| Issue |
184754
|
| Summary |
[cuda] Clang cuda compile don't support cooperative_groups::this_cluster()
|
| Labels |
cuda
|
| Assignees |
|
| Reporter |
woruyu
|
the code like this, https://godbolt.org/z/3G87YEPq3, godbolt shows that nvcc works fine! However, clang will report error `error: no member named 'this_cluster' in namespace 'cooperative_groups'`
```
// clang++ -x cuda --cuda-gpu-arch=sm_100 -S -emit-llvm --cuda-device-only test.cpp
#include <cooperative_groups.h>
__global__ void cluster_kernel(int *v)
{
namespace cg = cooperative_groups;
auto cluster = cg::this_cluster();
v[1] = cluster.size();
}
```
![]()
_______________________________________________
llvm-bugs mailing list
[email protected]
https://lists.llvm.org/cgi-bin/mailman/listinfo/llvm-bugs
