frozenleaves/pytorch
1
0
Fork 0
mirror of https://github.com/pytorch/pytorch.git synced 2025-10-20 21:14:14 +08:00
Code Issues Packages Projects Releases Wiki Activity

[symm_mem] Update CMakeList to reflect code moving a dedicated folder (#155823)

Browse Source 
We moved all symm_mem code into a folder ([CudaDMAConnectivity](https://github.com/pytorch/pytorch/pull/155573)) but somehow forgot update for CudaDMAConnectivity in the CMakeList.

Users see errors: RuntimeError: DMA connectivity detector for cuda over nvlink is not available while torch.distributed.init_process_group(backend=backend). So this PR should fix it.
Pull Request resolved: https://github.com/pytorch/pytorch/pull/155823
Approved by: https://github.com/Skylion007
This commit is contained in:
fduwjj
2025-06-12 21:33:01 +00:00
committed by PyTorch MergeBot
parent 70bb34929a
commit a87dfc7480
1 changed files with 1 additions and 1 deletions
Download Patch File Download Diff File Expand all files Collapse all files

2
caffe2/CMakeLists.txt
View File

@ -572,10 +572,10 @@ if(USE_CUDA)
if(NOT WIN32)
append_filelist("libtorch_cuda_distributed_extra_sources" Caffe2_GPU_SRCS)
set_source_files_properties(
${TORCH_SRC_DIR}/csrc/distributed/c10d/CudaDMAConnectivity.cpp
${TORCH_SRC_DIR}/csrc/distributed/c10d/ProcessGroupNCCL.cpp
${TORCH_SRC_DIR}/csrc/distributed/c10d/cuda/utils.cpp
${TORCH_SRC_DIR}/csrc/distributed/c10d/intra_node_comm.cpp
${TORCH_SRC_DIR}/csrc/distributed/c10d/symm_mem/CudaDMAConnectivity.cpp
${TORCH_SRC_DIR}/csrc/distributed/c10d/symm_mem/CUDASymmetricMemory.cu
${TORCH_SRC_DIR}/csrc/distributed/c10d/symm_mem/CUDASymmetricMemoryOps.cu
${TORCH_SRC_DIR}/csrc/distributed/c10d/symm_mem/CUDASymmetricMemoryUtils.cpp