biondizzle/vllm
1
0
Fork 0
Code Issues Pull Requests Actions 2 Packages Projects Releases Wiki Activity

fix CUDA_check redefinition in #17918 (#18287)

Browse Source 
Signed-off-by: Lucia Fang <fanglu@fb.com>
Co-authored-by: Lucia (Lu) Fang <fanglu@meta.com>
This commit is contained in:
Lucia Fang
2025-05-19 13:42:35 -07:00
committed by GitHub
parent dc1440cf9f
commit 258bf621d5
2 changed files with 5 additions and 12 deletions
Download Patch File Download Diff File Expand all files Collapse all files

9
csrc/cutlass_extensions/common.hpp
View File

@@ -15,15 +15,6 @@
cutlassGetStatusString(error)); \
}
/**
* Panic wrapper for unwinding CUDA runtime errors
*/
#define CUDA_CHECK(status) \
{ \
cudaError_t error = status; \
TORCH_CHECK(error == cudaSuccess, cudaGetErrorString(error)); \
}
inline int get_cuda_max_shared_memory_per_block_opt_in(int const device) {
int max_shared_mem_per_block_opt_in = 0;
cudaDeviceGetAttribute(&max_shared_mem_per_block_opt_in,

2
csrc/sparse/cutlass/sparse_scaled_mm_c3x.cuh
View File

@@ -8,6 +8,8 @@
#include <ATen/cuda/CUDAContext.h>
#include "cuda_utils.h"
#include "cutlass/cutlass.h"
#include "cutlass/gemm/device/gemm_universal_adapter.h"