[libc-commits] [libc] [libc] Expose the AMDGPU code object version (PR #152931)

[via libc-commits]

llvmbot wrote:


<!--LLVM PR SUMMARY COMMENT-->

@llvm/pr-subscribers-libc

Author: Joseph Huber (jhuber6)

<details>
<summary>Changes</summary>

Summary:
People might want to modify this to run on older machines. Expose this
so it can be set with
`-DRUNTIMES_amdgcn-amd-amdhsa_LIBC_GPU_CODE_OBJECT_VERSION`.


---
Full diff: https://github.com/llvm/llvm-project/pull/152931.diff


1 Files Affected:

- (modified) libc/cmake/modules/prepare_libc_gpu_build.cmake (+4-6) 


``````````diff
diff --git a/libc/cmake/modules/prepare_libc_gpu_build.cmake b/libc/cmake/modules/prepare_libc_gpu_build.cmake
index f8f5a954e5e91..df1de1454c3eb 100644
--- a/libc/cmake/modules/prepare_libc_gpu_build.cmake
+++ b/libc/cmake/modules/prepare_libc_gpu_build.cmake
@@ -100,12 +100,10 @@ if(NOT TARGET libc.utils.gpu.loader AND gpu_loader_executable)
   )
 endif()
 
-if(LIBC_TARGET_ARCHITECTURE_IS_AMDGPU)
-  # The AMDGPU environment uses different code objects to encode the ABI for
-  # kernel calls and intrinsic functions. We want to specify this manually to
-  # conform to whatever the test suite was built to handle.
-  set(LIBC_GPU_CODE_OBJECT_VERSION 6)
-endif()
+# The AMDGPU environment uses different code objects to encode the ABI for
+# kernel calls and intrinsic functions. We want to expose this to conform to
+# whatever the test suite was built to handle.
+set(LIBC_GPU_CODE_OBJECT_VERSION "6" CACHE STRING "AMDGPU Code object ABI to use")
 
 if(LIBC_TARGET_ARCHITECTURE_IS_NVPTX)
   # FIXME: This is a hack required to keep the CUDA package from trying to find

``````````

</details>


https://github.com/llvm/llvm-project/pull/152931