_uSampler2DArray_sampler, vDirRef.xyz, vDirRef.w, int2(-1, -1))) + uSamplerCube.SampleCmpLevelZero(_uSamplerCube_sampler, vDirRef.xyz, vDirRef.w)) + uSampler2D.SampleCmp(_uSampler2D_sampler, SPIRV_Cross_projectTextureCoordinate(_80.xyz), vDirRef.z, int2(1, 1))) + uSampler2D.SampleCmpLevelZero(_uSampler2D_sampler, SPIRV_Cross_projectTextureCoordinate(_80.xyz), vDirRef.z, int2(1, 1));
}
SPIRV_Cross_Output main(SPIRV_Cross_Input stage_input)

3
reference/opt/shaders-hlsl/frag/sampler-array.frag
View File

@ -17,7 +17,8 @@ struct SPIRV_Cross_Input
void frag_main()
{
uImage[vIndex][int2(gl_FragCoord.xy)] = ((uCombined[vIndex].Sample(_uCombined_sampler[vIndex], vTex) + uTex[vIndex].Sample(uSampler[vIndex], vTex)) + (uCombined[vIndex + 1].Sample(_uCombined_sampler[vIndex + 1], vTex))) + (uTex[vIndex + 1].Sample(uSampler[vIndex + 1], vTex));
int _72 = vIndex + 1;
uImage[vIndex][int2(gl_FragCoord.xy)] = ((uCombined[vIndex].Sample(_uCombined_sampler[vIndex], vTex) + uTex[vIndex].Sample(uSampler[vIndex], vTex)) + uCombined[_72].Sample(_uCombined_sampler[_72], vTex)) + uTex[_72].Sample(uSampler[_72], vTex);
}
void main(SPIRV_Cross_Input stage_input)

39
reference/opt/shaders-hlsl/frag/sampler-image-arrays.frag Normal file
View File

@ -0,0 +1,39 @@
Texture2D<float4> uSampler[4] : register(t0);
SamplerState _uSampler_sampler[4] : register(s0);
Texture2D<float4> uTextures[4] : register(t8);
SamplerState uSamplers[4] : register(s4);
static int vIndex;
static float2 vTex;
static float4 FragColor;
struct SPIRV_Cross_Input
{
nointerpolation float2 vTex : TEXCOORD0;
nointerpolation int vIndex : TEXCOORD1;
};
struct SPIRV_Cross_Output
{
float4 FragColor : SV_Target0;
};
void frag_main()
{
FragColor = 0.0f.xxxx;
FragColor += uTextures[2].Sample(uSamplers[1], vTex);
FragColor += uSampler[vIndex].Sample(_uSampler_sampler[vIndex], vTex);
FragColor += uSampler[vIndex].Sample(_uSampler_sampler[vIndex], vTex + 0.100000001490116119384765625f.xx);
FragColor += uSampler[vIndex].Sample(_uSampler_sampler[vIndex], vTex + 0.20000000298023223876953125f.xx);
FragColor += uSampler[3].Sample(_uSampler_sampler[3], vTex + 0.300000011920928955078125f.xx);
}
SPIRV_Cross_Output main(SPIRV_Cross_Input stage_input)
{
vIndex = stage_input.vIndex;
vTex = stage_input.vTex;
frag_main();
SPIRV_Cross_Output stage_output;
stage_output.FragColor = FragColor;
return stage_output;
}

33
reference/opt/shaders-hlsl/frag/spec-constant.frag
View File

@ -1,33 +0,0 @@
static const float a = 1.0f;
static const float b = 2.0f;
static const int c = 3;
static const int d = 4;
struct Foo
{
float elems[(d + 2)];
};
static float4 FragColor;
struct SPIRV_Cross_Output
{
float4 FragColor : SV_Target0;
};
void frag_main()
{
float vec0[(c + 3)][8];
vec0[0][0] = 10.0f;
Foo foo;
foo.elems[c] = 10.0f;
FragColor = (((a + b).xxxx + vec0[0][0].xxxx) + 20.0f.xxxx) + foo.elems[c].xxxx;
}
SPIRV_Cross_Output main()
{
frag_main();
SPIRV_Cross_Output stage_output;
stage_output.FragColor = FragColor;
return stage_output;
}

33
reference/opt/shaders-hlsl/frag/tex-sampling-ms.frag Normal file
View File

@ -0,0 +1,33 @@
Texture2DMS<float4> uTex : register(t0);
SamplerState _uTex_sampler : register(s0);
static float4 gl_FragCoord;
static float4 FragColor;
struct SPIRV_Cross_Input
{
float4 gl_FragCoord : SV_Position;
};
struct SPIRV_Cross_Output
{
float4 FragColor : SV_Target0;
};
void frag_main()
{
int2 _22 = int2(gl_FragCoord.xy);
FragColor = uTex.Load(_22, 0);
FragColor += uTex.Load(_22, 1);
FragColor += uTex.Load(_22, 2);
FragColor += uTex.Load(_22, 3);
}
SPIRV_Cross_Output main(SPIRV_Cross_Input stage_input)
{
gl_FragCoord = stage_input.gl_FragCoord;
frag_main();
SPIRV_Cross_Output stage_output;
stage_output.FragColor = FragColor;
return stage_output;
}

58
reference/opt/shaders-hlsl/frag/tex-sampling.frag
View File

@ -1,27 +1,27 @@
Texture1D<float4> tex1d;
SamplerState _tex1d_sampler;
Texture2D<float4> tex2d;
SamplerState _tex2d_sampler;
Texture3D<float4> tex3d;
SamplerState _tex3d_sampler;
TextureCube<float4> texCube;
SamplerState _texCube_sampler;
Texture1D<float4> tex1dShadow;
SamplerComparisonState _tex1dShadow_sampler;
Texture2D<float4> tex2dShadow;
SamplerComparisonState _tex2dShadow_sampler;
TextureCube<float4> texCubeShadow;
SamplerComparisonState _texCubeShadow_sampler;
Texture1DArray<float4> tex1dArray;
SamplerState _tex1dArray_sampler;
Texture2DArray<float4> tex2dArray;
SamplerState _tex2dArray_sampler;
TextureCubeArray<float4> texCubeArray;
SamplerState _texCubeArray_sampler;
Texture2D<float4> separateTex2d;
SamplerState samplerNonDepth;
Texture2D<float4> separateTex2dDepth;
SamplerComparisonState samplerDepth;
Texture1D<float4> tex1d : register(t0);
SamplerState _tex1d_sampler : register(s0);
Texture2D<float4> tex2d : register(t1);
SamplerState _tex2d_sampler : register(s1);
Texture3D<float4> tex3d : register(t2);
SamplerState _tex3d_sampler : register(s2);
TextureCube<float4> texCube : register(t3);
SamplerState _texCube_sampler : register(s3);
Texture1D<float4> tex1dShadow : register(t4);
SamplerComparisonState _tex1dShadow_sampler : register(s4);
Texture2D<float4> tex2dShadow : register(t5);
SamplerComparisonState _tex2dShadow_sampler : register(s5);
TextureCube<float4> texCubeShadow : register(t6);
SamplerComparisonState _texCubeShadow_sampler : register(s6);
Texture1DArray<float4> tex1dArray : register(t7);
SamplerState _tex1dArray_sampler : register(s7);
Texture2DArray<float4> tex2dArray : register(t8);
SamplerState _tex2dArray_sampler : register(s8);
TextureCubeArray<float4> texCubeArray : register(t9);
SamplerState _texCubeArray_sampler : register(s9);
Texture2D<float4> separateTex2d : register(t12);
SamplerState samplerNonDepth : register(s11);
Texture2D<float4> separateTex2dDepth : register(t13);
SamplerComparisonState samplerDepth : register(s10);
static float texCoord1d;
static float2 texCoord2d;
@ -60,14 +60,8 @@ float3 SPIRV_Cross_projectTextureCoordinate(float4 coord)
void frag_main()
{
float4 _162 = (((((((((((((((((((tex1d.Sample(_tex1d_sampler, texCoord1d) + tex1d.Sample(_tex1d_sampler, texCoord1d, 1)) + tex1d.SampleLevel(_tex1d_sampler, texCoord1d, 2.0f)) + tex1d.SampleGrad(_tex1d_sampler, texCoord1d, 1.0f, 2.0f)) + tex1d.Sample(_tex1d_sampler, SPIRV_Cross_projectTextureCoordinate(float2(texCoord1d, 2.0f)))) + tex1d.SampleBias(_tex1d_sampler, texCoord1d, 1.0f)) + tex2d.Sample(_tex2d_sampler, texCoord2d)) + tex2d.Sample(_tex2d_sampler, texCoord2d, int2(1, 2))) + tex2d.SampleLevel(_tex2d_sampler, texCoord2d, 2.0f)) + tex2d.SampleGrad(_tex2d_sampler, texCoord2d, float2(1.0f, 2.0f), float2(3.0f, 4.0f))) + tex2d.Sample(_tex2d_sampler, SPIRV_Cross_projectTextureCoordinate(float3(texCoord2d, 2.0f)))) + tex2d.SampleBias(_tex2d_sampler, texCoord2d, 1.0f)) + tex3d.Sample(_tex3d_sampler, texCoord3d)) + tex3d.Sample(_tex3d_sampler, texCoord3d, int3(1, 2, 3))) + tex3d.SampleLevel(_tex3d_sampler, texCoord3d, 2.0f)) + tex3d.SampleGrad(_tex3d_sampler, texCoord3d, float3(1.0f, 2.0f, 3.0f), float3(4.0f, 5.0f, 6.0f))) + tex3d.Sample(_tex3d_sampler, SPIRV_Cross_projectTextureCoordinate(float4(texCoord3d, 2.0f)))) + tex3d.SampleBias(_tex3d_sampler, texCoord3d, 1.0f)) + texCube.Sample(_texCube_sampler, texCoord3d)) + texCube.SampleLevel(_texCube_sampler, texCoord3d, 2.0f)) + texCube.SampleBias(_texCube_sampler, texCoord3d, 1.0f);
float _178 = _162.w + tex1dShadow.SampleCmp(_tex1dShadow_sampler, float3(texCoord1d, 0.0f, 0.0f).x, 0.0f);
float4 _327 = _162;
_327.w = _178;
float _193 = _178 + tex2dShadow.SampleCmp(_tex2dShadow_sampler, float3(texCoord2d, 0.0f).xy, 0.0f);
float4 _331 = _327;
_331.w = _193;
float4 _335 = _331;
_335.w = _193 + texCubeShadow.SampleCmp(_texCubeShadow_sampler, float4(texCoord3d, 0.0f).xyz, 0.0f);
float4 _335 = _162;
_335.w = ((_162.w + tex1dShadow.SampleCmp(_tex1dShadow_sampler, float3(texCoord1d, 0.0f, 0.0f).x, 0.0f)) + tex2dShadow.SampleCmp(_tex2dShadow_sampler, float3(texCoord2d, 0.0f).xy, 0.0f)) + texCubeShadow.SampleCmp(_texCubeShadow_sampler, float4(texCoord3d, 0.0f).xyz, 0.0f);
float4 _308 = ((((((((((((((_335 + tex1dArray.Sample(_tex1dArray_sampler, texCoord2d)) + tex2dArray.Sample(_tex2dArray_sampler, texCoord3d)) + texCubeArray.Sample(_texCubeArray_sampler, texCoord4d)) + tex2d.GatherRed(_tex2d_sampler, texCoord2d)) + tex2d.GatherRed(_tex2d_sampler, texCoord2d)) + tex2d.GatherGreen(_tex2d_sampler, texCoord2d)) + tex2d.GatherBlue(_tex2d_sampler, texCoord2d)) + tex2d.GatherAlpha(_tex2d_sampler, texCoord2d)) + tex2d.GatherRed(_tex2d_sampler, texCoord2d, int2(1, 1))) + tex2d.GatherRed(_tex2d_sampler, texCoord2d, int2(1, 1))) + tex2d.GatherGreen(_tex2d_sampler, texCoord2d, int2(1, 1))) + tex2d.GatherBlue(_tex2d_sampler, texCoord2d, int2(1, 1))) + tex2d.GatherAlpha(_tex2d_sampler, texCoord2d, int2(1, 1))) + tex2d.Load(int3(int2(1, 2), 0))) + separateTex2d.Sample(samplerNonDepth, texCoord2d);
float4 _339 = _308;
_339.w = _308.w + separateTex2dDepth.SampleCmp(samplerDepth, texCoord3d.xy, texCoord3d.z);

2
reference/opt/shaders-hlsl/frag/unary-enclose.frag
View File

@ -15,7 +15,7 @@ struct SPIRV_Cross_Output
void frag_main()
{
FragColor = -(-vIn);
FragColor = vIn;
}
SPIRV_Cross_Output main(SPIRV_Cross_Input stage_input)

12
reference/opt/shaders-hlsl/vert/locations.vert
View File

@ -5,6 +5,8 @@ struct Foo
float3 c;
};
static const Foo _71 = { 1.0f.xxx, 1.0f.xxx, 1.0f.xxx };
static float4 gl_Position;
static float4 Input2;
static float4 Input4;
@ -40,8 +42,6 @@ struct SPIRV_Cross_Output
float4 gl_Position : SV_Position;
};
Foo _70;
void vert_main()
{
gl_Position = ((1.0f.xxxx + Input2) + Input4) + Input0;
@ -49,13 +49,7 @@ void vert_main()
vLocation1 = 1.0f;
vLocation2[0] = 2.0f;
vLocation2[1] = 2.0f;
Foo _65 = _70;
_65.a = 1.0f.xxx;
Foo _67 = _65;
_67.b = 1.0f.xxx;
Foo _69 = _67;
_69.c = 1.0f.xxx;
vLocation4 = _69;
vLocation4 = _71;
vLocation9 = 9.0f;
vout.color = 2.0f.xxx;
vout.foo = 4.0f.xxx;

35
reference/opt/shaders-hlsl/vert/read-from-row-major-array.vert Normal file
View File

@ -0,0 +1,35 @@
cbuffer _104 : register(b0)
{
column_major float2x3 _104_var[3][4] : packoffset(c0);
};
static float4 gl_Position;
static float4 a_position;
static float v_vtxResult;
struct SPIRV_Cross_Input
{
float4 a_position : TEXCOORD0;
};
struct SPIRV_Cross_Output
{
float v_vtxResult : TEXCOORD0;
float4 gl_Position : SV_Position;
};
void vert_main()
{
gl_Position = a_position;
v_vtxResult = ((float(abs(_104_var[0][0][0].x - 2.0f) < 0.0500000007450580596923828125f) * float(abs(_104_var[0][0][0].y - 6.0f) < 0.0500000007450580596923828125f)) * float(abs(_104_var[0][0][0].z - (-6.0f)) < 0.0500000007450580596923828125f)) * ((float(abs(_104_var[0][0][1].x) < 0.0500000007450580596923828125f) * float(abs(_104_var[0][0][1].y - 5.0f) < 0.0500000007450580596923828125f)) * float(abs(_104_var[0][0][1].z - 5.0f) < 0.0500000007450580596923828125f));
}
SPIRV_Cross_Output main(SPIRV_Cross_Input stage_input)
{
a_position = stage_input.a_position;
vert_main();
SPIRV_Cross_Output stage_output;
stage_output.gl_Position = gl_Position;
stage_output.v_vtxResult = v_vtxResult;
return stage_output;
}

2
reference/opt/shaders-hlsl/vert/return-array.vert
View File

@ -1,5 +1,3 @@
static const float4 _20[2] = { 10.0f.xxxx, 20.0f.xxxx };
static float4 gl_Position;
static float4 vInput0;
static float4 vInput1;

18
reference/opt/shaders-msl/asm/comp/bitcast_sar.asm.comp
View File

@ -17,13 +17,15 @@ struct _4
kernel void main0(device _3& _5 [[buffer(0)]], device _4& _6 [[buffer(1)]])
{
_6._m0 = uint4(int4(_5._m1) >> _5._m0);
_6._m0 = uint4(_5._m0 >> int4(_5._m1));
_6._m0 = uint4(int4(_5._m1) >> int4(_5._m1));
_6._m0 = uint4(_5._m0 >> _5._m0);
_6._m1 = int4(_5._m1) >> int4(_5._m1);
_6._m1 = _5._m0 >> _5._m0;
_6._m1 = int4(_5._m1) >> _5._m0;
_6._m1 = _5._m0 >> int4(_5._m1);
int4 _22 = _5._m0;
uint4 _23 = _5._m1;
_6._m0 = uint4(int4(_23) >> _22);
_6._m0 = uint4(_22 >> int4(_23));
_6._m0 = uint4(int4(_23) >> int4(_23));
_6._m0 = uint4(_22 >> _22);
_6._m1 = int4(_23) >> int4(_23);
_6._m1 = _22 >> _22;
_6._m1 = int4(_23) >> _22;
_6._m1 = _22 >> int4(_23);
}

18
reference/opt/shaders-msl/asm/comp/bitcast_sdiv.asm.comp
View File

@ -17,13 +17,15 @@ struct _4
kernel void main0(device _3& _5 [[buffer(0)]], device _4& _6 [[buffer(1)]])
{
_6._m0 = uint4(int4(_5._m1) / _5._m0);
_6._m0 = uint4(_5._m0 / int4(_5._m1));
_6._m0 = uint4(int4(_5._m1) / int4(_5._m1));
_6._m0 = uint4(_5._m0 / _5._m0);
_6._m1 = int4(_5._m1) / int4(_5._m1);
_6._m1 = _5._m0 / _5._m0;
_6._m1 = int4(_5._m1) / _5._m0;
_6._m1 = _5._m0 / int4(_5._m1);
int4 _22 = _5._m0;
uint4 _23 = _5._m1;
_6._m0 = uint4(int4(_23) / _22);
_6._m0 = uint4(_22 / int4(_23));
_6._m0 = uint4(int4(_23) / int4(_23));
_6._m0 = uint4(_22 / _22);
_6._m1 = int4(_23) / int4(_23);
_6._m1 = _22 / _22;
_6._m1 = int4(_23) / _22;
_6._m1 = _22 / int4(_23);
}

18
reference/opt/shaders-msl/asm/comp/bitcast_slr.asm.comp
View File

@ -17,13 +17,15 @@ struct _4
kernel void main0(device _3& _5 [[buffer(0)]], device _4& _6 [[buffer(1)]])
{
_6._m0 = _5._m1 >> uint4(_5._m0);
_6._m0 = uint4(_5._m0) >> _5._m1;
_6._m0 = _5._m1 >> _5._m1;
_6._m0 = uint4(_5._m0) >> uint4(_5._m0);
_6._m1 = int4(_5._m1 >> _5._m1);
_6._m1 = int4(uint4(_5._m0) >> uint4(_5._m0));
_6._m1 = int4(_5._m1 >> uint4(_5._m0));
_6._m1 = int4(uint4(_5._m0) >> _5._m1);
int4 _22 = _5._m0;
uint4 _23 = _5._m1;
_6._m0 = _23 >> uint4(_22);
_6._m0 = uint4(_22) >> _23;
_6._m0 = _23 >> _23;
_6._m0 = uint4(_22) >> uint4(_22);
_6._m1 = int4(_23 >> _23);
_6._m1 = int4(uint4(_22) >> uint4(_22));
_6._m1 = int4(_23 >> uint4(_22));
_6._m1 = int4(uint4(_22) >> _23);
}

2
reference/shaders-msl/asm/comp/storage-buffer-basic.asm.comp → reference/opt/shaders-msl/asm/comp/storage-buffer-basic.invalid.asm.comp
View File

@ -16,7 +16,7 @@ struct _6
kernel void main0(device _6& _8 [[buffer(0)]], device _6& _9 [[buffer(1)]], uint3 gl_WorkGroupID [[threadgroup_position_in_grid]])
{
uint3 _23 = gl_WorkGroupSize;
_8._m0[gl_WorkGroupID.x] = _9._m0[gl_WorkGroupID.x] + _8._m0[gl_WorkGroupID.x];
uint3 _23 = gl_WorkGroupSize;
}

23
reference/opt/shaders-msl/asm/frag/combined-sampler-reuse.asm.frag Normal file
View File

@ -0,0 +1,23 @@
#include <metal_stdlib>
#include <simd/simd.h>
using namespace metal;
struct main0_out
{
float4 FragColor [[color(0)]];
};
struct main0_in
{
float2 vUV [[user(locn0)]];
};
fragment main0_out main0(main0_in in [[stage_in]], texture2d<float> uTex [[texture(1)]], sampler uSampler [[sampler(0)]])
{
main0_out out = {};
out.FragColor = uTex.sample(uSampler, in.vUV);
out.FragColor += uTex.sample(uSampler, in.vUV, int2(1));
return out;
}

26
reference/opt/shaders-msl/asm/frag/default-member-names.asm.frag
View File

@ -3,29 +3,7 @@
using namespace metal;
struct _9
{
float _m0;
};
struct _10
{
float _m0;
float _m1;
float _m2;
float _m3;
float _m4;
float _m5;
float _m6;
float _m7;
float _m8;
float _m9;
float _m10;
float _m11;
_9 _m12;
};
constant _10 _51 = {};
constant float _57 = {};
struct main0_out
{
@ -35,7 +13,7 @@ struct main0_out
fragment main0_out main0()
{
main0_out out = {};
out.m_3 = float4(_51._m0, _51._m1, _51._m2, _51._m3);
out.m_3 = float4(_57);
return out;
}

2
reference/shaders-msl/asm/vert/empty-struct-composite.asm.vert → reference/opt/shaders-msl/asm/frag/empty-struct.asm.frag
View File

@ -3,7 +3,7 @@
using namespace metal;
vertex void main0()
fragment void main0()
{
}

12
reference/opt/shaders-msl/asm/frag/frem.asm.frag
View File

@ -3,17 +3,17 @@
using namespace metal;
struct main0_in
{
float4 vB [[user(locn1)]];
float4 vA [[user(locn0)]];
};
struct main0_out
{
float4 FragColor [[color(0)]];
};
struct main0_in
{
float4 vA [[user(locn0)]];
float4 vB [[user(locn1)]];
};
fragment main0_out main0(main0_in in [[stage_in]])
{
main0_out out = {};

2
reference/opt/shaders-msl/asm/frag/function-overload-alias.asm.frag
View File

@ -11,7 +11,7 @@ struct main0_out
fragment main0_out main0()
{
main0_out out = {};
out.FragColor = (((float4(1.0) + float4(1.0)) + (float3(1.0).xyzz + float4(1.0))) + (float4(1.0) + float4(2.0))) + (float2(1.0).xyxy + float4(2.0));
out.FragColor = float4(10.0);
return out;
}

17
reference/opt/shaders-msl/asm/frag/image-extract-reuse.asm.frag Normal file
View File

@ -0,0 +1,17 @@
#include <metal_stdlib>
#include <simd/simd.h>
using namespace metal;
struct main0_out
{
int2 Size [[color(0)]];
};
fragment main0_out main0(texture2d<float> uTexture [[texture(0)]], sampler uTextureSmplr [[sampler(0)]])
{
main0_out out = {};
out.Size = int2(uTexture.get_width(), uTexture.get_height()) + int2(uTexture.get_width(1), uTexture.get_height(1));
return out;
}

50
reference/opt/shaders-msl/asm/frag/implicit-read-dep-phi.asm.frag Normal file
View File

@ -0,0 +1,50 @@
#include <metal_stdlib>
#include <simd/simd.h>
using namespace metal;
struct main0_out
{
float4 FragColor [[color(0)]];
};
struct main0_in
{
float4 v0 [[user(locn0)]];
};
fragment main0_out main0(main0_in in [[stage_in]], texture2d<float> uImage [[texture(0)]], sampler uImageSmplr [[sampler(0)]])
{
main0_out out = {};
float phi;
float4 _36;
int _51;
_51 = 0;
phi = 1.0;
_36 = float4(1.0, 2.0, 1.0, 2.0);
for (;;)
{
out.FragColor = _36;
if (_51 < 4)
{
if (in.v0[_51] > 0.0)
{
float2 _48 = float2(phi);
_51++;
phi += 2.0;
_36 = uImage.sample(uImageSmplr, _48, level(0.0));
continue;
}
else
{
break;
}
}
else
{
break;
}
}
return out;
}

6
reference/opt/shaders-msl/asm/frag/op-constant-null.asm.frag
View File

@ -3,12 +3,6 @@
using namespace metal;
struct D
{
float4 a;
float b;
};
struct main0_out
{
float FragColor [[color(0)]];

3
reference/opt/shaders-msl/asm/frag/phi-loop-variable.asm.frag
View File

@ -5,8 +5,5 @@ using namespace metal;
fragment void main0()
{
for (int _22 = 35; _22 >= 0; _22--)
{
}
}

21
reference/shaders-msl/frag/in_block_assign.noopt.frag → reference/opt/shaders-msl/asm/frag/srem.asm.frag
View File

@ -3,28 +3,21 @@
using namespace metal;
struct VOUT
{
float4 a;
};
struct main0_in
{
float4 VOUT_a [[user(locn0)]];
};
struct main0_out
{
float4 FragColor [[color(0)]];
};
struct main0_in
{
int4 vA [[user(locn0)]];
int4 vB [[user(locn1)]];
};
fragment main0_out main0(main0_in in [[stage_in]])
{
main0_out out = {};
VOUT tmp;
tmp.a = in.VOUT_a;
tmp.a += float4(1.0);
out.FragColor = tmp.a;
out.FragColor = float4(in.vA - in.vB * (in.vA / in.vB));
return out;
}

17
reference/opt/shaders-msl/asm/frag/texel-fetch-no-lod.asm.frag Normal file
View File

@ -0,0 +1,17 @@
#include <metal_stdlib>
#include <simd/simd.h>
using namespace metal;
struct main0_out
{
float4 FragColor [[color(0)]];
};
fragment main0_out main0(texture2d<float> uTexture [[texture(0)]], sampler uTextureSmplr [[sampler(0)]], float4 gl_FragCoord [[position]])
{
main0_out out = {};
out.FragColor = uTexture.read(uint2(int2(gl_FragCoord.xy)), 0);
return out;
}

23
reference/opt/shaders-msl/asm/frag/undef-variable-store.asm.frag
View File

@ -3,9 +3,6 @@
using namespace metal;
constant float4 _38 = {};
constant float4 _50 = {};
struct main0_out
{
float4 _entryPointOutput [[color(0)]];
@ -14,25 +11,7 @@ struct main0_out
fragment main0_out main0()
{
main0_out out = {};
float4 _51;
_51 = _50;
float4 _52;
for (;;)
{
if (0.0 != 0.0)
{
_52 = float4(1.0, 0.0, 0.0, 1.0);
break;
}
else
{
_52 = float4(1.0, 1.0, 0.0, 1.0);
break;
}
_52 = _38;
break;
}
out._entryPointOutput = _52;
out._entryPointOutput = float4(1.0, 1.0, 0.0, 1.0);
return out;
}

26
reference/opt/shaders-msl/asm/frag/unreachable.asm.frag
View File

@ -3,36 +3,34 @@
using namespace metal;
constant float4 _21 = {};
struct main0_in
{
int counter [[user(locn0)]];
};
struct main0_out
{
float4 FragColor [[color(0)]];
};
struct main0_in
{
int counter [[user(locn0)]];
};
fragment main0_out main0(main0_in in [[stage_in]])
{
main0_out out = {};
float4 _33;
do
bool _29;
for (;;)
{
if (in.counter == 10)
_29 = in.counter == 10;
if (_29)
{
_33 = float4(10.0);
break;
}
else
{
_33 = float4(30.0);
break;
}
} while (false);
out.FragColor = _33;
}
bool4 _35 = bool4(_29);
out.FragColor = float4(_35.x ? float4(10.0).x : float4(30.0).x, _35.y ? float4(10.0).y : float4(30.0).y, _35.z ? float4(10.0).z : float4(30.0).z, _35.w ? float4(10.0).w : float4(30.0).w);
return out;
}

190
reference/opt/shaders-msl/asm/frag/vector-shuffle-oom.asm.frag
View File

@ -95,212 +95,212 @@ struct main0_out
float4 m_5 [[color(0)]];
};
fragment main0_out main0(constant _6& _7 [[buffer(0)]], constant _10& _11 [[buffer(1)]], constant _18& _19 [[buffer(2)]], texture2d<float> _8 [[texture(0)]], texture2d<float> _12 [[texture(1)]], texture2d<float> _14 [[texture(2)]], sampler _9 [[sampler(0)]], sampler _13 [[sampler(1)]], sampler _15 [[sampler(2)]], float4 gl_FragCoord [[position]])
fragment main0_out main0(constant _6& _7 [[buffer(0)]], constant _18& _19 [[buffer(1)]], constant _10& _11 [[buffer(2)]], texture2d<float> _14 [[texture(4)]], texture2d<float> _12 [[texture(13)]], texture2d<float> _8 [[texture(14)]], sampler _15 [[sampler(3)]], sampler _13 [[sampler(5)]], sampler _9 [[sampler(6)]], float4 gl_FragCoord [[position]])
{
main0_out out = {};
_28 _77 = _74;
_77._m0 = float4(0.0);
float2 _82 = gl_FragCoord.xy * _19._m23.xy;
float4 _88 = _7._m2 * _7._m0.xyxy;
float2 _97 = clamp(_82 + (float3(0.0, -2.0, 0.5).xy * _7._m0.xy), _88.xy, _88.zw);
float3 _109 = _11._m5 * clamp(_8.sample(_9, _97, level(0.0)).w * _7._m1, 0.0, 1.0);
float2 _95 = _88.xy;
float2 _96 = _88.zw;
float2 _97 = clamp(_82 + (float2(0.0, -2.0) * _7._m0.xy), _95, _96);
float3 _109 = float3(_11._m5) * clamp(_8.sample(_9, _97, level(0.0)).w * _7._m1, 0.0, 1.0);
float4 _113 = _12.sample(_13, _97, level(0.0));
float _114 = _113.y;
float3 _129;
if (_113.y > 0.0)
if (_114 > 0.0)
{
_129 = _109 + (_14.sample(_15, _97, level(0.0)).xyz * clamp(_113.y * _113.z, 0.0, 1.0));
_129 = _109 + (_14.sample(_15, _97, level(0.0)).xyz * clamp(_114 * _113.z, 0.0, 1.0));
}
else
{
_129 = _109;
}
float3 _133 = float4(0.0).xyz + (_129 * 0.5);
float4 _134 = float4(_133.x, _133.y, _133.z, float4(0.0).w);
_28 _135 = _77;
_135._m0 = _134;
float2 _144 = clamp(_82 + (float3(-1.0, -1.0, 0.5).xy * _7._m0.xy), _88.xy, _88.zw);
float3 _156 = _11._m5 * clamp(_8.sample(_9, _144, level(0.0)).w * _7._m1, 0.0, 1.0);
float3 _130 = _129 * 0.5;
float4 _134 = float4(_130.x, _130.y, _130.z, float4(0.0).w);
float2 _144 = clamp(_82 + (float2(-1.0) * _7._m0.xy), _95, _96);
float3 _156 = float3(_11._m5) * clamp(_8.sample(_9, _144, level(0.0)).w * _7._m1, 0.0, 1.0);
float4 _160 = _12.sample(_13, _144, level(0.0));
float _161 = _160.y;
float3 _176;
if (_160.y > 0.0)
if (_161 > 0.0)
{
_176 = _156 + (_14.sample(_15, _144, level(0.0)).xyz * clamp(_160.y * _160.z, 0.0, 1.0));
_176 = _156 + (_14.sample(_15, _144, level(0.0)).xyz * clamp(_161 * _160.z, 0.0, 1.0));
}
else
{
_176 = _156;
}
float3 _180 = _134.xyz + (_176 * 0.5);
float3 _177 = _176 * 0.5;
float3 _180 = _134.xyz + _177;
float4 _181 = float4(_180.x, _180.y, _180.z, _134.w);
_28 _182 = _135;
_182._m0 = _181;
float2 _191 = clamp(_82 + (float3(0.0, -1.0, 0.75).xy * _7._m0.xy), _88.xy, _88.zw);
float3 _203 = _11._m5 * clamp(_8.sample(_9, _191, level(0.0)).w * _7._m1, 0.0, 1.0);
float2 _191 = clamp(_82 + (float2(0.0, -1.0) * _7._m0.xy), _95, _96);
float3 _203 = float3(_11._m5) * clamp(_8.sample(_9, _191, level(0.0)).w * _7._m1, 0.0, 1.0);
float4 _207 = _12.sample(_13, _191, level(0.0));
float _208 = _207.y;
float3 _223;
if (_207.y > 0.0)
if (_208 > 0.0)
{
_223 = _203 + (_14.sample(_15, _191, level(0.0)).xyz * clamp(_207.y * _207.z, 0.0, 1.0));
_223 = _203 + (_14.sample(_15, _191, level(0.0)).xyz * clamp(_208 * _207.z, 0.0, 1.0));
}
else
{
_223 = _203;
}
float3 _227 = _181.xyz + (_223 * 0.75);
float3 _224 = _223 * 0.75;
float3 _227 = _181.xyz + _224;
float4 _228 = float4(_227.x, _227.y, _227.z, _181.w);
_28 _229 = _182;
_229._m0 = _228;
float2 _238 = clamp(_82 + (float3(1.0, -1.0, 0.5).xy * _7._m0.xy), _88.xy, _88.zw);
float3 _250 = _11._m5 * clamp(_8.sample(_9, _238, level(0.0)).w * _7._m1, 0.0, 1.0);
float2 _238 = clamp(_82 + (float2(1.0, -1.0) * _7._m0.xy), _95, _96);
float3 _250 = float3(_11._m5) * clamp(_8.sample(_9, _238, level(0.0)).w * _7._m1, 0.0, 1.0);
float4 _254 = _12.sample(_13, _238, level(0.0));
float _255 = _254.y;
float3 _270;
if (_254.y > 0.0)
if (_255 > 0.0)
{
_270 = _250 + (_14.sample(_15, _238, level(0.0)).xyz * clamp(_254.y * _254.z, 0.0, 1.0));
_270 = _250 + (_14.sample(_15, _238, level(0.0)).xyz * clamp(_255 * _254.z, 0.0, 1.0));
}
else
{
_270 = _250;
}
float3 _274 = _228.xyz + (_270 * 0.5);
float3 _271 = _270 * 0.5;
float3 _274 = _228.xyz + _271;
float4 _275 = float4(_274.x, _274.y, _274.z, _228.w);
_28 _276 = _229;
_276._m0 = _275;
float2 _285 = clamp(_82 + (float3(-2.0, 0.0, 0.5).xy * _7._m0.xy), _88.xy, _88.zw);
float3 _297 = _11._m5 * clamp(_8.sample(_9, _285, level(0.0)).w * _7._m1, 0.0, 1.0);
float2 _285 = clamp(_82 + (float2(-2.0, 0.0) * _7._m0.xy), _95, _96);
float3 _297 = float3(_11._m5) * clamp(_8.sample(_9, _285, level(0.0)).w * _7._m1, 0.0, 1.0);
float4 _301 = _12.sample(_13, _285, level(0.0));
float _302 = _301.y;
float3 _317;
if (_301.y > 0.0)
if (_302 > 0.0)
{
_317 = _297 + (_14.sample(_15, _285, level(0.0)).xyz * clamp(_301.y * _301.z, 0.0, 1.0));
_317 = _297 + (_14.sample(_15, _285, level(0.0)).xyz * clamp(_302 * _301.z, 0.0, 1.0));
}
else
{
_317 = _297;
}
float3 _321 = _275.xyz + (_317 * 0.5);
float3 _318 = _317 * 0.5;
float3 _321 = _275.xyz + _318;
float4 _322 = float4(_321.x, _321.y, _321.z, _275.w);
_28 _323 = _276;
_323._m0 = _322;
float2 _332 = clamp(_82 + (float3(-1.0, 0.0, 0.75).xy * _7._m0.xy), _88.xy, _88.zw);
float3 _344 = _11._m5 * clamp(_8.sample(_9, _332, level(0.0)).w * _7._m1, 0.0, 1.0);
float2 _332 = clamp(_82 + (float2(-1.0, 0.0) * _7._m0.xy), _95, _96);
float3 _344 = float3(_11._m5) * clamp(_8.sample(_9, _332, level(0.0)).w * _7._m1, 0.0, 1.0);
float4 _348 = _12.sample(_13, _332, level(0.0));
float _349 = _348.y;
float3 _364;
if (_348.y > 0.0)
if (_349 > 0.0)
{
_364 = _344 + (_14.sample(_15, _332, level(0.0)).xyz * clamp(_348.y * _348.z, 0.0, 1.0));
_364 = _344 + (_14.sample(_15, _332, level(0.0)).xyz * clamp(_349 * _348.z, 0.0, 1.0));
}
else
{
_364 = _344;
}
float3 _368 = _322.xyz + (_364 * 0.75);
float3 _365 = _364 * 0.75;
float3 _368 = _322.xyz + _365;
float4 _369 = float4(_368.x, _368.y, _368.z, _322.w);
_28 _370 = _323;
_370._m0 = _369;
float2 _379 = clamp(_82 + (float3(0.0, 0.0, 1.0).xy * _7._m0.xy), _88.xy, _88.zw);
float3 _391 = _11._m5 * clamp(_8.sample(_9, _379, level(0.0)).w * _7._m1, 0.0, 1.0);
float2 _379 = clamp(_82, _95, _96);
float3 _391 = float3(_11._m5) * clamp(_8.sample(_9, _379, level(0.0)).w * _7._m1, 0.0, 1.0);
float4 _395 = _12.sample(_13, _379, level(0.0));
float _396 = _395.y;
float3 _411;
if (_395.y > 0.0)
if (_396 > 0.0)
{
_411 = _391 + (_14.sample(_15, _379, level(0.0)).xyz * clamp(_395.y * _395.z, 0.0, 1.0));
_411 = _391 + (_14.sample(_15, _379, level(0.0)).xyz * clamp(_396 * _395.z, 0.0, 1.0));
}
else
{
_411 = _391;
}
float3 _415 = _369.xyz + (_411 * 1.0);
float3 _412 = _411 * 1.0;
float3 _415 = _369.xyz + _412;
float4 _416 = float4(_415.x, _415.y, _415.z, _369.w);
_28 _417 = _370;
_417._m0 = _416;
float2 _426 = clamp(_82 + (float3(1.0, 0.0, 0.75).xy * _7._m0.xy), _88.xy, _88.zw);
float3 _438 = _11._m5 * clamp(_8.sample(_9, _426, level(0.0)).w * _7._m1, 0.0, 1.0);
float2 _426 = clamp(_82 + (float2(1.0, 0.0) * _7._m0.xy), _95, _96);
float3 _438 = float3(_11._m5) * clamp(_8.sample(_9, _426, level(0.0)).w * _7._m1, 0.0, 1.0);
float4 _442 = _12.sample(_13, _426, level(0.0));
float _443 = _442.y;
float3 _458;
if (_442.y > 0.0)
if (_443 > 0.0)
{
_458 = _438 + (_14.sample(_15, _426, level(0.0)).xyz * clamp(_442.y * _442.z, 0.0, 1.0));
_458 = _438 + (_14.sample(_15, _426, level(0.0)).xyz * clamp(_443 * _442.z, 0.0, 1.0));
}
else
{
_458 = _438;
}
float3 _462 = _416.xyz + (_458 * 0.75);
float3 _459 = _458 * 0.75;
float3 _462 = _416.xyz + _459;
float4 _463 = float4(_462.x, _462.y, _462.z, _416.w);
_28 _464 = _417;
_464._m0 = _463;
float2 _473 = clamp(_82 + (float3(2.0, 0.0, 0.5).xy * _7._m0.xy), _88.xy, _88.zw);
float3 _485 = _11._m5 * clamp(_8.sample(_9, _473, level(0.0)).w * _7._m1, 0.0, 1.0);
float2 _473 = clamp(_82 + (float2(2.0, 0.0) * _7._m0.xy), _95, _96);
float3 _485 = float3(_11._m5) * clamp(_8.sample(_9, _473, level(0.0)).w * _7._m1, 0.0, 1.0);
float4 _489 = _12.sample(_13, _473, level(0.0));
float _490 = _489.y;
float3 _505;
if (_489.y > 0.0)
if (_490 > 0.0)
{
_505 = _485 + (_14.sample(_15, _473, level(0.0)).xyz * clamp(_489.y * _489.z, 0.0, 1.0));
_505 = _485 + (_14.sample(_15, _473, level(0.0)).xyz * clamp(_490 * _489.z, 0.0, 1.0));
}
else
{
_505 = _485;
}
float3 _509 = _463.xyz + (_505 * 0.5);
float3 _506 = _505 * 0.5;
float3 _509 = _463.xyz + _506;
float4 _510 = float4(_509.x, _509.y, _509.z, _463.w);
_28 _511 = _464;
_511._m0 = _510;
float2 _520 = clamp(_82 + (float3(-1.0, 1.0, 0.5).xy * _7._m0.xy), _88.xy, _88.zw);
float3 _532 = _11._m5 * clamp(_8.sample(_9, _520, level(0.0)).w * _7._m1, 0.0, 1.0);
float2 _520 = clamp(_82 + (float2(-1.0, 1.0) * _7._m0.xy), _95, _96);
float3 _532 = float3(_11._m5) * clamp(_8.sample(_9, _520, level(0.0)).w * _7._m1, 0.0, 1.0);
float4 _536 = _12.sample(_13, _520, level(0.0));
float _537 = _536.y;
float3 _552;
if (_536.y > 0.0)
if (_537 > 0.0)
{
_552 = _532 + (_14.sample(_15, _520, level(0.0)).xyz * clamp(_536.y * _536.z, 0.0, 1.0));
_552 = _532 + (_14.sample(_15, _520, level(0.0)).xyz * clamp(_537 * _536.z, 0.0, 1.0));
}
else
{
_552 = _532;
}
float3 _556 = _510.xyz + (_552 * 0.5);
float3 _553 = _552 * 0.5;
float3 _556 = _510.xyz + _553;
float4 _557 = float4(_556.x, _556.y, _556.z, _510.w);
_28 _558 = _511;
_558._m0 = _557;
float2 _567 = clamp(_82 + (float3(0.0, 1.0, 0.75).xy * _7._m0.xy), _88.xy, _88.zw);
float3 _579 = _11._m5 * clamp(_8.sample(_9, _567, level(0.0)).w * _7._m1, 0.0, 1.0);
float2 _567 = clamp(_82 + (float2(0.0, 1.0) * _7._m0.xy), _95, _96);
float3 _579 = float3(_11._m5) * clamp(_8.sample(_9, _567, level(0.0)).w * _7._m1, 0.0, 1.0);
float4 _583 = _12.sample(_13, _567, level(0.0));
float _584 = _583.y;
float3 _599;
if (_583.y > 0.0)
if (_584 > 0.0)
{
_599 = _579 + (_14.sample(_15, _567, level(0.0)).xyz * clamp(_583.y * _583.z, 0.0, 1.0));
_599 = _579 + (_14.sample(_15, _567, level(0.0)).xyz * clamp(_584 * _583.z, 0.0, 1.0));
}
else
{
_599 = _579;
}
float3 _603 = _557.xyz + (_599 * 0.75);
float3 _600 = _599 * 0.75;
float3 _603 = _557.xyz + _600;
float4 _604 = float4(_603.x, _603.y, _603.z, _557.w);
_28 _605 = _558;
_605._m0 = _604;
float2 _614 = clamp(_82 + (float3(1.0, 1.0, 0.5).xy * _7._m0.xy), _88.xy, _88.zw);
float3 _626 = _11._m5 * clamp(_8.sample(_9, _614, level(0.0)).w * _7._m1, 0.0, 1.0);
float2 _614 = clamp(_82 + _7._m0.xy, _95, _96);
float3 _626 = float3(_11._m5) * clamp(_8.sample(_9, _614, level(0.0)).w * _7._m1, 0.0, 1.0);
float4 _630 = _12.sample(_13, _614, level(0.0));
float _631 = _630.y;
float3 _646;
if (_630.y > 0.0)
if (_631 > 0.0)
{
_646 = _626 + (_14.sample(_15, _614, level(0.0)).xyz * clamp(_630.y * _630.z, 0.0, 1.0));
_646 = _626 + (_14.sample(_15, _614, level(0.0)).xyz * clamp(_631 * _630.z, 0.0, 1.0));
}
else
{
_646 = _626;
}
float3 _650 = _604.xyz + (_646 * 0.5);
float3 _647 = _646 * 0.5;
float3 _650 = _604.xyz + _647;
float4 _651 = float4(_650.x, _650.y, _650.z, _604.w);
_28 _652 = _605;
_652._m0 = _651;
float2 _661 = clamp(_82 + (float3(0.0, 2.0, 0.5).xy * _7._m0.xy), _88.xy, _88.zw);
float3 _673 = _11._m5 * clamp(_8.sample(_9, _661, level(0.0)).w * _7._m1, 0.0, 1.0);
float2 _661 = clamp(_82 + (float2(0.0, 2.0) * _7._m0.xy), _95, _96);
float3 _673 = float3(_11._m5) * clamp(_8.sample(_9, _661, level(0.0)).w * _7._m1, 0.0, 1.0);
float4 _677 = _12.sample(_13, _661, level(0.0));
float _678 = _677.y;
float3 _693;
if (_677.y > 0.0)
if (_678 > 0.0)
{
_693 = _673 + (_14.sample(_15, _661, level(0.0)).xyz * clamp(_677.y * _677.z, 0.0, 1.0));
_693 = _673 + (_14.sample(_15, _661, level(0.0)).xyz * clamp(_678 * _677.z, 0.0, 1.0));
}
else
{
@ -308,10 +308,8 @@ fragment main0_out main0(constant _6& _7 [[buffer(0)]], constant _10& _11 [[buff
}
float3 _697 = _651.xyz + (_693 * 0.5);
float4 _698 = float4(_697.x, _697.y, _697.z, _651.w);
_28 _699 = _652;
_699._m0 = _698;
float3 _702 = _698.xyz / float3(((((((((((((0.0 + 0.5) + 0.5) + 0.75) + 0.5) + 0.5) + 0.75) + 1.0) + 0.75) + 0.5) + 0.5) + 0.75) + 0.5) + 0.5);
_28 _704 = _699;
float3 _702 = _698.xyz * float3(0.125);
_28 _704 = _74;
_704._m0 = float4(_702.x, _702.y, _702.z, _698.w);
_28 _705 = _704;
_705._m0.w = 1.0;

0
reference/opt/shaders-msl/asm/vert/empty-struct-composite.asm.vert → reference/opt/shaders-msl/asm/vert/packing-test.asm.vert
View File

35
reference/opt/shaders-msl/asm/vert/spec-constant-op-composite.asm.vert Normal file
View File

@ -0,0 +1,35 @@
#include <metal_stdlib>
#include <simd/simd.h>
using namespace metal;
constant int _7_tmp [[function_constant(201)]];
constant int _7 = is_function_constant_defined(_7_tmp) ? _7_tmp : -10;
constant uint _8_tmp [[function_constant(202)]];
constant uint _8 = is_function_constant_defined(_8_tmp) ? _8_tmp : 100u;
constant int _20 = (_7 + 2);
constant uint _25 = (_8 % 5u);
constant int4 _30 = int4(20, 30, _20, _20);
constant int2 _32 = int2(_30.y, _30.x);
constant int _33 = _30.y;
struct main0_out
{
int m_4 [[user(locn0)]];
float4 gl_Position [[position]];
};
vertex main0_out main0()
{
main0_out out = {};
float4 _64 = float4(0.0);
_64.y = float(_20);
float4 _68 = _64;
_68.z = float(_25);
float4 _52 = _68 + float4(_30);
float2 _56 = _52.xy + float2(_32);
out.gl_Position = float4(_56.x, _56.y, _52.z, _52.w);
out.m_4 = _33;
return out;
}

18
reference/opt/shaders-msl/comp/access-private-workgroup-in-function.comp Normal file
View File

@ -0,0 +1,18 @@
#include <metal_stdlib>
#include <simd/simd.h>
using namespace metal;
kernel void main0(uint gl_LocalInvocationIndex [[thread_index_in_threadgroup]])
{
threadgroup int u;
u = 50;
if (gl_LocalInvocationIndex == 0u)
{
}
else
{
u = 20;
}
}

72
reference/opt/shaders-msl/comp/atomic.comp
View File

@ -12,25 +12,59 @@ struct SSBO
int i32;
};
kernel void main0(device SSBO& ssbo [[buffer(0)]])
kernel void main0(device SSBO& ssbo [[buffer(2)]])
{
uint _16 = atomic_fetch_add_explicit((volatile device atomic_uint*)&(ssbo.u32), 1u, memory_order_relaxed);
uint _18 = atomic_fetch_or_explicit((volatile device atomic_uint*)&(ssbo.u32), 1u, memory_order_relaxed);
uint _20 = atomic_fetch_xor_explicit((volatile device atomic_uint*)&(ssbo.u32), 1u, memory_order_relaxed);
uint _22 = atomic_fetch_and_explicit((volatile device atomic_uint*)&(ssbo.u32), 1u, memory_order_relaxed);
uint _24 = atomic_fetch_min_explicit((volatile device atomic_uint*)&(ssbo.u32), 1u, memory_order_relaxed);
uint _26 = atomic_fetch_max_explicit((volatile device atomic_uint*)&(ssbo.u32), 1u, memory_order_relaxed);
uint _28 = atomic_exchange_explicit((volatile device atomic_uint*)&(ssbo.u32), 1u, memory_order_relaxed);
uint _30 = 10u;
uint _32 = atomic_compare_exchange_weak_explicit((volatile device atomic_uint*)&(ssbo.u32), &(_30), 2u, memory_order_relaxed, memory_order_relaxed);
int _36 = atomic_fetch_add_explicit((volatile device atomic_int*)&(ssbo.i32), 1, memory_order_relaxed);
int _38 = atomic_fetch_or_explicit((volatile device atomic_int*)&(ssbo.i32), 1, memory_order_relaxed);
int _40 = atomic_fetch_xor_explicit((volatile device atomic_int*)&(ssbo.i32), 1, memory_order_relaxed);
int _42 = atomic_fetch_and_explicit((volatile device atomic_int*)&(ssbo.i32), 1, memory_order_relaxed);
int _44 = atomic_fetch_min_explicit((volatile device atomic_int*)&(ssbo.i32), 1, memory_order_relaxed);
int _46 = atomic_fetch_max_explicit((volatile device atomic_int*)&(ssbo.i32), 1, memory_order_relaxed);
int _48 = atomic_exchange_explicit((volatile device atomic_int*)&(ssbo.i32), 1, memory_order_relaxed);
int _50 = 10;
int _52 = atomic_compare_exchange_weak_explicit((volatile device atomic_int*)&(ssbo.i32), &(_50), 2, memory_order_relaxed, memory_order_relaxed);
threadgroup uint shared_u32;
threadgroup int shared_i32;
uint _16 = atomic_fetch_add_explicit((volatile device atomic_uint*)&ssbo.u32, 1u, memory_order_relaxed);
uint _18 = atomic_fetch_or_explicit((volatile device atomic_uint*)&ssbo.u32, 1u, memory_order_relaxed);
uint _20 = atomic_fetch_xor_explicit((volatile device atomic_uint*)&ssbo.u32, 1u, memory_order_relaxed);
uint _22 = atomic_fetch_and_explicit((volatile device atomic_uint*)&ssbo.u32, 1u, memory_order_relaxed);
uint _24 = atomic_fetch_min_explicit((volatile device atomic_uint*)&ssbo.u32, 1u, memory_order_relaxed);
uint _26 = atomic_fetch_max_explicit((volatile device atomic_uint*)&ssbo.u32, 1u, memory_order_relaxed);
uint _28 = atomic_exchange_explicit((volatile device atomic_uint*)&ssbo.u32, 1u, memory_order_relaxed);
uint _32;
do
{
_32 = 10u;
} while (!atomic_compare_exchange_weak_explicit((volatile device atomic_uint*)&ssbo.u32, &_32, 2u, memory_order_relaxed, memory_order_relaxed));
int _36 = atomic_fetch_add_explicit((volatile device atomic_int*)&ssbo.i32, 1, memory_order_relaxed);
int _38 = atomic_fetch_or_explicit((volatile device atomic_int*)&ssbo.i32, 1, memory_order_relaxed);
int _40 = atomic_fetch_xor_explicit((volatile device atomic_int*)&ssbo.i32, 1, memory_order_relaxed);
int _42 = atomic_fetch_and_explicit((volatile device atomic_int*)&ssbo.i32, 1, memory_order_relaxed);
int _44 = atomic_fetch_min_explicit((volatile device atomic_int*)&ssbo.i32, 1, memory_order_relaxed);
int _46 = atomic_fetch_max_explicit((volatile device atomic_int*)&ssbo.i32, 1, memory_order_relaxed);
int _48 = atomic_exchange_explicit((volatile device atomic_int*)&ssbo.i32, 1, memory_order_relaxed);
int _52;
do
{
_52 = 10;
} while (!atomic_compare_exchange_weak_explicit((volatile device atomic_int*)&ssbo.i32, &_52, 2, memory_order_relaxed, memory_order_relaxed));
shared_u32 = 10u;
shared_i32 = 10;
uint _57 = atomic_fetch_add_explicit((volatile threadgroup atomic_uint*)&shared_u32, 1u, memory_order_relaxed);
uint _58 = atomic_fetch_or_explicit((volatile threadgroup atomic_uint*)&shared_u32, 1u, memory_order_relaxed);
uint _59 = atomic_fetch_xor_explicit((volatile threadgroup atomic_uint*)&shared_u32, 1u, memory_order_relaxed);
uint _60 = atomic_fetch_and_explicit((volatile threadgroup atomic_uint*)&shared_u32, 1u, memory_order_relaxed);
uint _61 = atomic_fetch_min_explicit((volatile threadgroup atomic_uint*)&shared_u32, 1u, memory_order_relaxed);
uint _62 = atomic_fetch_max_explicit((volatile threadgroup atomic_uint*)&shared_u32, 1u, memory_order_relaxed);
uint _63 = atomic_exchange_explicit((volatile threadgroup atomic_uint*)&shared_u32, 1u, memory_order_relaxed);
uint _64;
do
{
_64 = 10u;
} while (!atomic_compare_exchange_weak_explicit((volatile threadgroup atomic_uint*)&shared_u32, &_64, 2u, memory_order_relaxed, memory_order_relaxed));
int _65 = atomic_fetch_add_explicit((volatile threadgroup atomic_int*)&shared_i32, 1, memory_order_relaxed);
int _66 = atomic_fetch_or_explicit((volatile threadgroup atomic_int*)&shared_i32, 1, memory_order_relaxed);
int _67 = atomic_fetch_xor_explicit((volatile threadgroup atomic_int*)&shared_i32, 1, memory_order_relaxed);
int _68 = atomic_fetch_and_explicit((volatile threadgroup atomic_int*)&shared_i32, 1, memory_order_relaxed);
int _69 = atomic_fetch_min_explicit((volatile threadgroup atomic_int*)&shared_i32, 1, memory_order_relaxed);
int _70 = atomic_fetch_max_explicit((volatile threadgroup atomic_int*)&shared_i32, 1, memory_order_relaxed);
int _71 = atomic_exchange_explicit((volatile threadgroup atomic_int*)&shared_i32, 1, memory_order_relaxed);
int _72;
do
{
_72 = 10;
} while (!atomic_compare_exchange_weak_explicit((volatile threadgroup atomic_int*)&shared_i32, &_72, 2, memory_order_relaxed, memory_order_relaxed));
}

22
reference/opt/shaders-msl/comp/bake_gradient.comp
View File

@ -1,22 +0,0 @@
#include <metal_stdlib>
#include <simd/simd.h>
using namespace metal;
constant uint3 gl_WorkGroupSize = uint3(8u, 8u, 1u);
struct UBO
{
float4 uInvSize;
float4 uScale;
};
kernel void main0(uint3 gl_GlobalInvocationID [[thread_position_in_grid]], constant UBO& _46 [[buffer(0)]], texture2d<float> uHeight [[texture(0)]], sampler uHeightSmplr [[sampler(0)]], texture2d<float> uDisplacement [[texture(1)]], sampler uDisplacementSmplr [[sampler(1)]], texture2d<float, access::write> iHeightDisplacement [[texture(2)]], texture2d<float, access::write> iGradJacobian [[texture(3)]])
{
float4 _59 = (float2(gl_GlobalInvocationID.xy) * _46.uInvSize.xy).xyxy + (_46.uInvSize * 0.5);
float2 _157 = ((uDisplacement.sample(uDisplacementSmplr, _59.zw, level(0.0), int2(1, 0)).xy - uDisplacement.sample(uDisplacementSmplr, _59.zw, level(0.0), int2(-1, 0)).xy) * 0.60000002384185791015625) * _46.uScale.z;
float2 _161 = ((uDisplacement.sample(uDisplacementSmplr, _59.zw, level(0.0), int2(0, 1)).xy - uDisplacement.sample(uDisplacementSmplr, _59.zw, level(0.0), int2(0, -1)).xy) * 0.60000002384185791015625) * _46.uScale.z;
iHeightDisplacement.write(float4(uHeight.sample(uHeightSmplr, _59.xy, level(0.0)).x, 0.0, 0.0, 0.0), uint2(int2(gl_GlobalInvocationID.xy)));
iGradJacobian.write(float4((_46.uScale.xy * 0.5) * float2(uHeight.sample(uHeightSmplr, _59.xy, level(0.0), int2(1, 0)).x - uHeight.sample(uHeightSmplr, _59.xy, level(0.0), int2(-1, 0)).x, uHeight.sample(uHeightSmplr, _59.xy, level(0.0), int2(0, 1)).x - uHeight.sample(uHeightSmplr, _59.xy, level(0.0), int2(0, -1)).x), ((1.0 + _157.x) * (1.0 + _161.y)) - (_157.y * _161.x), 0.0), uint2(int2(gl_GlobalInvocationID.xy)));
}

4
reference/opt/shaders-msl/comp/basic.comp
View File

@ -21,12 +21,12 @@ struct SSBO3
uint counter;
};
kernel void main0(device SSBO& _23 [[buffer(0)]], device SSBO2& _45 [[buffer(1)]], device SSBO3& _48 [[buffer(2)]], uint3 gl_GlobalInvocationID [[thread_position_in_grid]])
kernel void main0(const device SSBO& _23 [[buffer(0)]], device SSBO2& _45 [[buffer(1)]], device SSBO3& _48 [[buffer(2)]], uint3 gl_GlobalInvocationID [[thread_position_in_grid]])
{
float4 _29 = _23.in_data[gl_GlobalInvocationID.x];
if (dot(_29, float4(1.0, 5.0, 6.0, 2.0)) > 8.19999980926513671875)
{
uint _52 = atomic_fetch_add_explicit((volatile device atomic_uint*)&(_48.counter), 1u, memory_order_relaxed);
uint _52 = atomic_fetch_add_explicit((volatile device atomic_uint*)&_48.counter, 1u, memory_order_relaxed);
_45.out_data[_52] = _29;
}
}

2
reference/opt/shaders-msl/comp/builtins.comp
View File

@ -3,6 +3,8 @@
using namespace metal;
constant uint3 gl_WorkGroupSize = uint3(8u, 4u, 2u);
kernel void main0(uint3 gl_LocalInvocationID [[thread_position_in_threadgroup]], uint3 gl_GlobalInvocationID [[thread_position_in_grid]], uint gl_LocalInvocationIndex [[thread_index_in_threadgroup]], uint3 gl_NumWorkGroups [[threadgroups_per_grid]], uint3 gl_WorkGroupID [[threadgroup_position_in_grid]])
{
}

2
reference/opt/shaders-msl/comp/coherent-block.comp
View File

@ -8,7 +8,7 @@ struct SSBO
float4 value;
};
kernel void main0(device SSBO& _10 [[buffer(0)]])
kernel void main0(device SSBO& _10 [[buffer(1)]])
{
_10.value = float4(20.0);
}

2
reference/opt/shaders-msl/comp/coherent-image.comp
View File

@ -8,7 +8,7 @@ struct SSBO
int4 value;
};
kernel void main0(device SSBO& _10 [[buffer(0)]], texture2d<int> uImage [[texture(0)]])
kernel void main0(device SSBO& _10 [[buffer(1)]], texture2d<int> uImage [[texture(3)]])
{
_10.value = uImage.read(uint2(int2(10)));
}

35
reference/opt/shaders-msl/comp/composite-construct.comp Normal file
View File

@ -0,0 +1,35 @@
#pragma clang diagnostic ignored "-Wmissing-prototypes"
#include <metal_stdlib>
#include <simd/simd.h>
using namespace metal;
struct SSBO0
{
float4 as[1];
};
// Implementation of an array copy function to cover GLSL's ability to copy an array via assignment.
template<typename T, uint N>
void spvArrayCopy(thread T (&dst)[N], thread const T (&src)[N])
{
for (uint i = 0; i < N; dst[i] = src[i], i++);
}
// An overload for constant arrays.
template<typename T, uint N>
void spvArrayCopyConstant(thread T (&dst)[N], constant T (&src)[N])
{
for (uint i = 0; i < N; dst[i] = src[i], i++);
}
kernel void main0(device SSBO0& _16 [[buffer(0)]], device SSBO0& _32 [[buffer(1)]], uint3 gl_GlobalInvocationID [[thread_position_in_grid]], uint gl_LocalInvocationIndex [[thread_index_in_threadgroup]])
{
float4 _37[2] = { _16.as[gl_GlobalInvocationID.x], _32.as[gl_GlobalInvocationID.x] };
float4 values[2];
spvArrayCopy(values, _37);
_16.as[0] = values[gl_LocalInvocationIndex];
_32.as[1] = float4(40.0);
}

4
reference/opt/shaders-msl/comp/culling.comp
View File

@ -23,12 +23,12 @@ struct SSBO3
uint count;
};
kernel void main0(device SSBO& _22 [[buffer(0)]], device SSBO2& _38 [[buffer(1)]], device SSBO3& _41 [[buffer(2)]], uint3 gl_GlobalInvocationID [[thread_position_in_grid]])
kernel void main0(const device SSBO& _22 [[buffer(0)]], device SSBO2& _38 [[buffer(1)]], device SSBO3& _41 [[buffer(2)]], uint3 gl_GlobalInvocationID [[thread_position_in_grid]])
{
float _28 = _22.in_data[gl_GlobalInvocationID.x];
if (_28 > 12.0)
{
uint _45 = atomic_fetch_add_explicit((volatile device atomic_uint*)&(_41.count), 1u, memory_order_relaxed);
uint _45 = atomic_fetch_add_explicit((volatile device atomic_uint*)&_41.count, 1u, memory_order_relaxed);
_38.out_data[_45] = _28;
}
}

7
reference/opt/shaders-msl/comp/defer-parens.comp
View File

@ -12,10 +12,11 @@ struct SSBO
kernel void main0(device SSBO& _13 [[buffer(0)]])
{
float4 _17 = _13.data;
_13.data = float4(_17.x, _17.yz + float2(10.0), _17.w);
float2 _28 = _17.yz + float2(10.0);
_13.data = float4(_17.x, _28, _17.w);
_13.data = (_17 + _17) + _17;
_13.data = (_17.yz + float2(10.0)).xxyy;
_13.data = float4((_17.yz + float2(10.0)).y);
_13.data = _28.xxyy;
_13.data = float4(_28.y);
_13.data = float4((_17.zw + float2(10.0))[_13.index]);
}

3
reference/opt/shaders-msl/comp/dowhile.comp
View File

@ -14,7 +14,7 @@ struct SSBO2
float4 out_data[1];
};
kernel void main0(device SSBO& _28 [[buffer(0)]], device SSBO2& _52 [[buffer(1)]], uint3 gl_GlobalInvocationID [[thread_position_in_grid]])
kernel void main0(const device SSBO& _28 [[buffer(0)]], device SSBO2& _52 [[buffer(1)]], uint3 gl_GlobalInvocationID [[thread_position_in_grid]])
{
int i = 0;
float4 _56;
@ -27,7 +27,6 @@ kernel void main0(device SSBO& _28 [[buffer(0)]], device SSBO2& _52 [[buffer(1)]
if (i < 16)
{
_56 = _42;
continue;
}
else
{

10
reference/opt/shaders-msl/comp/image-cube-array-load-store.comp Normal file
View File

@ -0,0 +1,10 @@
#include <metal_stdlib>
#include <simd/simd.h>
using namespace metal;
kernel void main0(texturecube_array<float> uImageIn [[texture(0)]], texturecube_array<float, access::write> uImageOut [[texture(1)]])
{
uImageOut.write(uImageIn.read(uint2(int3(9, 7, 11).xy), uint(int3(9, 7, 11).z) % 6u, uint(int3(9, 7, 11).z) / 6u), uint2(int3(9, 7, 11).xy), uint(int3(9, 7, 11).z) % 6u, uint(int3(9, 7, 11).z) / 6u);
}

3
reference/opt/shaders-msl/comp/image.comp
View File

@ -5,6 +5,7 @@ using namespace metal;
kernel void main0(texture2d<float> uImageIn [[texture(0)]], texture2d<float, access::write> uImageOut [[texture(1)]], uint3 gl_GlobalInvocationID [[thread_position_in_grid]])
{
uImageOut.write(uImageIn.read(uint2((int2(gl_GlobalInvocationID.xy) + int2(uImageIn.get_width(), uImageIn.get_height())))), uint2(int2(gl_GlobalInvocationID.xy)));
int2 _23 = int2(gl_GlobalInvocationID.xy);
uImageOut.write(uImageIn.read(uint2((_23 + int2(uImageIn.get_width(), uImageIn.get_height())))), uint2(_23));
}

2
reference/opt/shaders-msl/comp/inverse.comp
View File

@ -114,7 +114,7 @@ float2x2 spvInverse2x2(float2x2 m)
return (det != 0.0f) ? (adj * (1.0f / det)) : m;
}
kernel void main0(device MatrixOut& _15 [[buffer(0)]], device MatrixIn& _20 [[buffer(1)]])
kernel void main0(device MatrixOut& _15 [[buffer(0)]], const device MatrixIn& _20 [[buffer(1)]])
{
_15.m2out = spvInverse2x2(_20.m2in);
_15.m3out = spvInverse3x3(_20.m3in);

2
reference/opt/shaders-msl/comp/mat3.comp
View File

@ -8,7 +8,7 @@ struct SSBO2
float3x3 out_data[1];
};
kernel void main0(device SSBO2& _22 [[buffer(0)]], uint3 gl_GlobalInvocationID [[thread_position_in_grid]])
kernel void main0(device SSBO2& _22 [[buffer(1)]], uint3 gl_GlobalInvocationID [[thread_position_in_grid]])
{
_22.out_data[gl_GlobalInvocationID.x] = float3x3(float3(10.0), float3(20.0), float3(40.0));
}

2
reference/opt/shaders-msl/comp/mod.comp
View File

@ -22,7 +22,7 @@ Tx mod(Tx x, Ty y)
return x - y * floor(x / y);
}
kernel void main0(device SSBO& _23 [[buffer(0)]], device SSBO2& _33 [[buffer(1)]], uint3 gl_GlobalInvocationID [[thread_position_in_grid]])
kernel void main0(const device SSBO& _23 [[buffer(0)]], device SSBO2& _33 [[buffer(1)]], uint3 gl_GlobalInvocationID [[thread_position_in_grid]])
{
_33.out_data[gl_GlobalInvocationID.x] = mod(_23.in_data[gl_GlobalInvocationID.x], _33.out_data[gl_GlobalInvocationID.x]);
_33.out_data[gl_GlobalInvocationID.x] = as_type<float4>(as_type<uint4>(_23.in_data[gl_GlobalInvocationID.x]) % as_type<uint4>(_33.out_data[gl_GlobalInvocationID.x]));

2
reference/opt/shaders-msl/comp/modf.comp
View File

@ -13,7 +13,7 @@ struct SSBO2
float4 out_data[1];
};
kernel void main0(device SSBO& _23 [[buffer(0)]], device SSBO2& _35 [[buffer(1)]], uint3 gl_GlobalInvocationID [[thread_position_in_grid]])
kernel void main0(const device SSBO& _23 [[buffer(0)]], device SSBO2& _35 [[buffer(1)]], uint3 gl_GlobalInvocationID [[thread_position_in_grid]])
{
float4 i;
float4 _31 = modf(_23.in_data[gl_GlobalInvocationID.x], i);

28
reference/opt/shaders-msl/comp/packing-test-1.comp Normal file
View File

@ -0,0 +1,28 @@
#include <metal_stdlib>
#include <simd/simd.h>
using namespace metal;
constant uint3 gl_WorkGroupSize = uint3(32u, 1u, 1u);
struct T1
{
packed_float3 a;
float b;
};
struct Buffer0
{
T1 buf0[1];
};
struct Buffer1
{
float buf1[1];
};
kernel void main0(device Buffer0& _15 [[buffer(1)]], device Buffer1& _34 [[buffer(2)]], uint3 gl_GlobalInvocationID [[thread_position_in_grid]])
{
_34.buf1[gl_GlobalInvocationID.x] = _15.buf0[0].b;
}

28
reference/opt/shaders-msl/comp/packing-test-2.comp Normal file
View File

@ -0,0 +1,28 @@
#include <metal_stdlib>
#include <simd/simd.h>
using namespace metal;
constant uint3 gl_WorkGroupSize = uint3(32u, 1u, 1u);
struct T1
{
packed_float3 a;
float b;
};
struct Buffer0
{
T1 buf0[1];
};
struct Buffer1
{
float buf1[1];
};
kernel void main0(device Buffer0& _14 [[buffer(1)]], device Buffer1& _24 [[buffer(2)]], uint3 gl_GlobalInvocationID [[thread_position_in_grid]])
{
_24.buf1[gl_GlobalInvocationID.x] = _14.buf0[0].b;
}

2
reference/opt/shaders-msl/comp/read-write-only.comp
View File

@ -21,7 +21,7 @@ struct SSBO1
float4 data3;
};
kernel void main0(device SSBO2& _10 [[buffer(0)]], device SSBO0& _15 [[buffer(1)]], device SSBO1& _21 [[buffer(2)]])
kernel void main0(const device SSBO0& _15 [[buffer(0)]], device SSBO1& _21 [[buffer(1)]], device SSBO2& _10 [[buffer(2)]])
{
_10.data4 = _15.data0 + _21.data2;
_10.data5 = _15.data1 + _21.data3;

33
reference/opt/shaders-msl/comp/return.comp
View File

@ -1,33 +0,0 @@
#include <metal_stdlib>
#include <simd/simd.h>
using namespace metal;
struct SSBO2
{
float4 out_data[1];
};
constant int _69 = {};
kernel void main0(device SSBO2& _27 [[buffer(0)]], uint3 gl_GlobalInvocationID [[thread_position_in_grid]])
{
if (gl_GlobalInvocationID.x == 2u)
{
_27.out_data[gl_GlobalInvocationID.x] = float4(20.0);
}
else
{
if (gl_GlobalInvocationID.x == 4u)
{
_27.out_data[gl_GlobalInvocationID.x] = float4(10.0);
return;
}
}
for (int _68 = 0; _68 < 20; _68 = _69 + 1)
{
return;
}
_27.out_data[gl_GlobalInvocationID.x] = float4(10.0);
}

22
reference/opt/shaders-msl/comp/rmw-matrix.comp Normal file
View File

@ -0,0 +1,22 @@
#include <metal_stdlib>
#include <simd/simd.h>
using namespace metal;
struct SSBO
{
float a;
float4 b;
float4x4 c;
float a1;
float4 b1;
float4x4 c1;
};
kernel void main0(device SSBO& _11 [[buffer(0)]])
{
_11.a *= _11.a1;
_11.b *= _11.b1;
_11.c = _11.c * _11.c1;
}

3
reference/opt/shaders-msl/comp/rmw-opt.comp
View File

@ -20,7 +20,6 @@ kernel void main0(device SSBO& _9 [[buffer(0)]])
_9.a ^= 10;
_9.a %= 40;
_9.a |= 1;
bool _65 = false && true;
_9.a = int(_65 && (true || _65));
_9.a = 0;
}

2
reference/opt/shaders-msl/comp/shared-array-of-arrays.comp
View File

@ -15,6 +15,6 @@ kernel void main0(device SSBO& _67 [[buffer(0)]], uint3 gl_LocalInvocationID [[t
threadgroup float foo[4][4];
foo[gl_LocalInvocationID.x][gl_LocalInvocationID.y] = float(gl_LocalInvocationIndex);
threadgroup_barrier(mem_flags::mem_threadgroup);
_67.out_data[gl_GlobalInvocationID.x] = (((0.0 + foo[gl_LocalInvocationID.x][0]) + foo[gl_LocalInvocationID.x][1]) + foo[gl_LocalInvocationID.x][2]) + foo[gl_LocalInvocationID.x][3];
_67.out_data[gl_GlobalInvocationID.x] = ((foo[gl_LocalInvocationID.x][0] + foo[gl_LocalInvocationID.x][1]) + foo[gl_LocalInvocationID.x][2]) + foo[gl_LocalInvocationID.x][3];
}

4
reference/opt/shaders-msl/comp/shared.comp
View File

@ -15,11 +15,11 @@ struct SSBO2
float out_data[1];
};
kernel void main0(device SSBO& _22 [[buffer(0)]], device SSBO2& _44 [[buffer(1)]], uint3 gl_GlobalInvocationID [[thread_position_in_grid]], uint gl_LocalInvocationIndex [[thread_index_in_threadgroup]])
kernel void main0(const device SSBO& _22 [[buffer(0)]], device SSBO2& _44 [[buffer(1)]], uint3 gl_GlobalInvocationID [[thread_position_in_grid]], uint gl_LocalInvocationIndex [[thread_index_in_threadgroup]])
{
threadgroup float sShared[4];
sShared[gl_LocalInvocationIndex] = _22.in_data[gl_GlobalInvocationID.x];
threadgroup_barrier(mem_flags::mem_threadgroup);
_44.out_data[gl_GlobalInvocationID.x] = sShared[(4u - gl_LocalInvocationIndex) - 1u];
_44.out_data[gl_GlobalInvocationID.x] = sShared[3u - gl_LocalInvocationIndex];
}

2
reference/opt/shaders-msl/comp/struct-layout.comp
View File

@ -18,7 +18,7 @@ struct SSBO
Foo in_data[1];
};
kernel void main0(device SSBO2& _23 [[buffer(0)]], device SSBO& _30 [[buffer(1)]], uint3 gl_GlobalInvocationID [[thread_position_in_grid]])
kernel void main0(const device SSBO& _30 [[buffer(0)]], device SSBO2& _23 [[buffer(1)]], uint3 gl_GlobalInvocationID [[thread_position_in_grid]])
{
_23.out_data[gl_GlobalInvocationID.x].m = _30.in_data[gl_GlobalInvocationID.x].m * _30.in_data[gl_GlobalInvocationID.x].m;
}

8
reference/opt/shaders-msl/comp/struct-nested.comp
View File

@ -18,12 +18,8 @@ struct dstbuffer
s2 test[1];
};
constant s2 _31 = {};
kernel void main0(device dstbuffer& _19 [[buffer(0)]])
kernel void main0(device dstbuffer& _19 [[buffer(1)]])
{
s2 _30 = _31;
_30.b.a = 0;
_19.test[0].b.a = _30.b.a;
_19.test[0].b.a = 0;
}

75
reference/opt/shaders-msl/comp/struct-packing.comp
View File

@ -67,39 +67,60 @@ struct SSBO1
float array[1];
};
struct S0_1
{
float2 a[1];
float b;
};
struct Content_1
{
S0_1 m0s[1];
S1 m1s[1];
S2 m2s[1];
S0_1 m0;
S1 m1;
S2 m2;
S3 m3;
char pad7[4];
float m4;
S4 m3s[8];
};
struct SSBO0
{
Content content;
Content content1[2];
Content content2;
Content_1 content;
Content_1 content1[2];
Content_1 content2;
float array[1];
};
kernel void main0(device SSBO1& ssbo_430 [[buffer(0)]], device SSBO0& ssbo_140 [[buffer(1)]])
kernel void main0(device SSBO0& ssbo_140 [[buffer(0)]], device SSBO1& ssbo_430 [[buffer(1)]])
{
ssbo_430.content.m0s[0].a[0] = ssbo_140.content.m0s[0].a[0];
ssbo_430.content.m0s[0].b = ssbo_140.content.m0s[0].b;
ssbo_430.content.m1s[0].a = ssbo_140.content.m1s[0].a;
ssbo_430.content.m1s[0].b = ssbo_140.content.m1s[0].b;
ssbo_430.content.m2s[0].a[0] = ssbo_140.content.m2s[0].a[0];
ssbo_430.content.m2s[0].b = ssbo_140.content.m2s[0].b;
ssbo_430.content.m0.a[0] = ssbo_140.content.m0.a[0];
ssbo_430.content.m0.b = ssbo_140.content.m0.b;
ssbo_430.content.m1.a = ssbo_140.content.m1.a;
ssbo_430.content.m1.b = ssbo_140.content.m1.b;
ssbo_430.content.m2.a[0] = ssbo_140.content.m2.a[0];
ssbo_430.content.m2.b = ssbo_140.content.m2.b;
ssbo_430.content.m3.a = ssbo_140.content.m3.a;
ssbo_430.content.m3.b = ssbo_140.content.m3.b;
ssbo_430.content.m4 = ssbo_140.content.m4;
ssbo_430.content.m3s[0].c = ssbo_140.content.m3s[0].c;
ssbo_430.content.m3s[1].c = ssbo_140.content.m3s[1].c;
ssbo_430.content.m3s[2].c = ssbo_140.content.m3s[2].c;
ssbo_430.content.m3s[3].c = ssbo_140.content.m3s[3].c;
ssbo_430.content.m3s[4].c = ssbo_140.content.m3s[4].c;
ssbo_430.content.m3s[5].c = ssbo_140.content.m3s[5].c;
ssbo_430.content.m3s[6].c = ssbo_140.content.m3s[6].c;
ssbo_430.content.m3s[7].c = ssbo_140.content.m3s[7].c;
Content_1 _60 = ssbo_140.content;
ssbo_430.content.m0s[0].a[0] = _60.m0s[0].a[0];
ssbo_430.content.m0s[0].b = _60.m0s[0].b;
ssbo_430.content.m1s[0].a = _60.m1s[0].a;
ssbo_430.content.m1s[0].b = _60.m1s[0].b;
ssbo_430.content.m2s[0].a[0] = _60.m2s[0].a[0];
ssbo_430.content.m2s[0].b = _60.m2s[0].b;
ssbo_430.content.m0.a[0] = _60.m0.a[0];
ssbo_430.content.m0.b = _60.m0.b;
ssbo_430.content.m1.a = _60.m1.a;
ssbo_430.content.m1.b = _60.m1.b;
ssbo_430.content.m2.a[0] = _60.m2.a[0];
ssbo_430.content.m2.b = _60.m2.b;
ssbo_430.content.m3.a = _60.m3.a;
ssbo_430.content.m3.b = _60.m3.b;
ssbo_430.content.m4 = _60.m4;
ssbo_430.content.m3s[0].c = _60.m3s[0].c;
ssbo_430.content.m3s[1].c = _60.m3s[1].c;
ssbo_430.content.m3s[2].c = _60.m3s[2].c;
ssbo_430.content.m3s[3].c = _60.m3s[3].c;
ssbo_430.content.m3s[4].c = _60.m3s[4].c;
ssbo_430.content.m3s[5].c = _60.m3s[5].c;
ssbo_430.content.m3s[6].c = _60.m3s[6].c;
ssbo_430.content.m3s[7].c = _60.m3s[7].c;
ssbo_430.content.m1.a = ssbo_430.content.m3.a * ssbo_430.m6[1][1];
}

73
reference/opt/shaders-msl/comp/torture-loop.comp
View File

@ -14,66 +14,29 @@ struct SSBO2
float4 out_data[1];
};
constant uint _98 = {};
kernel void main0(device SSBO& _24 [[buffer(0)]], device SSBO2& _89 [[buffer(1)]], uint3 gl_GlobalInvocationID [[thread_position_in_grid]])
kernel void main0(const device SSBO& _24 [[buffer(0)]], device SSBO2& _89 [[buffer(1)]], uint3 gl_GlobalInvocationID [[thread_position_in_grid]])
{
float4 _93;
int _94;
_93 = _24.in_data[gl_GlobalInvocationID.x];
_94 = 0;
int _40;
float4 _46;
int _48;
for (;;)
float4 _99;
_99 = _24.in_data[gl_GlobalInvocationID.x];
for (int _93 = 0; (_93 + 1) < 10; )
{
_40 = _94 + 1;
if (_40 < 10)
_99 *= 2.0;
_93 += 2;
continue;
}
float4 _98;
_98 = _99;
float4 _103;
for (uint _94 = 0u; _94 < 16u; _98 = _103, _94++)
{
_103 = _98;
for (uint _100 = 0u; _100 < 30u; )
{
_46 = _93 * 2.0;
_48 = _40 + 1;
_93 = _46;
_94 = _48;
_103 = _24.mvp * _103;
_100++;
continue;
}
else
{
break;
}
}
float4 _95;
int _96;
_95 = _93;
_96 = _40;
float4 _100;
uint _101;
uint _99;
for (uint _97 = 0u; _97 < 16u; _95 = _100, _96++, _97++, _99 = _101)
{
_100 = _95;
_101 = 0u;
float4 _71;
for (; _101 < 30u; _100 = _71, _101++)
{
_71 = _24.mvp * _100;
}
}
int _102;
_102 = _96;
int _83;
for (;;)
{
_83 = _102 + 1;
if (_83 > 10)
{
_102 = _83;
continue;
}
else
{
break;
}
}
_89.out_data[gl_GlobalInvocationID.x] = _95;
_89.out_data[gl_GlobalInvocationID.x] = _98;
}

14
reference/opt/shaders-msl/comp/type-alias.comp
View File

@ -8,28 +8,18 @@ struct S0
float4 a;
};
struct S1
{
float4 a;
};
struct SSBO0
{
S0 s0s[1];
};
struct SSBO1
{
S1 s1s[1];
};
struct SSBO2
{
float4 outputs[1];
};
kernel void main0(device SSBO0& _36 [[buffer(0)]], device SSBO1& _55 [[buffer(1)]], device SSBO2& _66 [[buffer(2)]], uint3 gl_GlobalInvocationID [[thread_position_in_grid]])
kernel void main0(device SSBO0& _36 [[buffer(0)]], device SSBO0& _55 [[buffer(1)]], device SSBO2& _66 [[buffer(2)]], uint3 gl_GlobalInvocationID [[thread_position_in_grid]])
{
_66.outputs[gl_GlobalInvocationID.x] = _36.s0s[gl_GlobalInvocationID.x].a + _55.s1s[gl_GlobalInvocationID.x].a;
_66.outputs[gl_GlobalInvocationID.x] = _36.s0s[gl_GlobalInvocationID.x].a + _55.s0s[gl_GlobalInvocationID.x].a;
}

9
reference/opt/shaders-msl/comp/udiv.comp
View File

@ -8,13 +8,8 @@ struct SSBO2
uint outputs[1];
};
struct SSBO
kernel void main0(device SSBO2& _23 [[buffer(0)]], device SSBO2& _10 [[buffer(1)]], uint3 gl_GlobalInvocationID [[thread_position_in_grid]])
{
uint inputs[1];
};
kernel void main0(device SSBO2& _10 [[buffer(0)]], device SSBO& _23 [[buffer(1)]], uint3 gl_GlobalInvocationID [[thread_position_in_grid]])
{
_10.outputs[gl_GlobalInvocationID.x] = _23.inputs[gl_GlobalInvocationID.x] / 29u;
_10.outputs[gl_GlobalInvocationID.x] = _23.outputs[gl_GlobalInvocationID.x] / 29u;
}

2
reference/opt/shaders-msl/desktop-only/frag/image-ms.desktop.frag
View File

@ -3,7 +3,7 @@
using namespace metal;
fragment void main0(texture2d_ms<float> uImageMS [[texture(0)]], texture2d_array<float, access::read_write> uImageArray [[texture(1)]], texture2d<float, access::write> uImage [[texture(2)]])
fragment void main0(texture2d<float, access::write> uImage [[texture(0)]], texture2d_array<float, access::read_write> uImageArray [[texture(1)]], texture2d_ms<float> uImageMS [[texture(2)]])
{
uImage.write(uImageMS.read(uint2(int2(1, 2)), 2), uint2(int2(2, 3)));
uImageArray.write(uImageArray.read(uint2(int3(1, 2, 4).xy), uint(int3(1, 2, 4).z)), uint2(int3(2, 3, 7).xy), uint(int3(2, 3, 7).z));

12
reference/opt/shaders-msl/desktop-only/vert/basic.desktop.sso.vert
View File

@ -8,18 +8,18 @@ struct UBO
float4x4 uMVP;
};
struct main0_in
{
float3 aNormal [[attribute(1)]];
float4 aVertex [[attribute(0)]];
};
struct main0_out
{
float3 vNormal [[user(locn0)]];
float4 gl_Position [[position]];
};
struct main0_in
{
float4 aVertex [[attribute(0)]];
float3 aNormal [[attribute(1)]];
};
vertex main0_out main0(main0_in in [[stage_in]], constant UBO& _16 [[buffer(0)]])
{
main0_out out = {};

12
reference/opt/shaders-msl/flatten/basic.flatten.vert
View File

@ -8,18 +8,18 @@ struct UBO
float4x4 uMVP;
};
struct main0_in
{
float3 aNormal [[attribute(1)]];
float4 aVertex [[attribute(0)]];
};
struct main0_out
{
float3 vNormal [[user(locn0)]];
float4 gl_Position [[position]];
};
struct main0_in
{
float4 aVertex [[attribute(0)]];
float3 aNormal [[attribute(1)]];
};
vertex main0_out main0(main0_in in [[stage_in]], constant UBO& _16 [[buffer(0)]])
{
main0_out out = {};

10
reference/opt/shaders-msl/flatten/multiindex.flatten.vert
View File

@ -8,16 +8,16 @@ struct UBO
float4 Data[3][5];
};
struct main0_in
{
int2 aIndex [[attribute(0)]];
};
struct main0_out
{
float4 gl_Position [[position]];
};
struct main0_in
{
int2 aIndex [[attribute(0)]];
};
vertex main0_out main0(main0_in in [[stage_in]], constant UBO& _20 [[buffer(0)]])
{
main0_out out = {};

12
reference/opt/shaders-msl/flatten/push-constant.flatten.vert
View File

@ -10,18 +10,18 @@ struct PushMe
float Arr[4];
};
struct main0_in
{
float4 Pos [[attribute(1)]];
float2 Rot [[attribute(0)]];
};
struct main0_out
{
float2 vRot [[user(locn0)]];
float4 gl_Position [[position]];
};
struct main0_in
{
float2 Rot [[attribute(0)]];
float4 Pos [[attribute(1)]];
};
vertex main0_out main0(main0_in in [[stage_in]], constant PushMe& registers [[buffer(0)]])
{
main0_out out = {};

10
reference/opt/shaders-msl/flatten/rowmajor.flatten.vert
View File

@ -10,16 +10,16 @@ struct UBO
float2x4 uMVP;
};
struct main0_in
{
float4 aVertex [[attribute(0)]];
};
struct main0_out
{
float4 gl_Position [[position]];
};
struct main0_in
{
float4 aVertex [[attribute(0)]];
};
vertex main0_out main0(main0_in in [[stage_in]], constant UBO& _18 [[buffer(0)]])
{
main0_out out = {};

14
reference/opt/shaders-msl/flatten/struct.flatten.vert
View File

@ -16,24 +16,24 @@ struct UBO
Light light;
};
struct main0_in
{
float3 aNormal [[attribute(1)]];
float4 aVertex [[attribute(0)]];
};
struct main0_out
{
float4 vColor [[user(locn0)]];
float4 gl_Position [[position]];
};
struct main0_in
{
float4 aVertex [[attribute(0)]];
float3 aNormal [[attribute(1)]];
};
vertex main0_out main0(main0_in in [[stage_in]], constant UBO& _18 [[buffer(0)]])
{
main0_out out = {};
out.gl_Position = _18.uMVP * in.aVertex;
out.vColor = float4(0.0);
float3 _39 = in.aVertex.xyz - _18.light.Position;
float3 _39 = in.aVertex.xyz - float3(_18.light.Position);
out.vColor += ((_18.light.Color * clamp(1.0 - (length(_39) / _18.light.Radius), 0.0, 1.0)) * dot(in.aNormal, normalize(_39)));
return out;
}

Some files were not shown because too many files have changed in this diff Show More