colin
/
onnx-mlir
1
0
Fork 0
Code Issues Pull Requests Packages Projects Releases Wiki Activity
onnx-mlir/README.md

107 lines
3.2 KiB
Markdown
Raw Normal View History

[NFC] Change ONNF based names to ONNX-MLIR (#32) * Rename onnf to onnx-mlir. * Change workspace name.
2020-03-17 21:16:33 +08:00
# ONNX MLIR
The Open Neural Network Exchange implementation in MLIR.
add minimum documentation
2019-12-24 16:59:48 +08:00
Switch to new status badge.
2020-03-02 20:37:33 +08:00
[![CircleCI](https://circleci.com/gh/onnx/onnx-mlir/tree/master.svg?style=svg)](https://circleci.com/gh/onnx/onnx-mlir/tree/master)
add minimum documentation
2019-12-24 16:59:48 +08:00
Update README.md
2020-02-14 04:51:39 +08:00
## Prerequisites
Update README.md
2020-02-14 04:52:53 +08:00
```
Update README.md
2020-02-14 04:51:39 +08:00
gcc >= 6.4
libprotoc >= 3.11.0
cmake >= 3.15.4
Update README.md
2020-02-14 04:52:53 +08:00
```
Update README.md
2020-02-14 04:51:39 +08:00
add minimum documentation
2019-12-24 16:59:48 +08:00
## Installation
[RFC] Doc-check utility. (#12) * 1. Implement doc-check utility. * 1. Move ONNF installation script to a standalone script file. * 1. Modify build script to install llvm-project next to ONNF. The build script used to install llvm-project inside ONNF, which didn't make sense. * 1. Check out code to ONNF directory. * 1. Pass path parameter correctly. * 1. Debugging buildbot. * 1. Remove debug code. * 1. Update installation instructions in README.md. 2. Enforce consistency with scripts used in testing using doc-check. * 1. Fix error with respect to syntax to build multiple CMake targets. * 1. Move doc-check to doc_check. 2. Remove directive_config in top-level driver. * 1. Build onnf and check-mlir-lit separately because only CMake 3.15+ supports building multiple targets in one cmake --build run. * 1. Use new env variables to locate LLVM-Project. * 1. Documentation nits. * 1. Prettify buildbot scripts. * 1. Fix build script error. * 1. Support exclude_dirs in DocCheck. 2. Add README for DocCheck. * 1. Mark python3 interpreter as required. 2. Use imported interpreter target. * 1. Automatically deduce doc file extension in DocCheckCtx. 2. Rename ctx.open -> ctx.open_doc since it should only be used to open doc file. 3. Always read line in parser, instead of reading lines in driver and then passing it to parser.py. * 1. Rename parser -> doc_parser due to name conflict with python built-in module. 2. Explose doc_check module directory first before importing; otherwise if the doc_check utility is invoked by other script, importing will not work correctly. * 1. Keep renaming parser -> doc_parser. 2. Explicitly define a default configuration parser that parses the configuration into a python dictionary. * 1. Add test for doc-check. 2. Exclude doc-check tests from project dock-check because base directory is different. * 1. Raise ValueError if directive configuration fails to parse. 2. Format code. * Shorten test case documentation. Show example of using same-as-file directive, check with DocCheck. * 1. Shorten test case documentation. 2. More documentation, check documentation with DocCheck. * 1. Add copyright notice. * 1. Make documentation clearer. 2. Prettify build-scripts. * 1. Provide more documentation. 2. Fix some non-compliance with pep8 recommendations. Co-authored-by: Gheorghe-Teodor Bercea <gt.bercea@gmail.com>
2020-01-10 07:35:52 +08:00
Firstly, install MLIR (as a part of LLVM-Project):
[same-as-file]: <> (utils/install-mlir.sh)
``` bash
git clone https://github.com/llvm/llvm-project.git
[NFC] Change ONNF based names to ONNX-MLIR (#32) * Rename onnf to onnx-mlir. * Change workspace name.
2020-03-17 21:16:33 +08:00
# Check out a specific branch that is known to work with ONNX MLIR.
Move to more recent LLVM commit ID (#64) * Move to more recent LLVM commit ID * Update LLVM cache version from V9 to V10 * Update to latest LLVM commit id from master, roll back conditions in util scripts * Update circlci LLVM cache tag to ensure ci updates builds with latest LLVM commit id * Update README.md to have matching LLVM commit id * Update doc/Dialtects/onnx.md
2020-04-02 00:38:34 +08:00
cd llvm-project && git checkout 07e462526d0cbae40b320e1a4307ce11e197fb0a && cd ..
[RFC] Doc-check utility. (#12) * 1. Implement doc-check utility. * 1. Move ONNF installation script to a standalone script file. * 1. Modify build script to install llvm-project next to ONNF. The build script used to install llvm-project inside ONNF, which didn't make sense. * 1. Check out code to ONNF directory. * 1. Pass path parameter correctly. * 1. Debugging buildbot. * 1. Remove debug code. * 1. Update installation instructions in README.md. 2. Enforce consistency with scripts used in testing using doc-check. * 1. Fix error with respect to syntax to build multiple CMake targets. * 1. Move doc-check to doc_check. 2. Remove directive_config in top-level driver. * 1. Build onnf and check-mlir-lit separately because only CMake 3.15+ supports building multiple targets in one cmake --build run. * 1. Use new env variables to locate LLVM-Project. * 1. Documentation nits. * 1. Prettify buildbot scripts. * 1. Fix build script error. * 1. Support exclude_dirs in DocCheck. 2. Add README for DocCheck. * 1. Mark python3 interpreter as required. 2. Use imported interpreter target. * 1. Automatically deduce doc file extension in DocCheckCtx. 2. Rename ctx.open -> ctx.open_doc since it should only be used to open doc file. 3. Always read line in parser, instead of reading lines in driver and then passing it to parser.py. * 1. Rename parser -> doc_parser due to name conflict with python built-in module. 2. Explose doc_check module directory first before importing; otherwise if the doc_check utility is invoked by other script, importing will not work correctly. * 1. Keep renaming parser -> doc_parser. 2. Explicitly define a default configuration parser that parses the configuration into a python dictionary. * 1. Add test for doc-check. 2. Exclude doc-check tests from project dock-check because base directory is different. * 1. Raise ValueError if directive configuration fails to parse. 2. Format code. * Shorten test case documentation. Show example of using same-as-file directive, check with DocCheck. * 1. Shorten test case documentation. 2. More documentation, check documentation with DocCheck. * 1. Add copyright notice. * 1. Make documentation clearer. 2. Prettify build-scripts. * 1. Provide more documentation. 2. Fix some non-compliance with pep8 recommendations. Co-authored-by: Gheorghe-Teodor Bercea <gt.bercea@gmail.com>
2020-01-10 07:35:52 +08:00
mkdir llvm-project/build
cd llvm-project/build
cmake -G Ninja ../llvm \
-DLLVM_ENABLE_PROJECTS=mlir \
-DLLVM_BUILD_EXAMPLES=ON \
-DLLVM_TARGETS_TO_BUILD="host" \
-DCMAKE_BUILD_TYPE=Release \
-DLLVM_ENABLE_ASSERTIONS=ON \
-DLLVM_ENABLE_RTTI=ON
Resolve buildbot issue. (#64) * 1. Try limit parallelism. * Limit build parallelism to 4. * Fix doc-check failure. Co-authored-by: Gheorghe-Teodor Bercea <gt.bercea@gmail.com>
2020-01-31 23:24:45 +08:00
cmake --build . --target -- ${MAKEFLAGS}
Enable e2e tests (#29) * Sync with latest MLIR. * Enable ONNX backend tests as a means to test ONNF lowering end-to-end. * Install ONNX using quiet mode. * Remove debug comments. * Install ONNX from third_party/onnx. * Check python version and fix pip command for installing ONNX. * Using --user install option to prevent permission denied. * Remove unused imports. * Try using stock ONNX pip package as there are more tests in them. * Pip got stuck building wheels, try sudo. * Use verbose install to debug. * Invalidate cache to build LLVM tools. * Fix mlir installation script location. * Debug to locate ONNF. * Sanity check. * Check out ONNF code first. * Use verbose LIT output. * 1. Update documentation to always use verbose LIT. 2. Update krnl ops to reflect new affine map attribute syntax. * See if conda exists * Install ONNX by manually cloning the repo. * Install cmake first. * Using sudo priviledge when installing. * Limit build parallelism. * Limit parallelism. * Larger memory. * Install onnx package with pip. * Build MLIR tools. * Invalidate cache. * Compile model.so with -fPIC. * Remove module dump to get concise debug output. * Print command before executing. * Use quiet install mode to reduce logging. * Use -relocation-model=pic to generate position independent code. * 1. Remove MAKEFLAGS because now buildbot has enough memory. 2. Run DocCheck as a last step. * 1. Add verbose mode for backtend test. * When dumping to LLVM bitcode, do not dump module IR, but print a message indicating that bitcode has been written to disk. * Do not pass MakeFlags to CMake. * Add more explaination for posible reasons of failing to identify tests.
2020-01-21 01:30:08 +08:00
cmake --build . --target check-mlir
[RFC] Doc-check utility. (#12) * 1. Implement doc-check utility. * 1. Move ONNF installation script to a standalone script file. * 1. Modify build script to install llvm-project next to ONNF. The build script used to install llvm-project inside ONNF, which didn't make sense. * 1. Check out code to ONNF directory. * 1. Pass path parameter correctly. * 1. Debugging buildbot. * 1. Remove debug code. * 1. Update installation instructions in README.md. 2. Enforce consistency with scripts used in testing using doc-check. * 1. Fix error with respect to syntax to build multiple CMake targets. * 1. Move doc-check to doc_check. 2. Remove directive_config in top-level driver. * 1. Build onnf and check-mlir-lit separately because only CMake 3.15+ supports building multiple targets in one cmake --build run. * 1. Use new env variables to locate LLVM-Project. * 1. Documentation nits. * 1. Prettify buildbot scripts. * 1. Fix build script error. * 1. Support exclude_dirs in DocCheck. 2. Add README for DocCheck. * 1. Mark python3 interpreter as required. 2. Use imported interpreter target. * 1. Automatically deduce doc file extension in DocCheckCtx. 2. Rename ctx.open -> ctx.open_doc since it should only be used to open doc file. 3. Always read line in parser, instead of reading lines in driver and then passing it to parser.py. * 1. Rename parser -> doc_parser due to name conflict with python built-in module. 2. Explose doc_check module directory first before importing; otherwise if the doc_check utility is invoked by other script, importing will not work correctly. * 1. Keep renaming parser -> doc_parser. 2. Explicitly define a default configuration parser that parses the configuration into a python dictionary. * 1. Add test for doc-check. 2. Exclude doc-check tests from project dock-check because base directory is different. * 1. Raise ValueError if directive configuration fails to parse. 2. Format code. * Shorten test case documentation. Show example of using same-as-file directive, check with DocCheck. * 1. Shorten test case documentation. 2. More documentation, check documentation with DocCheck. * 1. Add copyright notice. * 1. Make documentation clearer. 2. Prettify build-scripts. * 1. Provide more documentation. 2. Fix some non-compliance with pep8 recommendations. Co-authored-by: Gheorghe-Teodor Bercea <gt.bercea@gmail.com>
2020-01-10 07:35:52 +08:00
```
add minimum documentation
2019-12-24 16:59:48 +08:00
Two environment variables need to be set:
Update README.md (#23)
2020-01-14 00:40:51 +08:00
- LLVM_PROJ_SRC should point to the llvm-project src directory (e.g., llvm-project/).
- LLVM_PROJ_BUILD should point to the llvm-project build directory (e.g., llvm-project/build).
add minimum documentation
2019-12-24 16:59:48 +08:00
[NFC] Change ONNF based names to ONNX-MLIR (#32) * Rename onnf to onnx-mlir. * Change workspace name.
2020-03-17 21:16:33 +08:00
To build ONNX-MLIR, use the following command:
[RFC] Doc-check utility. (#12) * 1. Implement doc-check utility. * 1. Move ONNF installation script to a standalone script file. * 1. Modify build script to install llvm-project next to ONNF. The build script used to install llvm-project inside ONNF, which didn't make sense. * 1. Check out code to ONNF directory. * 1. Pass path parameter correctly. * 1. Debugging buildbot. * 1. Remove debug code. * 1. Update installation instructions in README.md. 2. Enforce consistency with scripts used in testing using doc-check. * 1. Fix error with respect to syntax to build multiple CMake targets. * 1. Move doc-check to doc_check. 2. Remove directive_config in top-level driver. * 1. Build onnf and check-mlir-lit separately because only CMake 3.15+ supports building multiple targets in one cmake --build run. * 1. Use new env variables to locate LLVM-Project. * 1. Documentation nits. * 1. Prettify buildbot scripts. * 1. Fix build script error. * 1. Support exclude_dirs in DocCheck. 2. Add README for DocCheck. * 1. Mark python3 interpreter as required. 2. Use imported interpreter target. * 1. Automatically deduce doc file extension in DocCheckCtx. 2. Rename ctx.open -> ctx.open_doc since it should only be used to open doc file. 3. Always read line in parser, instead of reading lines in driver and then passing it to parser.py. * 1. Rename parser -> doc_parser due to name conflict with python built-in module. 2. Explose doc_check module directory first before importing; otherwise if the doc_check utility is invoked by other script, importing will not work correctly. * 1. Keep renaming parser -> doc_parser. 2. Explicitly define a default configuration parser that parses the configuration into a python dictionary. * 1. Add test for doc-check. 2. Exclude doc-check tests from project dock-check because base directory is different. * 1. Raise ValueError if directive configuration fails to parse. 2. Format code. * Shorten test case documentation. Show example of using same-as-file directive, check with DocCheck. * 1. Shorten test case documentation. 2. More documentation, check documentation with DocCheck. * 1. Add copyright notice. * 1. Make documentation clearer. 2. Prettify build-scripts. * 1. Provide more documentation. 2. Fix some non-compliance with pep8 recommendations. Co-authored-by: Gheorghe-Teodor Bercea <gt.bercea@gmail.com>
2020-01-10 07:35:52 +08:00
[NFC] Change ONNF based names to ONNX-MLIR (#32) * Rename onnf to onnx-mlir. * Change workspace name.
2020-03-17 21:16:33 +08:00
[same-as-file]: <> ({"ref": "utils/install-onnx-mlir.sh", "skip-doc": 2})
add minimum documentation
2019-12-24 16:59:48 +08:00
```
[NFC] Change ONNF based names to ONNX-MLIR (#32) * Rename onnf to onnx-mlir. * Change workspace name.
2020-03-17 21:16:33 +08:00
git clone --recursive git@github.com:onnx/onnx-mlir.git
[RFC] Doc-check utility. (#12) * 1. Implement doc-check utility. * 1. Move ONNF installation script to a standalone script file. * 1. Modify build script to install llvm-project next to ONNF. The build script used to install llvm-project inside ONNF, which didn't make sense. * 1. Check out code to ONNF directory. * 1. Pass path parameter correctly. * 1. Debugging buildbot. * 1. Remove debug code. * 1. Update installation instructions in README.md. 2. Enforce consistency with scripts used in testing using doc-check. * 1. Fix error with respect to syntax to build multiple CMake targets. * 1. Move doc-check to doc_check. 2. Remove directive_config in top-level driver. * 1. Build onnf and check-mlir-lit separately because only CMake 3.15+ supports building multiple targets in one cmake --build run. * 1. Use new env variables to locate LLVM-Project. * 1. Documentation nits. * 1. Prettify buildbot scripts. * 1. Fix build script error. * 1. Support exclude_dirs in DocCheck. 2. Add README for DocCheck. * 1. Mark python3 interpreter as required. 2. Use imported interpreter target. * 1. Automatically deduce doc file extension in DocCheckCtx. 2. Rename ctx.open -> ctx.open_doc since it should only be used to open doc file. 3. Always read line in parser, instead of reading lines in driver and then passing it to parser.py. * 1. Rename parser -> doc_parser due to name conflict with python built-in module. 2. Explose doc_check module directory first before importing; otherwise if the doc_check utility is invoked by other script, importing will not work correctly. * 1. Keep renaming parser -> doc_parser. 2. Explicitly define a default configuration parser that parses the configuration into a python dictionary. * 1. Add test for doc-check. 2. Exclude doc-check tests from project dock-check because base directory is different. * 1. Raise ValueError if directive configuration fails to parse. 2. Format code. * Shorten test case documentation. Show example of using same-as-file directive, check with DocCheck. * 1. Shorten test case documentation. 2. More documentation, check documentation with DocCheck. * 1. Add copyright notice. * 1. Make documentation clearer. 2. Prettify build-scripts. * 1. Provide more documentation. 2. Fix some non-compliance with pep8 recommendations. Co-authored-by: Gheorghe-Teodor Bercea <gt.bercea@gmail.com>
2020-01-10 07:35:52 +08:00
# Export environment variables pointing to LLVM-Projects.
export LLVM_PROJ_SRC=$(pwd)/llvm-project/
export LLVM_PROJ_BUILD=$(pwd)/llvm-project/build
[NFC] Change ONNF based names to ONNX-MLIR (#32) * Rename onnf to onnx-mlir. * Change workspace name.
2020-03-17 21:16:33 +08:00
mkdir onnx-mlir/build && cd onnx-mlir/build
add minimum documentation
2019-12-24 16:59:48 +08:00
cmake ..
[NFC] Change ONNF based names to ONNX-MLIR (#32) * Rename onnf to onnx-mlir. * Change workspace name.
2020-03-17 21:16:33 +08:00
cmake --build . --target onnx-mlir
[RFC] Doc-check utility. (#12) * 1. Implement doc-check utility. * 1. Move ONNF installation script to a standalone script file. * 1. Modify build script to install llvm-project next to ONNF. The build script used to install llvm-project inside ONNF, which didn't make sense. * 1. Check out code to ONNF directory. * 1. Pass path parameter correctly. * 1. Debugging buildbot. * 1. Remove debug code. * 1. Update installation instructions in README.md. 2. Enforce consistency with scripts used in testing using doc-check. * 1. Fix error with respect to syntax to build multiple CMake targets. * 1. Move doc-check to doc_check. 2. Remove directive_config in top-level driver. * 1. Build onnf and check-mlir-lit separately because only CMake 3.15+ supports building multiple targets in one cmake --build run. * 1. Use new env variables to locate LLVM-Project. * 1. Documentation nits. * 1. Prettify buildbot scripts. * 1. Fix build script error. * 1. Support exclude_dirs in DocCheck. 2. Add README for DocCheck. * 1. Mark python3 interpreter as required. 2. Use imported interpreter target. * 1. Automatically deduce doc file extension in DocCheckCtx. 2. Rename ctx.open -> ctx.open_doc since it should only be used to open doc file. 3. Always read line in parser, instead of reading lines in driver and then passing it to parser.py. * 1. Rename parser -> doc_parser due to name conflict with python built-in module. 2. Explose doc_check module directory first before importing; otherwise if the doc_check utility is invoked by other script, importing will not work correctly. * 1. Keep renaming parser -> doc_parser. 2. Explicitly define a default configuration parser that parses the configuration into a python dictionary. * 1. Add test for doc-check. 2. Exclude doc-check tests from project dock-check because base directory is different. * 1. Raise ValueError if directive configuration fails to parse. 2. Format code. * Shorten test case documentation. Show example of using same-as-file directive, check with DocCheck. * 1. Shorten test case documentation. 2. More documentation, check documentation with DocCheck. * 1. Add copyright notice. * 1. Make documentation clearer. 2. Prettify build-scripts. * 1. Provide more documentation. 2. Fix some non-compliance with pep8 recommendations. Co-authored-by: Gheorghe-Teodor Bercea <gt.bercea@gmail.com>
2020-01-10 07:35:52 +08:00
# Run FileCheck tests:
Enable e2e tests (#29) * Sync with latest MLIR. * Enable ONNX backend tests as a means to test ONNF lowering end-to-end. * Install ONNX using quiet mode. * Remove debug comments. * Install ONNX from third_party/onnx. * Check python version and fix pip command for installing ONNX. * Using --user install option to prevent permission denied. * Remove unused imports. * Try using stock ONNX pip package as there are more tests in them. * Pip got stuck building wheels, try sudo. * Use verbose install to debug. * Invalidate cache to build LLVM tools. * Fix mlir installation script location. * Debug to locate ONNF. * Sanity check. * Check out ONNF code first. * Use verbose LIT output. * 1. Update documentation to always use verbose LIT. 2. Update krnl ops to reflect new affine map attribute syntax. * See if conda exists * Install ONNX by manually cloning the repo. * Install cmake first. * Using sudo priviledge when installing. * Limit build parallelism. * Limit parallelism. * Larger memory. * Install onnx package with pip. * Build MLIR tools. * Invalidate cache. * Compile model.so with -fPIC. * Remove module dump to get concise debug output. * Print command before executing. * Use quiet install mode to reduce logging. * Use -relocation-model=pic to generate position independent code. * 1. Remove MAKEFLAGS because now buildbot has enough memory. 2. Run DocCheck as a last step. * 1. Add verbose mode for backtend test. * When dumping to LLVM bitcode, do not dump module IR, but print a message indicating that bitcode has been written to disk. * Do not pass MakeFlags to CMake. * Add more explaination for posible reasons of failing to identify tests.
2020-01-21 01:30:08 +08:00
export LIT_OPTS=-v
Name change for tests, to be check-onnx-(lit | backend) (#62)
2020-03-31 22:06:14 +08:00
cmake --build . --target check-onnx-lit
add minimum documentation
2019-12-24 16:59:48 +08:00
```
[NFC] Change ONNF based names to ONNX-MLIR (#32) * Rename onnf to onnx-mlir. * Change workspace name.
2020-03-17 21:16:33 +08:00
After the above commands succeed, an `onnx-mlir` executable should appear in the `bin` directory.
add minimum documentation
2019-12-24 16:59:48 +08:00
[NFC] Change ONNF based names to ONNX-MLIR (#32) * Rename onnf to onnx-mlir. * Change workspace name.
2020-03-17 21:16:33 +08:00
## Using ONNX MLIR
add minimum documentation
2019-12-24 16:59:48 +08:00
[NFC] Change ONNF based names to ONNX-MLIR (#32) * Rename onnf to onnx-mlir. * Change workspace name.
2020-03-17 21:16:33 +08:00
The usage of `onnx-mlir` is as such:
add minimum documentation
2019-12-24 16:59:48 +08:00
```
[NFC] Change ONNF based names to ONNX-MLIR (#32) * Rename onnf to onnx-mlir. * Change workspace name.
2020-03-17 21:16:33 +08:00
OVERVIEW: ONNX MLIR modular optimizer driver
add minimum documentation
2019-12-24 16:59:48 +08:00
[NFC] Change ONNF based names to ONNX-MLIR (#32) * Rename onnf to onnx-mlir. * Change workspace name.
2020-03-17 21:16:33 +08:00
USAGE: onnx-mlir [options] <input file>
add minimum documentation
2019-12-24 16:59:48 +08:00
OPTIONS:
Generic Options:
--help - Display available options (--help-hidden for more)
--help-list - Display list of available options (--help-list-hidden for more)
--version - Display the version of this program
[NFC] Change ONNF based names to ONNX-MLIR (#32) * Rename onnf to onnx-mlir. * Change workspace name.
2020-03-17 21:16:33 +08:00
ONNX MLIR Options:
add minimum documentation
2019-12-24 16:59:48 +08:00
These are frontend options.
Choose target to emit:
--EmitONNXIR - Ingest ONNX and emit corresponding ONNX dialect.
--EmitMLIR - Lower model to MLIR built-in transformation dialect.
--EmitLLVMIR - Lower model to LLVM IR (LLVM dialect).
--EmitLLVMBC - Lower model to LLVM IR and emit (to file) LLVM bitcode for model.
```
## Example
For example, to lower an ONNX model (e.g., add.onnx) to ONNX dialect, use the following command:
```
[NFC] Change ONNF based names to ONNX-MLIR (#32) * Rename onnf to onnx-mlir. * Change workspace name.
2020-03-17 21:16:33 +08:00
./onnx-mlir --EmitONNXIR add.onnx
add minimum documentation
2019-12-24 16:59:48 +08:00
```
The output should look like:
```
module {
func @main_graph(%arg0: tensor<10x10x10xf32>, %arg1: tensor<10x10x10xf32>) -> tensor<10x10x10xf32> {
%0 = "onnx.Add"(%arg0, %arg1) : (tensor<10x10x10xf32>, tensor<10x10x10xf32>) -> tensor<10x10x10xf32>
return %0 : tensor<10x10x10xf32>
}
}
```
Update README.md
2020-02-14 02:43:19 +08:00
## Troubleshooting
Update README.md
2020-02-14 02:46:39 +08:00
If the latest LLVM project fails to work due to the latest changes to the MLIR subproject please consider using a slightly older version of LLVM. One such version, which we use, can be found [here](https://github.com/clang-ykt/llvm-project).