Skip to content

GitLab

Commit 2593dd60 authored by turneram's avatar turneram
Browse files

Formatting

parent d1e27426
Hide whitespace changes
Inline Side-by-side
Showing with 8 additions and 8 deletions
src/targets/gpu/jit/ck_gemm.cpp
View file @ 2593dd60
......@@ -53,18 +53,18 @@ namespace gpu {
// extern "C" {
// __global__ void ck_gemm_kernel(void* a_p, void* b_p, void* c_p)
// __global__ void ck_gemm_kernel(void* a_p, void* b_p, void* c_p)
// {
// // hipDeviceProp_t hdp{};
// // printf("Shared mem: %i\n", int(hdp.sharedMemPerBlock));
// // make_tensors()(a_p, b_p, c_p)([](auto&&... xs) {
// // ck_gemm(xs...);
// // make_tensors()(a_p, b_p, c_p)([](auto&&... xs) {
// // ck_gemm(xs...);
// // });
// make_tensors()(a_p, b_p, c_p)([](auto a_t, auto b_t, auto c_t) {
// __shared__ float p_shared_block[512]; //[(a_t.get_shape().elements() + b_t.get_shape().elements()) * 2];
// ck_gemm(a_t, b_t, c_t, p_shared_block);
// make_tensors()(a_p, b_p, c_p)([](auto a_t, auto b_t, auto c_t) {
// __shared__ float p_shared_block[512]; //[(a_t.get_shape().elements() +
// b_t.get_shape().elements()) * 2]; ck_gemm(a_t, b_t, c_t, p_shared_block);
// // make_tensors()(p_shared_block)([&](auto p_t) {
// // ck_gemm(a_t, b_t, c_t, p_t);
// // ck_gemm(a_t, b_t, c_t, p_t);
// // });
// });
// }
......
src/targets/gpu/kernels/include/migraphx/kernels/ck_gemm.hpp
View file @ 2593dd60
......@@ -283,7 +283,7 @@ __device__ void ck_gemm(const T& a_t, const U& b_t, const V& c_t, float* p_t)
constexpr bool HasMainKBlockLoop = true;
constexpr bool HasDoubleTailKBlockLoop = true;
auto num_bytes = GridwiseGemm::GetSharedMemoryNumberOfByte();
auto num_bytes = GridwiseGemm::GetSharedMemoryNumberOfByte();
printf("Bytes: %i\n", int(num_bytes));
GridwiseGemm::Run(a_t.data(),
b_t.data(),
......
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