mirror of
https://github.com/ROCm/composable_kernel.git
synced 2026-05-13 09:45:56 +00:00
implicit gemm v1r2: only load 1d filter
This commit is contained in:
@@ -12,7 +12,7 @@ __device__ void threadwise_4d_tensor_pointwise_operation_unary(Desc, Float* __re
|
||||
constexpr auto desc = Desc{};
|
||||
|
||||
#if 0
|
||||
if(threadIdx.x == 0)
|
||||
if(get_thread_local_1d_id() == 0)
|
||||
{
|
||||
print_ConstantTensorDescriptor(desc, "threadwise_4d_tensor_op_unary: ");
|
||||
}
|
||||
@@ -218,7 +218,7 @@ __device__ void threadwise_4d_tensor_shift_down(Desc, Float* __restrict__ p, IDi
|
||||
constexpr auto desc = Desc{};
|
||||
|
||||
#if 0
|
||||
if(threadIdx.x == 0)
|
||||
if(get_thread_local_1d_id() == 0)
|
||||
{
|
||||
print_ConstantTensorDescriptor(desc, "threadwise_4d_tensor_shift_down: ");
|
||||
}
|
||||
|
||||
Reference in New Issue
Block a user