[libc] Configure CMAKE_REQUIRED_FLAGS so the GPU can use flag checks #95424
Conversation
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
|
@llvm/pr-subscribers-libc Author: Joseph Huber (jhuber6) ChangesSummary: Full diff: https://github.com/llvm/llvm-project/pull/95424.diff 2 Files Affected:
diff --git a/libc/cmake/modules/prepare_libc_gpu_build.cmake b/libc/cmake/modules/prepare_libc_gpu_build.cmake
index dc8beb14fd7f4..f1275e34a42e8 100644
--- a/libc/cmake/modules/prepare_libc_gpu_build.cmake
+++ b/libc/cmake/modules/prepare_libc_gpu_build.cmake
@@ -16,6 +16,17 @@ if(NOT LLVM_LIBC_FULL_BUILD)
"GPU.")
endif()
+# Set the required flags globally so standard CMake utilities can compile.
+if(LIBC_TARGET_TRIPLE)
+ set(CMAKE_REQUIRED_FLAGS "--target=${explicit_target_triple}")
+endif()
+if(LIBC_TARGET_ARCHITECTURE_IS_AMDGPU)
+ set(CMAKE_REQUIRED_FLAGS "${CMAKE_REQUIRED_FLAGS} -nogpulib")
+elseif(LIBC_TARGET_ARCHITECTURE_IS_NVPTX)
+ set(CMAKE_REQUIRED_FLAGS
+ "${CMAKE_REQUIRED_FLAGS} -flto -c -Wno-unused-command-line-argument")
+endif()
+
# Identify the program used to package multiple images into a single binary.
get_filename_component(compiler_path ${CMAKE_CXX_COMPILER} DIRECTORY)
if(TARGET clang-offload-packager)
@@ -56,39 +67,9 @@ endif()
set(LIBC_GPU_TEST_ARCHITECTURE "" CACHE STRING "Architecture for the GPU tests")
if(LIBC_TARGET_ARCHITECTURE_IS_AMDGPU)
- # Identify any locally installed AMD GPUs on the system using 'amdgpu-arch'.
- if(TARGET amdgpu-arch)
- get_target_property(LIBC_AMDGPU_ARCH amdgpu-arch LOCATION)
- else()
- find_program(LIBC_AMDGPU_ARCH
- NAMES amdgpu-arch NO_DEFAULT_PATH
- PATHS ${LLVM_BINARY_DIR}/bin ${compiler_path})
- endif()
- if(LIBC_AMDGPU_ARCH)
- execute_process(COMMAND ${LIBC_AMDGPU_ARCH}
- OUTPUT_VARIABLE arch_tool_output
- ERROR_QUIET OUTPUT_STRIP_TRAILING_WHITESPACE)
- if(arch_tool_output MATCHES "^gfx[0-9]+")
- set(PLATFORM_HAS_GPU TRUE)
- endif()
- endif()
+ check_cxx_compiler_flag(-mcpu=native PLATFORM_HAS_GPU)
elseif(LIBC_TARGET_ARCHITECTURE_IS_NVPTX)
- # Identify any locally installed NVIDIA GPUs on the system using 'nvptx-arch'.
- if(TARGET nvptx-arch)
- get_target_property(LIBC_NVPTX_ARCH nvptx-arch LOCATION)
- else()
- find_program(LIBC_NVPTX_ARCH
- NAMES nvptx-arch NO_DEFAULT_PATH
- PATHS ${LLVM_BINARY_DIR}/bin ${compiler_path})
- endif()
- if(LIBC_NVPTX_ARCH)
- execute_process(COMMAND ${LIBC_NVPTX_ARCH}
- OUTPUT_VARIABLE arch_tool_output
- ERROR_QUIET OUTPUT_STRIP_TRAILING_WHITESPACE)
- if(arch_tool_output MATCHES "^sm_[0-9]+")
- set(PLATFORM_HAS_GPU TRUE)
- endif()
- endif()
+ check_cxx_compiler_flag(-march=native PLATFORM_HAS_GPU)
endif()
set(gpu_test_architecture "")
diff --git a/llvm/runtimes/CMakeLists.txt b/llvm/runtimes/CMakeLists.txt
index e60c9dd6041b4..b6ed497fa122a 100644
--- a/llvm/runtimes/CMakeLists.txt
+++ b/llvm/runtimes/CMakeLists.txt
@@ -487,7 +487,7 @@ if(build_runtimes)
if(TARGET amdhsa-loader)
list(APPEND libc_cmake_args
"-DRUNTIMES_amdgcn-amd-amdhsa_LIBC_GPU_LOADER_EXECUTABLE=$<TARGET_FILE:amdhsa-loader>")
- list(APPEND extra_deps amdhsa-loader amdgpu-arch)
+ list(APPEND extra_deps amdhsa-loader)
endif()
list(APPEND libc_cmake_args "-DRUNTIMES_amdgcn-amd-amdhsa_LLVM_LIBC_FULL_BUILD=ON")
endif()
@@ -495,7 +495,7 @@ if(build_runtimes)
if(TARGET nvptx-loader)
list(APPEND libc_cmake_args
"-DRUNTIMES_nvptx64-nvidia-cuda_LIBC_GPU_LOADER_EXECUTABLE=$<TARGET_FILE:nvptx-loader>")
- list(APPEND extra_deps nvptx-loader nvptx-arch)
+ list(APPEND extra_deps nvptx-loader)
endif()
list(APPEND libc_cmake_args "-DRUNTIMES_nvptx64-nvidia-cuda_LLVM_LIBC_FULL_BUILD=ON")
endif()
|
lntue
approved these changes
Jun 13, 2024
Summary: This patch adds `CMAKE_REQUIRED_FLAGS` for the GPU build so checks like `check_cxx_compiler_flags` work as expected. This is required because we need to hack around the potential lack of `nvlink` and `ptxas` for NVPTX targets and the fact that the AMDGPU target needs `-nogpulib` to avoid errors on lack of ROCm. This makes a few of the checks pass and also allows us to just check `-mcpu=native` for architecture detection instead of finding the tools manually.
|
LLVM Buildbot has detected a new failure on builder Full details are available at: https://lab.llvm.org/buildbot/#/builders/163/builds/585 Here is the relevant piece of the build log for the reference: |
|
LLVM Buildbot has detected a new failure on builder Full details are available at: https://lab.llvm.org/buildbot/#/builders/123/builds/586 Here is the relevant piece of the build log for the reference: |
|
LLVM Buildbot has detected a new failure on builder Full details are available at: https://lab.llvm.org/buildbot/#/builders/18/builds/496 Here is the relevant piece of the build log for the reference: |
AlexisPerry
pushed a commit
to llvm-project-tlp/llvm-project
that referenced
this pull request
Jul 9, 2024
…lvm#95424) Summary: This patch adds `CMAKE_REQUIRED_FLAGS` for the GPU build so checks like `check_cxx_compiler_flags` work as expected. This is required because we need to hack around the potential lack of `nvlink` and `ptxas` for NVPTX targets and the fact that the AMDGPU target needs `-nogpulib` to avoid errors on lack of ROCm. This makes a few of the checks pass and also allows us to just check `-mcpu=native` for architecture detection instead of finding the tools manually.
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
Add this suggestion to a batch that can be applied as a single commit.
This suggestion is invalid because no changes were made to the code.
Suggestions cannot be applied while the pull request is closed.
Suggestions cannot be applied while viewing a subset of changes.
Only one suggestion per line can be applied in a batch.
Add this suggestion to a batch that can be applied as a single commit.
Applying suggestions on deleted lines is not supported.
You must change the existing code in this line in order to create a valid suggestion.
Outdated suggestions cannot be applied.
This suggestion has been applied or marked resolved.
Suggestions cannot be applied from pending reviews.
Suggestions cannot be applied on multi-line comments.
Suggestions cannot be applied while the pull request is queued to merge.
Suggestion cannot be applied right now. Please check back later.
Summary:
This patch adds
CMAKE_REQUIRED_FLAGSfor the GPU build so checks likecheck_cxx_compiler_flagswork as expected. This is required because weneed to hack around the potential lack of
nvlinkandptxasfor NVPTXtargets and the fact that the AMDGPU target needs
-nogpulibto avoiderrors on lack of ROCm. This makes a few of the checks pass and also
allows us to just check
-mcpu=nativefor architecture detectioninstead of finding the tools manually.