frozenleaves/pytorch
1
0
Fork 0
mirror of https://github.com/pytorch/pytorch.git synced 2025-10-20 12:54:11 +08:00
Code Issues Packages Projects Releases Wiki Activity

[CD] Do not enable GenAI on Windows (#163116)

Browse Source 
Follow up after https://github.com/pytorch/pytorch/pull/162209 as looks
like it causes some of the Windows builds to fail with
```
C:/actions-runner/_work/pytorch/pytorch/pytorch/third_party/fbgemm/fbgemm_gpu/experimental/gen_ai/src/quantize/common/include\fbgemm_gpu/quantize/utils.h(19): error C3861: '__builtin_clz': identifier not found
```

May be fixes https://github.com/pytorch/pytorch/issues/162881
Pull Request resolved: https://github.com/pytorch/pytorch/pull/163116
Approved by: https://github.com/wdvr, https://github.com/danielvegamyhre
This commit is contained in:
Nikita Shulga
2025-09-16 15:34:49 -07:00
committed by PyTorch MergeBot
parent bc38c5baa1
commit 6cfb080d84
1 changed files with 1 additions and 1 deletions
Download Patch File Download Diff File Expand all files Collapse all files

2
CMakeLists.txt
View File

@ -891,7 +891,7 @@ IF(USE_FBGEMM_GENAI AND USE_ROCM AND NOT "gfx942" IN_LIST PYTORCH_ROCM_ARCH)
endif()
# Set USE_FBGEMM_GENAI to ON for CUDA build on SM100.
if(USE_CUDA AND "$ENV{TORCH_CUDA_ARCH_LIST}" MATCHES "10.0" AND CMAKE_CUDA_COMPILER_VERSION VERSION_GREATER_EQUAL 12.8)
if(USE_CUDA AND "$ENV{TORCH_CUDA_ARCH_LIST}" MATCHES "10.0" AND CMAKE_CUDA_COMPILER_VERSION VERSION_GREATER_EQUAL 12.8 AND NOT WIN32)
message(STATUS "Setting USE_FBGEMM_GENAI to ON, doing CUDA build for SM100a")
set(USE_FBGEMM_GENAI ON)
endif()