We read every piece of feedback, and take your input very seriously.
To see all available qualifiers, see our documentation.
There was an error while loading. Please reload this page.
1 parent 9fcc4d8 commit 7af61dbCopy full SHA for 7af61db
csrc/cuda/utils.cuh
@@ -6,13 +6,10 @@
6
AT_ASSERTM(x.device().is_cuda(), #x " must be CUDA tensor")
7
#define CHECK_INPUT(x) AT_ASSERTM(x, "Input mismatch")
8
9
-#ifndef _WIN32
10
-__device__ __inline__ at::Half __shfl_sync(const unsigned mask,
11
- const at::Half var,
12
- const unsigned int srcLane) {
+__device__ __inline__ at::Half
+__shfl_sync(const unsigned mask, const at::Half var, const int srcLane) {
13
return __shfl_sync(mask, (__half)var, srcLane);
14
}
15
-#endif
16
17
__device__ __inline__ at::Half __shfl_down_sync(const unsigned mask,
18
const at::Half var,
0 commit comments