Skip to content

GitLab

Commit a43eabbc authored by fengzch's avatar fengzch
Browse files

fix: asm movmatrix.sync.aligned.m8n8.trans.b16 is not supported

parent 0343f43e
Hide whitespace changes
Inline Side-by-side
Showing with 4 additions and 3 deletions
src/kernels/zgemm/gemm_utils.cuh
View file @ a43eabbc
...@@ -199,9 +199,10 @@ __device__ __forceinline__ static void ldmatrix(const void *ptr, uint4 &out) { ...@@ -199,9 +199,10 @@ __device__ __forceinline__ static void ldmatrix(const void *ptr, uint4 &out) {
template<typename T> template<typename T>
__device__ __forceinline__ static T movmatrix(T x) { __device__ __forceinline__ static T movmatrix(T x) {
asm volatile("movmatrix.sync.aligned.m8n8.trans.b16 %0, %1;" // asm volatile("movmatrix.sync.aligned.m8n8.trans.b16 %0, %1;"
: "=r"(*reinterpret_cast<uint32_t *>(&x)) // : "=r"(*reinterpret_cast<uint32_t *>(&x))
: "r"(*reinterpret_cast<uint32_t *>(&x))); // : "r"(*reinterpret_cast<uint32_t *>(&x)));
printf("%s: asm movmatrix.sync.aligned.m8n8.trans.b16 is not supported in HIP yet!\n", __func__);
return x; return x;
} }
......
Markdown is supported
0% or .
You are about to add 0 people to the discussion. Proceed with caution.
Finish editing this message first!
Please register or to comment