Skip to content
New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

triton-mlir LLVM commit hash #828

Closed
makslevental opened this issue Oct 31, 2022 · 1 comment
Closed

triton-mlir LLVM commit hash #828

makslevental opened this issue Oct 31, 2022 · 1 comment

Comments

@makslevental
Copy link
Contributor

qq: which LLVM commit hash is the triton-mlir branch supposed to be built against?

@makslevental
Copy link
Contributor Author

makslevental commented Oct 31, 2022

closing as the correct thing to do is use setup.py, which lists llvm-14.0.0-329fda39c507 as the correct version

EDIT:

In case anyone lands here, here's the info on how to compile the CMake project

LLVM

commit f28c006a5895fc0e329fe15fead81e37457cb1d1 (HEAD, tag: llvmorg-14.0.6, llvm/release/14.x)

cmake -G "${CMAKE_GENERATOR}" \
    ../llvm \
    -DLLVM_ENABLE_PROJECTS="mlir" \
    -DLLVM_TARGETS_TO_BUILD="host" \
    -DLLVM_EXPERIMENTAL_TARGETS_TO_BUILD="NVPTX;AMDGPU" \
    -DLLVM_ENABLE_ASSERTIONS=ON \
    -DCMAKE_BUILD_TYPE=DEBUG \
    -DMLIR_ENABLE_BINDINGS_PYTHON=ON \
    -DLLVM_USE_LINKER=lld \
    -DCMAKE_C_COMPILER=clang \
    -DCMAKE_CXX_COMPILER=clang++ \
    -DBUILD_SHARED_LIBS=OFF

Triton

CMake

-DLLVM_LIBRARY_DIR=/home/mlevental/dev_projects/llvm-project/build/lib
-DTRITON_BUILD_PYTHON_MODULE=ON

and a slight bit of surgery on the triton source

diff --git a/CMakeLists.txt b/CMakeLists.txt
index b02044001..8afca31d6 100644
--- a/CMakeLists.txt
+++ b/CMakeLists.txt
@@ -230,6 +230,6 @@ if(TRITON_BUILD_PYTHON_MODULE AND NOT WIN32)
     target_link_libraries(triton ${CUTLASS_LIBRARIES} ${PYTHON_LDFLAGS})
 endif()
 
-add_subdirectory(test)
-
-add_subdirectory(unittest)
+#add_subdirectory(test)
+#
+#add_subdirectory(unittest)
diff --git a/bin/CMakeLists.txt b/bin/CMakeLists.txt
index 7fb68f48a..719fbf604 100644
--- a/bin/CMakeLists.txt
+++ b/bin/CMakeLists.txt
@@ -16,7 +16,7 @@ target_link_libraries(triton-opt PRIVATE
   ${dialect_libs}
   ${conversion_libs}
   # tests
-  TritonTestAnalysis
+#  TritonTestAnalysis
   # MLIR core
   MLIROptLib
   MLIRPass
diff --git a/bin/FileCheck/CMakeLists.txt b/bin/FileCheck/CMakeLists.txt
index 1b683be51..51bed53c4 100644
--- a/bin/FileCheck/CMakeLists.txt
+++ b/bin/FileCheck/CMakeLists.txt
@@ -1,2 +1,2 @@
-add_llvm_executable(FileCheck FileCheck.cpp)
-target_link_libraries(FileCheck PRIVATE LLVMFileCheck LLVMSupport)
\ No newline at end of file
+#add_llvm_executable(FileCheck FileCheck.cpp)
+#target_link_libraries(FileCheck PRIVATE LLVMFileCheck LLVMSupport)
\ No newline at end of file
diff --git a/bin/triton-opt.cpp b/bin/triton-opt.cpp
index 9f3b53b7a..b60c4ed04 100644
--- a/bin/triton-opt.cpp
+++ b/bin/triton-opt.cpp
@@ -23,10 +23,10 @@ int main(int argc, char **argv) {
   mlir::registerAllPasses();
   mlir::registerTritonPasses();
   mlir::registerTritonGPUPasses();
-  mlir::test::registerTestAliasPass();
-  mlir::test::registerTestAlignmentPass();
-  mlir::test::registerTestAllocationPass();
-  mlir::test::registerTestMembarPass();
+//  mlir::test::registerTestAliasPass();
+//  mlir::test::registerTestAlignmentPass();
+//  mlir::test::registerTestAllocationPass();
+//  mlir::test::registerTestMembarPass();
   mlir::triton::registerConvertTritonToTritonGPUPass();
   mlir::triton::registerConvertTritonGPUToLLVMPass();

ZzEeKkAa pushed a commit to ZzEeKkAa/triton that referenced this issue Aug 5, 2024
Resolves triton-lang#827

Signed-off-by: Ilya Enkovich <ilya.enkovich@intel.com>
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Labels
None yet
Projects
None yet
Development

No branches or pull requests

1 participant