Absolute include path (#281)

* ad gelu and fast_gelu * added GeLU and fast GeLU * clean up * add gemm+fastgelu example * add gemm+gelu instances * update profiler * clean up * clean up * adding gemm+bias+activation * clean * adding bias * clean * adding gemm multiple d * debugging * add gemm bias add fastgelu * rename, clean * refactoring; add readme * refactor * refactor * refactor * refactor * refactor * refactor * fix * fix * update example * update example * rename * update example * add ckProfiler * clean * clean * clean * clean * add client app example * update readme * delete obselete files * remove old client app * delete old file * cleaning * clean * remove half * fix header path * fix header path * fix header path * fix header path * fix header path * fix header path for all examples * fix header path * fix header path * fix header path * fix header path * fix header path * fix header path * fix header path * fix header path * fix header path * revert client app example * clean build * fix build * temporary disable client test on Jenkins * clean * clean * clean
2026-05-03 13:11:25 +00:00 · 2022-06-24 20:51:04 -05:00
parent a49115b95e
commit d1db6a0c3e
499 changed files with 3000 additions and 24130 deletions
--- a/include/ck/tensor_operation/gpu/block/blockwise_gemm_dl_v2r3.hpp
+++ b/include/ck/tensor_operation/gpu/block/blockwise_gemm_dl_v2r3.hpp
@@ -1,8 +1,9 @@
 #pragma once
-#include "common_header.hpp"
-#include "tensor_adaptor.hpp"
-#include "threadwise_tensor_slice_transfer_v4r1.hpp"
-#include "threadwise_contraction_dl.hpp"
+
+#include "ck/utility/common_header.hpp"
+#include "ck/tensor_description/tensor_adaptor.hpp"
+#include "ck/tensor_operation/gpu/thread/threadwise_tensor_slice_transfer_v4r1.hpp"
+#include "ck/tensor_operation/gpu/thread/threadwise_contraction_dl.hpp"

 namespace ck {

--- a/include/ck/tensor_operation/gpu/block/blockwise_gemm_xdlops.hpp
+++ b/include/ck/tensor_operation/gpu/block/blockwise_gemm_xdlops.hpp
@@ -1,9 +1,9 @@
 #pragma once
-#include "common_header.hpp"
-#include "threadwise_tensor_slice_transfer.hpp"
-#include "xdlops_gemm.hpp"
-#include "tensor_adaptor.hpp"
-#include "thread_group.hpp"
+
+#include "ck/utility/common_header.hpp"
+#include "ck/tensor_operation/gpu/thread/threadwise_tensor_slice_transfer.hpp"
+#include "ck/tensor_operation/gpu/warp/xdlops_gemm.hpp"
+#include "ck/tensor_description/tensor_adaptor.hpp"

 namespace ck {

--- a/include/ck/tensor_operation/gpu/block/blockwise_tensor_slice_transfer_v5r1.hpp
+++ b/include/ck/tensor_operation/gpu/block/blockwise_tensor_slice_transfer_v5r1.hpp
@@ -1,11 +1,10 @@
-#ifndef CK_BLOCKWISE_TENSOR_SLICE_TRANSFER_V5R1_HPP
-#define CK_BLOCKWISE_TENSOR_SLICE_TRANSFER_V5R1_HPP
+#pragma once

-#include "common_header.hpp"
-#include "tensor_descriptor.hpp"
-#include "tensor_descriptor_helper.hpp"
-#include "cluster_descriptor.hpp"
-#include "threadwise_tensor_slice_transfer_v5r1.hpp"
+#include "ck/utility/common_header.hpp"
+#include "ck/tensor_description/tensor_descriptor.hpp"
+#include "ck/tensor_description/tensor_descriptor_helper.hpp"
+#include "ck/tensor_description/cluster_descriptor.hpp"
+#include "ck/tensor_operation/gpu/thread/threadwise_tensor_slice_transfer_v5r1.hpp"

 namespace ck {

@@ -152,4 +151,3 @@ struct BlockwiseTensorSliceTransfer_v5r1
 };

 } // namespace ck
-#endif
--- a/include/ck/tensor_operation/gpu/block/reduction_functions_blockwise.hpp
+++ b/include/ck/tensor_operation/gpu/block/reduction_functions_blockwise.hpp
@@ -1,35 +1,8 @@
-/*******************************************************************************
- *
- * MIT License
- *
- * Copyright (c) 2020 Advanced Micro Devices, Inc.
- *
- * Permission is hereby granted, free of charge, to any person obtaining a copy
- * of this software and associated documentation files (the "Software"), to deal
- * in the Software without restriction, including without limitation the rights
- * to use, copy, modify, merge, publish, distribute, sublicense, and/or sell
- * copies of the Software, and to permit persons to whom the Software is
- * furnished to do so, subject to the following conditions:
- *
- * The above copyright notice and this permission notice shall be included in all
- * copies or substantial portions of the Software.
- *
- * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
- * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
- * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE
- * AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER
- * LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM,
- * OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN THE
- * SOFTWARE.
- *
- *******************************************************************************/
-#ifndef CK_REDUCTION_FUNCTIONS_BLOCKWISE_HPP
-#define CK_REDUCTION_FUNCTIONS_BLOCKWISE_HPP
+#pragma once

-#include "reduction_common.hpp"
-#include "reduction_functions_accumulate.hpp"
-
-#include "cluster_descriptor.hpp"
+#include "ck/tensor_description/cluster_descriptor.hpp"
+#include "ck/utility/reduction_common.hpp"
+#include "ck/utility/reduction_functions_accumulate.hpp"

 namespace ck {

@@ -193,6 +166,4 @@ struct PartitionedBlockwiseReductionWithIndex
    };
 };

-}; // end of namespace ck
-
-#endif
+} // namespace ck
--- a/include/ck/tensor_operation/gpu/block/thread_group_tensor_slice_transfer_v4r1.hpp
+++ b/include/ck/tensor_operation/gpu/block/thread_group_tensor_slice_transfer_v4r1.hpp
@@ -1,9 +1,10 @@
 #pragma once
-#include "common_header.hpp"
-#include "tensor_descriptor.hpp"
-#include "tensor_descriptor_helper.hpp"
-#include "cluster_descriptor.hpp"
-#include "threadwise_tensor_slice_transfer_v3r1.hpp"
+
+#include "ck/utility/common_header.hpp"
+#include "ck/tensor_description/tensor_descriptor.hpp"
+#include "ck/tensor_description/tensor_descriptor_helper.hpp"
+#include "ck/tensor_description/cluster_descriptor.hpp"
+#include "ck/tensor_operation/gpu/thread/threadwise_tensor_slice_transfer_v3r1.hpp"

 namespace ck {

--- a/include/ck/tensor_operation/gpu/block/thread_group_tensor_slice_transfer_v6r1.hpp
+++ b/include/ck/tensor_operation/gpu/block/thread_group_tensor_slice_transfer_v6r1.hpp
@@ -1,9 +1,10 @@
 #pragma once
-#include "common_header.hpp"
-#include "tensor_descriptor.hpp"
-#include "tensor_descriptor_helper.hpp"
-#include "cluster_descriptor.hpp"
-#include "threadwise_tensor_slice_transfer_v6r1.hpp"
+
+#include "ck/utility/common_header.hpp"
+#include "ck/tensor_description/tensor_descriptor.hpp"
+#include "ck/tensor_description/tensor_descriptor_helper.hpp"
+#include "ck/tensor_description/cluster_descriptor.hpp"
+#include "ck/tensor_operation/gpu/thread/threadwise_tensor_slice_transfer_v6r1.hpp"

 namespace ck {

--- a/include/ck/tensor_operation/gpu/block/thread_group_tensor_slice_transfer_v6r2.hpp
+++ b/include/ck/tensor_operation/gpu/block/thread_group_tensor_slice_transfer_v6r2.hpp
@@ -1,9 +1,10 @@
 #pragma once
-#include "common_header.hpp"
-#include "tensor_descriptor.hpp"
-#include "tensor_descriptor_helper.hpp"
-#include "cluster_descriptor.hpp"
-#include "threadwise_tensor_slice_transfer_v6r2.hpp"
+
+#include "ck/utility/common_header.hpp"
+#include "ck/tensor_description/tensor_descriptor.hpp"
+#include "ck/tensor_description/tensor_descriptor_helper.hpp"
+#include "ck/tensor_description/cluster_descriptor.hpp"
+#include "ck/tensor_operation/gpu/thread/threadwise_tensor_slice_transfer_v6r2.hpp"

 namespace ck {

--- a/include/ck/tensor_operation/gpu/block/thread_group_tensor_slice_transfer_v6r3.hpp
+++ b/include/ck/tensor_operation/gpu/block/thread_group_tensor_slice_transfer_v6r3.hpp
@@ -1,9 +1,10 @@
 #pragma once
-#include "common_header.hpp"
-#include "tensor_descriptor.hpp"
-#include "tensor_descriptor_helper.hpp"
-#include "cluster_descriptor.hpp"
-#include "threadwise_tensor_slice_transfer_v6r3.hpp"
+
+#include "ck/utility/common_header.hpp"
+#include "ck/tensor_description/tensor_descriptor.hpp"
+#include "ck/tensor_description/tensor_descriptor_helper.hpp"
+#include "ck/tensor_description/cluster_descriptor.hpp"
+#include "ck/tensor_operation/gpu/thread/threadwise_tensor_slice_transfer_v6r3.hpp"

 namespace ck {

--- a/include/ck/tensor_operation/gpu/block/thread_group_tensor_slice_transfer_v7.hpp
+++ b/include/ck/tensor_operation/gpu/block/thread_group_tensor_slice_transfer_v7.hpp
@@ -1,10 +1,10 @@
 #pragma once

-#include "common_header.hpp"
-#include "tensor_descriptor.hpp"
-#include "tensor_descriptor_helper.hpp"
-#include "cluster_descriptor.hpp"
-#include "threadwise_tensor_slice_transfer_v7.hpp"
+#include "ck/utility/common_header.hpp"
+#include "ck/tensor_description/tensor_descriptor.hpp"
+#include "ck/tensor_description/tensor_descriptor_helper.hpp"
+#include "ck/tensor_description/cluster_descriptor.hpp"
+#include "ck/tensor_operation/gpu/thread/threadwise_tensor_slice_transfer_v7.hpp"

 namespace ck {

--- a/include/ck/tensor_operation/gpu/device/device_5ary_elementwise.hpp
+++ b/include/ck/tensor_operation/gpu/device/device_5ary_elementwise.hpp
@@ -1,12 +1,16 @@
 #pragma once
+
 #include <iostream>
 #include <sstream>
-#include "device.hpp"
-#include "device_base.hpp"
-#include "gridwise_5ary_Elementwise_1d.hpp"
-#include "tensor_layout.hpp"
-#include "tensor_descriptor.hpp"
-#include "tensor_descriptor_helper.hpp"
+#include <vector>
+
+#include "ck/utility/common_header.hpp"
+#include "ck/tensor_description/tensor_descriptor.hpp"
+#include "ck/tensor_description/tensor_descriptor_helper.hpp"
+#include "ck/tensor_operation/gpu/device/device_base.hpp"
+#include "ck/tensor_operation/gpu/grid/gridwise_5ary_Elementwise_1d.hpp"
+#include "ck/device_utility/device_prop.hpp"
+#include "ck/device_utility/kernel_launch.hpp"

 namespace ck {
 namespace tensor_operation {
@@ -325,7 +329,7 @@ struct Device5AryElementwise : public BaseOperator

    static auto MakeInvoker() { return Invoker{}; }
    std::unique_ptr<BaseInvoker> MakeInvokerPointer() { return std::make_unique<Invoker>(); }
-}; // namespace device
+};

 } // namespace device
 } // namespace tensor_operation
--- a/include/ck/tensor_operation/gpu/device/device_base.hpp
+++ b/include/ck/tensor_operation/gpu/device/device_base.hpp
@@ -2,7 +2,7 @@

 #include <string>

-#include "stream_config.hpp"
+#include "ck/stream_config.hpp"

 namespace ck {
 namespace tensor_operation {
--- a/include/ck/tensor_operation/gpu/device/device_batched_gemm_reduce_xdl_cshuffle.hpp
+++ b/include/ck/tensor_operation/gpu/device/device_batched_gemm_reduce_xdl_cshuffle.hpp
@@ -1,14 +1,17 @@
 #pragma once
+
 #include <iostream>
 #include <sstream>
-#include "device.hpp"
-#include "device_gemm_reduce.hpp"
-#include "common_header.hpp"
-#include "tensor_layout.hpp"
-#include "tensor_descriptor.hpp"
-#include "tensor_descriptor_helper.hpp"
-#include "gridwise_gemm_reduce_xdl_cshuffle_v1.hpp"
-#include "gemm_specialization.hpp"
+
+#include "ck/utility/common_header.hpp"
+#include "ck/tensor_description/tensor_descriptor.hpp"
+#include "ck/tensor_description/tensor_descriptor_helper.hpp"
+#include "ck/tensor_operation/gpu/device/tensor_layout.hpp"
+#include "ck/tensor_operation/gpu/device/device_gemm_reduce.hpp"
+#include "ck/tensor_operation/gpu/device/gemm_specialization.hpp"
+#include "ck/tensor_operation/gpu/grid/gridwise_gemm_reduce_xdl_cshuffle_v1.hpp"
+#include "ck/device_utility/device_prop.hpp"
+#include "ck/device_utility/kernel_launch.hpp"

 namespace ck {
 namespace tensor_operation {
--- a/include/ck/tensor_operation/gpu/device/device_batched_gemm_xdl.hpp
+++ b/include/ck/tensor_operation/gpu/device/device_batched_gemm_xdl.hpp
@@ -1,16 +1,17 @@
-#ifndef DEVICE_BATCHED_GEMM_XDL_HPP
-#define DEVICE_BATCHED_GEMM_XDL_HPP
+#pragma once

 #include <iostream>
 #include <sstream>
-#include "device.hpp"
-#include "device_base.hpp"
-#include "device_gemm.hpp"
-#include "common_header.hpp"
-#include "tensor_layout.hpp"
-#include "tensor_descriptor.hpp"
-#include "tensor_descriptor_helper.hpp"
-#include "gridwise_gemm_xdlops_v2r3.hpp"
+
+#include "ck/utility/common_header.hpp"
+#include "ck/tensor_description/tensor_descriptor.hpp"
+#include "ck/tensor_description/tensor_descriptor_helper.hpp"
+#include "ck/tensor_operation/gpu/device/tensor_layout.hpp"
+#include "ck/tensor_operation/gpu/device/device_gemm.hpp"
+#include "ck/tensor_operation/gpu/device/gemm_specialization.hpp"
+#include "ck/tensor_operation/gpu/grid/gridwise_gemm_xdlops_v2r3.hpp"
+#include "ck/device_utility/device_prop.hpp"
+#include "ck/device_utility/kernel_launch.hpp"

 namespace ck {
 namespace tensor_operation {
@@ -616,4 +617,3 @@ struct DeviceBatchedGemmXdl
 } // namespace device
 } // namespace tensor_operation
 } // namespace ck
-#endif
--- a/include/ck/tensor_operation/gpu/device/device_binary_elementwise.hpp
+++ b/include/ck/tensor_operation/gpu/device/device_binary_elementwise.hpp
@@ -1,10 +1,12 @@
 #pragma once
+
 #include <iostream>
 #include <vector>

-#include "device.hpp"
-#include "device_base.hpp"
-#include "gridwise_binary_elementwise_1d.hpp"
+#include "ck/device_utility/device_prop.hpp"
+#include "ck/device_utility/kernel_launch.hpp"
+#include "ck/tensor_operation/gpu/device/device_base.hpp"
+#include "ck/tensor_operation/gpu/grid/gridwise_binary_elementwise_1d.hpp"

 namespace ck {
 namespace tensor_operation {
--- a/include/ck/tensor_operation/gpu/device/device_cgemm_4gemm_xdl_cshuffle.hpp
+++ b/include/ck/tensor_operation/gpu/device/device_cgemm_4gemm_xdl_cshuffle.hpp
@@ -1,42 +1,20 @@
-/*******************************************************************************
- *
- * MIT License
- *
- * Copyright (c) 2022 Advanced Micro Devices, Inc.
- *
- * Permission is hereby granted, free of charge, to any person obtaining a copy
- * of this software and associated documentation files (the "Software"), to deal
- * in the Software without restriction, including without limitation the rights
- * to use, copy, modify, merge, publish, distribute, sublicense, and/or sell
- * copies of the Software, and to permit persons to whom the Software is
- * furnished to do so, subject to the following conditions:
- *
- * The above copyright notice and this permission notice shall be included in all
- * copies or substantial portions of the Software.
- *
- * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
- * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
- * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE
- * AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER
- * LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM,
- * OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN THE
- * SOFTWARE.
- *
- *******************************************************************************/
 #pragma once
+
 #include <iostream>
 #include <sstream>
-#include "device.hpp"
-#include "device_gemm.hpp"
-#include "device_cgemm.hpp"
-#include "common_header.hpp"
-#include "tensor_layout.hpp"
-#include "tensor_descriptor.hpp"
-#include "tensor_descriptor_helper.hpp"
-#include "gridwise_gemm_xdl_cshuffle_v1.hpp"
-#include "binary_element_wise_operation.hpp"
-#include "gridwise_binary_elementwise_1d.hpp"
-#include "tensor_operation/gpu/device/gemm_specialization.hpp"
+
+#include "ck/utility/common_header.hpp"
+#include "ck/tensor_description/tensor_descriptor.hpp"
+#include "ck/tensor_description/tensor_descriptor_helper.hpp"
+#include "ck/tensor_operation/gpu/device/tensor_layout.hpp"
+#include "ck/tensor_operation/gpu/device/device_gemm.hpp"
+#include "ck/tensor_operation/gpu/device/device_cgemm.hpp"
+#include "ck/tensor_operation/gpu/device/gemm_specialization.hpp"
+#include "ck/tensor_operation/gpu/grid/gridwise_gemm_xdl_cshuffle_v1.hpp"
+#include "ck/tensor_operation/gpu/grid/gridwise_binary_elementwise_1d.hpp"
+#include "ck/tensor_operation/gpu/element/binary_element_wise_operation.hpp"
+#include "ck/device_utility/device_prop.hpp"
+#include "ck/device_utility/kernel_launch.hpp"

 namespace ck {
 namespace tensor_operation {
--- a/include/ck/tensor_operation/gpu/device/device_conv2d_backward_weight_xdl_c_shuffle_nhwc_kyxc_nhwk.hpp
+++ b/include/ck/tensor_operation/gpu/device/device_conv2d_backward_weight_xdl_c_shuffle_nhwc_kyxc_nhwk.hpp
@@ -1,17 +1,18 @@
-#ifndef DEVICE_CONV2D_WRW_XDL_C_SHUFFLE_NHWC_KYXC_NHWK_HPP
-#define DEVICE_CONV2D_WRW_XDL_C_SHUFFLE_NHWC_KYXC_NHWK_HPP
+#pragma once

 #include <iostream>
 #include <sstream>
-#include "device.hpp"
-#include "device_base.hpp"
-#include "device_conv_backward_weight.hpp"
-#include "convolution_forward_specialization.hpp"
-#include "common_header.hpp"
-#include "tensor_layout.hpp"
-#include "tensor_descriptor.hpp"
-#include "tensor_descriptor_helper.hpp"
-#include "gridwise_gemm_xdlops_bwd_weight.hpp"
+
+#include "ck/utility/common_header.hpp"
+#include "ck/tensor_description/tensor_descriptor.hpp"
+#include "ck/tensor_description/tensor_descriptor_helper.hpp"
+#include "ck/tensor_operation/gpu/device/tensor_layout.hpp"
+#include "ck/tensor_operation/gpu/device/device_conv_backward_weight.hpp"
+#include "ck/tensor_operation/gpu/device/convolution_backward_weight_specialization.hpp"
+#include "ck/tensor_operation/gpu/grid/gridwise_gemm_xdlops_bwd_weight.hpp"
+#include "ck/tensor_operation/gpu/grid/gridwise_unary_elementwise_1d.hpp"
+#include "ck/device_utility/device_prop.hpp"
+#include "ck/device_utility/kernel_launch.hpp"

 namespace ck {
 namespace tensor_operation {
@@ -773,4 +774,3 @@ struct DeviceConv2dBwdWeightXdl_C_Shuffle_Input_N_Hi_Wi_C_Weight_K_Y_X_C_Output_
 } // namespace device
 } // namespace tensor_operation
 } // namespace ck
-#endif
--- a/include/ck/tensor_operation/gpu/device/device_conv2d_bwd_data_xdl_nhwc_kyxc_nhwk.hpp
+++ b/include/ck/tensor_operation/gpu/device/device_conv2d_bwd_data_xdl_nhwc_kyxc_nhwk.hpp
@@ -1,17 +1,17 @@
-#ifndef DEVICE_CONV2D_BWD_DATA_XDL_NHWC_KYXC_NHWK_HPP
-#define DEVICE_CONV2D_BWD_DATA_XDL_NHWC_KYXC_NHWK_HPP
+#pragma once

 #include <iostream>
 #include <sstream>
-#include "device.hpp"
-#include "device_base.hpp"
-#include "device_conv_bwd_data.hpp"
-#include "convolution_backward_data_specialization.hpp"
-#include "common_header.hpp"
-#include "tensor_layout.hpp"
-#include "tensor_descriptor.hpp"
-#include "tensor_descriptor_helper.hpp"
-#include "gridwise_gemm_xdlops_v2r3.hpp"
+
+#include "ck/utility/common_header.hpp"
+#include "ck/tensor_description/tensor_descriptor.hpp"
+#include "ck/tensor_description/tensor_descriptor_helper.hpp"
+#include "ck/tensor_operation/gpu/device/tensor_layout.hpp"
+#include "ck/tensor_operation/gpu/device/device_conv_bwd_data.hpp"
+#include "ck/tensor_operation/gpu/device/convolution_backward_data_specialization.hpp"
+#include "ck/tensor_operation/gpu/grid/gridwise_gemm_xdlops_v2r3.hpp"
+#include "ck/device_utility/device_prop.hpp"
+#include "ck/device_utility/kernel_launch.hpp"

 namespace ck {
 namespace tensor_operation {
@@ -821,4 +821,3 @@ struct DeviceConv2dBwdDataXdl_Input_N_Hi_Wi_C_Weight_K_Y_X_C_Output_N_Ho_Wo_K
 } // namespace device
 } // namespace tensor_operation
 } // namespace ck
-#endif
--- a/include/ck/tensor_operation/gpu/device/device_conv2d_fwd_xdl_c_shuffle_bias_activation_add_nhwc_kyxc_nhwk.hpp
+++ b/include/ck/tensor_operation/gpu/device/device_conv2d_fwd_xdl_c_shuffle_bias_activation_add_nhwc_kyxc_nhwk.hpp
@@ -1,17 +1,17 @@
-#ifndef DEVICE_CONV2D_FWD_XDL_C_SHUFFLE_BIAS_ACTIVATION_ADD_NHWC_KYXC_NHWK_HPP
-#define DEVICE_CONV2D_FWD_XDL_C_SHUFFLE_BIAS_ACTIVATION_ADD_NHWC_KYXC_NHWK_HPP
+#pragma once

 #include <iostream>
 #include <sstream>
-#include "device.hpp"
-#include "device_base.hpp"
-#include "device_conv_fwd_bias_activation_add.hpp"
-#include "convolution_forward_specialization.hpp"
-#include "common_header.hpp"
-#include "tensor_layout.hpp"
-#include "tensor_descriptor.hpp"
-#include "tensor_descriptor_helper.hpp"
-#include "gridwise_gemm_xdlops_v3r3.hpp"
+
+#include "ck/utility/common_header.hpp"
+#include "ck/tensor_description/tensor_descriptor.hpp"
+#include "ck/tensor_description/tensor_descriptor_helper.hpp"
+#include "ck/tensor_operation/gpu/device/tensor_layout.hpp"
+#include "ck/tensor_operation/gpu/device/device_conv_fwd_bias_activation_add.hpp"
+#include "ck/tensor_operation/gpu/device/convolution_forward_specialization.hpp"
+#include "ck/tensor_operation/gpu/grid/gridwise_gemm_xdlops_v3r3.hpp"
+#include "ck/device_utility/device_prop.hpp"
+#include "ck/device_utility/kernel_launch.hpp"

 namespace ck {
 namespace tensor_operation {
@@ -963,4 +963,3 @@ struct
 } // namespace device
 } // namespace tensor_operation
 } // namespace ck
-#endif
--- a/include/ck/tensor_operation/gpu/device/device_conv2d_fwd_xdl_c_shuffle_bias_activation_nhwc_kyxc_nhwk.hpp
+++ b/include/ck/tensor_operation/gpu/device/device_conv2d_fwd_xdl_c_shuffle_bias_activation_nhwc_kyxc_nhwk.hpp
@@ -1,15 +1,18 @@
 #pragma once
+
 #include <iostream>
 #include <sstream>
-#include "device.hpp"
-#include "device_base.hpp"
-#include "device_conv_fwd_bias_activation.hpp"
-#include "convolution_forward_specialization.hpp"
-#include "common_header.hpp"
-#include "tensor_layout.hpp"
-#include "tensor_descriptor.hpp"
-#include "tensor_descriptor_helper.hpp"
-#include "gridwise_gemm_xdlops_v3r2.hpp"
+#include <vector>
+
+#include "ck/utility/common_header.hpp"
+#include "ck/tensor_description/tensor_descriptor.hpp"
+#include "ck/tensor_description/tensor_descriptor_helper.hpp"
+#include "ck/tensor_operation/gpu/device/tensor_layout.hpp"
+#include "ck/tensor_operation/gpu/device/device_conv_fwd_bias_activation.hpp"
+#include "ck/tensor_operation/gpu/device/convolution_forward_specialization.hpp"
+#include "ck/tensor_operation/gpu/grid/gridwise_gemm_xdlops_v3r2.hpp"
+#include "ck/device_utility/device_prop.hpp"
+#include "ck/device_utility/kernel_launch.hpp"

 namespace ck {
 namespace tensor_operation {
--- a/include/ck/tensor_operation/gpu/device/device_conv2d_fwd_xdl_c_shuffle_nhwc_kyxc_nhwk.hpp
+++ b/include/ck/tensor_operation/gpu/device/device_conv2d_fwd_xdl_c_shuffle_nhwc_kyxc_nhwk.hpp
@@ -1,17 +1,17 @@
-#ifndef DEVICE_CONV2D_FWD_XDL_C_SHUFFLE_NHWC_KYXC_NHWK_HPP
-#define DEVICE_CONV2D_FWD_XDL_C_SHUFFLE_NHWC_KYXC_NHWK_HPP
+#pragma once

 #include <iostream>
 #include <sstream>
-#include "device.hpp"
-#include "device_base.hpp"
-#include "device_conv_fwd.hpp"
-#include "convolution_forward_specialization.hpp"
-#include "common_header.hpp"
-#include "tensor_layout.hpp"
-#include "tensor_descriptor.hpp"
-#include "tensor_descriptor_helper.hpp"
-#include "gridwise_gemm_xdlops_v3r1.hpp"
+
+#include "ck/utility/common_header.hpp"
+#include "ck/tensor_description/tensor_descriptor.hpp"
+#include "ck/tensor_description/tensor_descriptor_helper.hpp"
+#include "ck/tensor_operation/gpu/device/tensor_layout.hpp"
+#include "ck/tensor_operation/gpu/device/device_conv_fwd.hpp"
+#include "ck/tensor_operation/gpu/device/convolution_forward_specialization.hpp"
+#include "ck/tensor_operation/gpu/grid/gridwise_gemm_xdlops_v3r1.hpp"
+#include "ck/device_utility/device_prop.hpp"
+#include "ck/device_utility/kernel_launch.hpp"

 namespace ck {
 namespace tensor_operation {
@@ -879,4 +879,3 @@ struct DeviceConv2dFwdXdl_C_Shuffle_Input_N_Hi_Wi_C_Weight_K_Y_X_C_Output_N_Ho_W
 } // namespace device
 } // namespace tensor_operation
 } // namespace ck
-#endif
--- a/include/ck/tensor_operation/gpu/device/device_conv2d_fwd_xdl_nhwc_kyxc_nhwk.hpp
+++ b/include/ck/tensor_operation/gpu/device/device_conv2d_fwd_xdl_nhwc_kyxc_nhwk.hpp
@@ -1,17 +1,17 @@
-#ifndef DEVICE_CONV2D_FWD_XDL_NHWC_KYXC_NHWK_HPP
-#define DEVICE_CONV2D_FWD_XDL_NHWC_KYXC_NHWK_HPP
+#pragma once

 #include <iostream>
 #include <sstream>
-#include "device.hpp"
-#include "device_base.hpp"
-#include "device_conv_fwd.hpp"
-#include "convolution_forward_specialization.hpp"
-#include "common_header.hpp"
-#include "tensor_layout.hpp"
-#include "tensor_descriptor.hpp"
-#include "tensor_descriptor_helper.hpp"
-#include "gridwise_gemm_xdlops_v2r3.hpp"
+
+#include "ck/utility/common_header.hpp"
+#include "ck/tensor_description/tensor_descriptor.hpp"
+#include "ck/tensor_description/tensor_descriptor_helper.hpp"
+#include "ck/tensor_operation/gpu/device/tensor_layout.hpp"
+#include "ck/tensor_operation/gpu/device/device_conv_fwd.hpp"
+#include "ck/tensor_operation/gpu/device/convolution_forward_specialization.hpp"
+#include "ck/tensor_operation/gpu/grid/gridwise_gemm_xdlops_v2r3.hpp"
+#include "ck/device_utility/device_prop.hpp"
+#include "ck/device_utility/kernel_launch.hpp"

 namespace ck {
 namespace tensor_operation {
@@ -714,9 +714,8 @@ struct DeviceConv2dFwdXdl_Input_N_Hi_Wi_C_Weight_K_Y_X_C_Output_N_Ho_Wo_K

        return str.str();
    }
-}; // namespace device
+};

 } // namespace device
 } // namespace tensor_operation
 } // namespace ck
-#endif
--- a/include/ck/tensor_operation/gpu/device/device_conv_backward_weight.hpp
+++ b/include/ck/tensor_operation/gpu/device/device_conv_backward_weight.hpp
@@ -1,8 +1,9 @@
-#ifndef DEVICE_CONV_WRW_HPP
-#define DEVICE_CONV_WRW_HPP
+#pragma once

+#include <vector>
 #include <iostream>
-#include "device_base.hpp"
+
+#include "ck/tensor_operation/gpu/device/device_base.hpp"

 namespace ck {
 namespace tensor_operation {
@@ -44,4 +45,3 @@ using DeviceConvBwdWeightPtr = std::unique_ptr<
 } // namespace device
 } // namespace tensor_operation
 } // namespace ck
-#endif
--- a/include/ck/tensor_operation/gpu/device/device_conv_bwd_data.hpp
+++ b/include/ck/tensor_operation/gpu/device/device_conv_bwd_data.hpp
@@ -1,9 +1,10 @@
-#ifndef DEVICE_CONV_BWD_DATA_HPP
-#define DEVICE_CONV_BWD_DATA_HPP
+#pragma once

+#include <vector>
 #include <iostream>
-#include "device_base.hpp"
-#include "element_wise_operation.hpp"
+
+#include "ck/tensor_operation/gpu/device/device_base.hpp"
+#include "ck/tensor_operation/gpu/element/element_wise_operation.hpp"

 namespace ck {
 namespace tensor_operation {
@@ -44,4 +45,3 @@ using DeviceConvBwdDataPtr = std::unique_ptr<
 } // namespace device
 } // namespace tensor_operation
 } // namespace ck
-#endif
--- a/include/ck/tensor_operation/gpu/device/device_conv_fwd.hpp
+++ b/include/ck/tensor_operation/gpu/device/device_conv_fwd.hpp
@@ -1,8 +1,9 @@
-#ifndef DEVICE_CONV_FWD_HPP
-#define DEVICE_CONV_FWD_HPP
+#pragma once

 #include <iostream>
-#include "device_base.hpp"
+#include <vector>
+
+#include "ck/tensor_operation/gpu/device/device_base.hpp"

 namespace ck {
 namespace tensor_operation {
@@ -43,4 +44,3 @@ using DeviceConvFwdPtr = std::unique_ptr<
 } // namespace device
 } // namespace tensor_operation
 } // namespace ck
-#endif
--- a/include/ck/tensor_operation/gpu/device/device_conv_fwd_bias_activation.hpp
+++ b/include/ck/tensor_operation/gpu/device/device_conv_fwd_bias_activation.hpp
@@ -1,8 +1,10 @@
-#ifndef DEVICE_CONV_FWD_BIAS_ACTIVATION_HPP
-#define DEVICE_CONV_FWD_BIAS_ACTIVATION_HPP
+#pragma once

+#include <vector>
 #include <iostream>
-#include "device_base.hpp"
+
+#include "ck/tensor_operation/gpu/device/device_base.hpp"
+#include "ck/tensor_operation/gpu/element/element_wise_operation.hpp"

 namespace ck {
 namespace tensor_operation {
@@ -46,4 +48,3 @@ using DeviceConvFwdBiasActivationPtr =
 } // namespace device
 } // namespace tensor_operation
 } // namespace ck
-#endif
--- a/include/ck/tensor_operation/gpu/device/device_conv_fwd_bias_activation_add.hpp
+++ b/include/ck/tensor_operation/gpu/device/device_conv_fwd_bias_activation_add.hpp
@@ -1,8 +1,9 @@
-#ifndef DEVICE_CONV_FWD_BIAS_ACTIVATION_ADD_HPP
-#define DEVICE_CONV_FWD_BIAS_ACTIVATION_ADD_HPP
+#pragma once

+#include <vector>
 #include <iostream>
-#include "device_base.hpp"
+
+#include "ck/tensor_operation/gpu/device/device_base.hpp"

 namespace ck {
 namespace tensor_operation {
@@ -47,4 +48,3 @@ using DeviceConvFwdBiasActivationAddPtr =
 } // namespace device
 } // namespace tensor_operation
 } // namespace ck
-#endif
--- a/include/ck/tensor_operation/gpu/device/device_convnd_backward_weight_xdl_c_shuffle_nhwc_kyxc_nhwk.hpp
+++ b/include/ck/tensor_operation/gpu/device/device_convnd_backward_weight_xdl_c_shuffle_nhwc_kyxc_nhwk.hpp
@@ -2,16 +2,17 @@

 #include <iostream>
 #include <sstream>
-#include "device.hpp"
-#include "device_base.hpp"
-#include "device_conv_backward_weight.hpp"
-#include "convolution_backward_weight_specialization.hpp"
-#include "common_header.hpp"
-#include "tensor_layout.hpp"
-#include "tensor_descriptor.hpp"
-#include "tensor_descriptor_helper.hpp"
-#include "gridwise_gemm_xdlops_bwd_weight.hpp"
-#include "gridwise_unary_elementwise_1d.hpp"
+
+#include "ck/utility/common_header.hpp"
+#include "ck/tensor_description/tensor_descriptor.hpp"
+#include "ck/tensor_description/tensor_descriptor_helper.hpp"
+#include "ck/tensor_operation/gpu/device/tensor_layout.hpp"
+#include "ck/tensor_operation/gpu/device/device_conv_backward_weight.hpp"
+#include "ck/tensor_operation/gpu/device/convolution_backward_weight_specialization.hpp"
+#include "ck/tensor_operation/gpu/grid/gridwise_gemm_xdlops_bwd_weight.hpp"
+#include "ck/tensor_operation/gpu/grid/gridwise_unary_elementwise_1d.hpp"
+#include "ck/device_utility/device_prop.hpp"
+#include "ck/device_utility/kernel_launch.hpp"

 namespace ck {
 namespace tensor_operation {
--- a/include/ck/tensor_operation/gpu/device/device_convnd_bwd_data_xdl_ndhwc_kzyxc_ndhwk.hpp
+++ b/include/ck/tensor_operation/gpu/device/device_convnd_bwd_data_xdl_ndhwc_kzyxc_ndhwk.hpp
@@ -1,17 +1,17 @@
-#ifndef DEVICE_CONVND_BWD_DATA_XDL_NDHWC_KZYXC_NDHWK_HPP
-#define DEVICE_CONVND_BWD_DATA_XDL_NDHWC_KZYXC_NDHWK_HPP
+#pragma once

 #include <iostream>
 #include <sstream>
-#include "device.hpp"
-#include "device_base.hpp"
-#include "device_conv_bwd_data.hpp"
-#include "convolution_backward_data_specialization.hpp"
-#include "common_header.hpp"
-#include "tensor_layout.hpp"
-#include "tensor_descriptor.hpp"
-#include "tensor_descriptor_helper.hpp"
-#include "gridwise_gemm_xdlops_v2r3.hpp"
+
+#include "ck/utility/common_header.hpp"
+#include "ck/tensor_description/tensor_descriptor.hpp"
+#include "ck/tensor_description/tensor_descriptor_helper.hpp"
+#include "ck/tensor_operation/gpu/device/tensor_layout.hpp"
+#include "ck/tensor_operation/gpu/device/device_conv_bwd_data.hpp"
+#include "ck/tensor_operation/gpu/device/convolution_backward_data_specialization.hpp"
+#include "ck/tensor_operation/gpu/grid/gridwise_gemm_xdlops_v2r3.hpp"
+#include "ck/device_utility/device_prop.hpp"
+#include "ck/device_utility/kernel_launch.hpp"

 namespace ck {
 namespace tensor_operation {
@@ -1546,4 +1546,3 @@ struct DeviceConvndBwdDataXdl_Input_N_Di_Hi_Wi_C_Weight_K_Z_Y_X_C_Output_N_Do_Ho
 } // namespace device
 } // namespace tensor_operation
 } // namespace ck
-#endif
--- a/include/ck/tensor_operation/gpu/device/device_convnd_fwd_xdl_nhwc_kyxc_nhwk.hpp
+++ b/include/ck/tensor_operation/gpu/device/device_convnd_fwd_xdl_nhwc_kyxc_nhwk.hpp
@@ -6,16 +6,15 @@
 #include <numeric>
 #include <sstream>

-#include "device.hpp"
-#include "device_prop.hpp"
-#include "device_base.hpp"
-#include "device_conv_fwd.hpp"
-#include "convolution_forward_specialization.hpp"
-#include "common_header.hpp"
-#include "tensor_layout.hpp"
-#include "tensor_descriptor.hpp"
-#include "tensor_descriptor_helper.hpp"
-#include "gridwise_gemm_xdlops_v2r3.hpp"
+#include "ck/utility/common_header.hpp"
+#include "ck/tensor_description/tensor_descriptor.hpp"
+#include "ck/tensor_description/tensor_descriptor_helper.hpp"
+#include "ck/tensor_operation/gpu/device/tensor_layout.hpp"
+#include "ck/tensor_operation/gpu/device/device_conv_fwd.hpp"
+#include "ck/tensor_operation/gpu/device/convolution_forward_specialization.hpp"
+#include "ck/tensor_operation/gpu/grid/gridwise_gemm_xdlops_v2r3.hpp"
+#include "ck/device_utility/device_prop.hpp"
+#include "ck/device_utility/kernel_launch.hpp"

 namespace ck {
 namespace tensor_operation {
--- a/include/ck/tensor_operation/gpu/device/device_gemm_bias.hpp
+++ b/include/ck/tensor_operation/gpu/device/device_gemm_bias.hpp
@@ -1,6 +1,8 @@
 #pragma once
+
 #include <iostream>
-#include "device_base.hpp"
+
+#include "ck/tensor_operation/gpu/device/device_base.hpp"

 namespace ck {
 namespace tensor_operation {
--- a/include/ck/tensor_operation/gpu/device/device_gemm_bias_activation.hpp
+++ b/include/ck/tensor_operation/gpu/device/device_gemm_bias_activation.hpp
@@ -1,8 +1,8 @@
-#ifndef DEVICE_GEMM_BIAS_ACTIVATION_HPP
-#define DEVICE_GEMM_BIAS_ACTIVATION_HPP
+#pragma once

 #include <iostream>
-#include "device_base.hpp"
+
+#include "ck/tensor_operation/gpu/device/device_base.hpp"

 namespace ck {
 namespace tensor_operation {
@@ -40,4 +40,3 @@ using DeviceGemmBiasActivationPtr = std::unique_ptr<
 } // namespace device
 } // namespace tensor_operation
 } // namespace ck
-#endif
--- a/include/ck/tensor_operation/gpu/device/device_gemm_bias_add_reduce_xdl_cshuffle.hpp
+++ b/include/ck/tensor_operation/gpu/device/device_gemm_bias_add_reduce_xdl_cshuffle.hpp
@@ -1,13 +1,17 @@
 #pragma once
+
 #include <iostream>
 #include <sstream>
-#include "device.hpp"
-#include "device_gemm_reduce.hpp"
-#include "tensor_layout.hpp"
-#include "tensor_descriptor.hpp"
-#include "tensor_descriptor_helper.hpp"
-#include "gridwise_gemm_bias_add_reduce_xdl_cshuffle_v1.hpp"
-#include "gemm_specialization.hpp"
+
+#include "ck/utility/common_header.hpp"
+#include "ck/tensor_description/tensor_descriptor.hpp"
+#include "ck/tensor_description/tensor_descriptor_helper.hpp"
+#include "ck/tensor_operation/gpu/device/tensor_layout.hpp"
+#include "ck/tensor_operation/gpu/device/device_gemm_reduce.hpp"
+#include "ck/tensor_operation/gpu/device/gemm_specialization.hpp"
+#include "ck/tensor_operation/gpu/grid/gridwise_gemm_bias_add_reduce_xdl_cshuffle_v1.hpp"
+#include "ck/device_utility/device_prop.hpp"
+#include "ck/device_utility/kernel_launch.hpp"

 namespace ck {
 namespace tensor_operation {
--- a/include/ck/tensor_operation/gpu/device/device_gemm_dl.hpp
+++ b/include/ck/tensor_operation/gpu/device/device_gemm_dl.hpp
@@ -3,17 +3,15 @@
 #include <iostream>
 #include <sstream>

-#include "device.hpp"
-#include "device_prop.hpp"
-#include "device_base.hpp"
-#include "device_gemm.hpp"
-#include "common_header.hpp"
-#include "tensor_layout.hpp"
-#include "tensor_descriptor.hpp"
-#include "tensor_descriptor_helper.hpp"
-#include "gemm_specialization.hpp"
-#include "element_wise_operation.hpp"
-#include "gridwise_gemm_dl_v1r3.hpp"
+#include "ck/utility/common_header.hpp"
+#include "ck/tensor_description/tensor_descriptor.hpp"
+#include "ck/tensor_description/tensor_descriptor_helper.hpp"
+#include "ck/tensor_operation/gpu/device/tensor_layout.hpp"
+#include "ck/tensor_operation/gpu/device/device_gemm.hpp"
+#include "ck/tensor_operation/gpu/device/gemm_specialization.hpp"
+#include "ck/tensor_operation/gpu/grid/gridwise_gemm_dl_v1r3.hpp"
+#include "ck/device_utility/device_prop.hpp"
+#include "ck/device_utility/kernel_launch.hpp"

 namespace ck {
 namespace tensor_operation {
--- a/include/ck/tensor_operation/gpu/device/device_gemm_multiple_d_xdl_cshuffle.hpp
+++ b/include/ck/tensor_operation/gpu/device/device_gemm_multiple_d_xdl_cshuffle.hpp
@@ -3,15 +3,15 @@
 #include <iostream>
 #include <sstream>

-#include "device.hpp"
-#include "device_gemm_multiple_d.hpp"
-#include "common_header.hpp"
-#include "tensor_layout.hpp"
-#include "tensor_descriptor.hpp"
-#include "tensor_descriptor_helper.hpp"
-#include "gridwise_gemm_multiple_d_xdl_cshuffle.hpp"
-#include "gemm_specialization.hpp"
-#include "device_prop.hpp"
+#include "ck/utility/common_header.hpp"
+#include "ck/tensor_description/tensor_descriptor.hpp"
+#include "ck/tensor_description/tensor_descriptor_helper.hpp"
+#include "ck/tensor_operation/gpu/device/tensor_layout.hpp"
+#include "ck/tensor_operation/gpu/device/device_gemm_multiple_d.hpp"
+#include "ck/tensor_operation/gpu/device/gemm_specialization.hpp"
+#include "ck/tensor_operation/gpu/grid/gridwise_gemm_multiple_d_xdl_cshuffle.hpp"
+#include "ck/device_utility/device_prop.hpp"
+#include "ck/device_utility/kernel_launch.hpp"

 namespace ck {

--- a/include/ck/tensor_operation/gpu/device/device_gemm_reduce_xdl_cshuffle.hpp
+++ b/include/ck/tensor_operation/gpu/device/device_gemm_reduce_xdl_cshuffle.hpp
@@ -1,14 +1,17 @@
 #pragma once
+
 #include <iostream>
 #include <sstream>
-#include "device.hpp"
-#include "device_gemm_reduce.hpp"
-#include "common_header.hpp"
-#include "tensor_layout.hpp"
-#include "tensor_descriptor.hpp"
-#include "tensor_descriptor_helper.hpp"
-#include "gridwise_gemm_reduce_xdl_cshuffle_v1.hpp"
-#include "gemm_specialization.hpp"
+
+#include "ck/utility/common_header.hpp"
+#include "ck/tensor_description/tensor_descriptor.hpp"
+#include "ck/tensor_description/tensor_descriptor_helper.hpp"
+#include "ck/tensor_operation/gpu/device/tensor_layout.hpp"
+#include "ck/tensor_operation/gpu/device/device_gemm_reduce.hpp"
+#include "ck/tensor_operation/gpu/device/gemm_specialization.hpp"
+#include "ck/tensor_operation/gpu/grid/gridwise_gemm_reduce_xdl_cshuffle_v1.hpp"
+#include "ck/device_utility/device_prop.hpp"
+#include "ck/device_utility/kernel_launch.hpp"

 namespace ck {
 namespace tensor_operation {
--- a/include/ck/tensor_operation/gpu/device/device_gemm_xdl.hpp
+++ b/include/ck/tensor_operation/gpu/device/device_gemm_xdl.hpp
@@ -2,16 +2,16 @@

 #include <iostream>
 #include <sstream>
-#include "device.hpp"
-#include "device_prop.hpp"
-#include "device_base.hpp"
-#include "device_gemm.hpp"
-#include "common_header.hpp"
-#include "tensor_layout.hpp"
-#include "tensor_descriptor.hpp"
-#include "tensor_descriptor_helper.hpp"
-#include "gridwise_gemm_xdlops_v2r3.hpp"
-#include "gemm_specialization.hpp"
+
+#include "ck/utility/common_header.hpp"
+#include "ck/tensor_description/tensor_descriptor.hpp"
+#include "ck/tensor_description/tensor_descriptor_helper.hpp"
+#include "ck/tensor_operation/gpu/device/tensor_layout.hpp"
+#include "ck/tensor_operation/gpu/device/device_gemm.hpp"
+#include "ck/tensor_operation/gpu/device/gemm_specialization.hpp"
+#include "ck/tensor_operation/gpu/grid/gridwise_gemm_xdlops_v2r3.hpp"
+#include "ck/device_utility/device_prop.hpp"
+#include "ck/device_utility/kernel_launch.hpp"

 namespace ck {
 namespace tensor_operation {
--- a/include/ck/tensor_operation/gpu/device/device_gemm_xdl_c_shuffle_bias_2d.hpp
+++ b/include/ck/tensor_operation/gpu/device/device_gemm_xdl_c_shuffle_bias_2d.hpp
@@ -1,13 +1,17 @@
 #pragma once
+
 #include <iostream>
 #include <sstream>
-#include "device.hpp"
-#include "device_gemm_bias.hpp"
-#include "common_header.hpp"
-#include "tensor_layout.hpp"
-#include "tensor_descriptor.hpp"
-#include "tensor_descriptor_helper.hpp"
-#include "gridwise_gemm_xdlops_v3r2.hpp"
+
+#include "ck/utility/common_header.hpp"
+#include "ck/tensor_description/tensor_descriptor.hpp"
+#include "ck/tensor_description/tensor_descriptor_helper.hpp"
+#include "ck/tensor_operation/gpu/device/tensor_layout.hpp"
+#include "ck/tensor_operation/gpu/device/device_gemm_bias.hpp"
+#include "ck/tensor_operation/gpu/device/gemm_specialization.hpp"
+#include "ck/tensor_operation/gpu/grid/gridwise_gemm_xdlops_v3r2.hpp"
+#include "ck/device_utility/device_prop.hpp"
+#include "ck/device_utility/kernel_launch.hpp"

 namespace ck {
 namespace tensor_operation {
--- a/include/ck/tensor_operation/gpu/device/device_gemm_xdl_c_shuffle_bias_activation.hpp
+++ b/include/ck/tensor_operation/gpu/device/device_gemm_xdl_c_shuffle_bias_activation.hpp
@@ -1,15 +1,17 @@
-#ifndef DEVICE_GEMM_XDL_C_SHUFFLE_BIAS_ACTIVATION_HPP
-#define DEVICE_GEMM_XDL_C_SHUFFLE_BIAS_ACTIVATION_HPP
+#pragma once

 #include <iostream>
 #include <sstream>
-#include "device.hpp"
-#include "device_gemm_bias_activation.hpp"
-#include "common_header.hpp"
-#include "tensor_layout.hpp"
-#include "tensor_descriptor.hpp"
-#include "tensor_descriptor_helper.hpp"
-#include "gridwise_gemm_xdlops_v3r2.hpp"
+
+#include "ck/utility/common_header.hpp"
+#include "ck/tensor_description/tensor_descriptor.hpp"
+#include "ck/tensor_description/tensor_descriptor_helper.hpp"
+#include "ck/tensor_operation/gpu/device/tensor_layout.hpp"
+#include "ck/tensor_operation/gpu/device/device_gemm_bias_activation.hpp"
+#include "ck/tensor_operation/gpu/device/gemm_specialization.hpp"
+#include "ck/tensor_operation/gpu/grid/gridwise_gemm_xdlops_v3r2.hpp"
+#include "ck/device_utility/device_prop.hpp"
+#include "ck/device_utility/kernel_launch.hpp"

 namespace ck {
 namespace tensor_operation {
@@ -513,4 +515,3 @@ struct DeviceGemmXdl_C_Shuffle_Bias_Activation
 } // namespace device
 } // namespace tensor_operation
 } // namespace ck
-#endif
--- a/include/ck/tensor_operation/gpu/device/device_gemm_xdl_c_shuffle_bias_activation_add.hpp
+++ b/include/ck/tensor_operation/gpu/device/device_gemm_xdl_c_shuffle_bias_activation_add.hpp
@@ -1,15 +1,17 @@
-#ifndef DEVICE_GEMM_XDL_C_SHUFFLE_BIAS_ACTIVATION_ADD_HPP
-#define DEVICE_GEMM_XDL_C_SHUFFLE_BIAS_ACTIVATION_ADD_HPP
+#pragma once

 #include <iostream>
 #include <sstream>
-#include "device.hpp"
-#include "device_gemm_bias_activation_add.hpp"
-#include "common_header.hpp"
-#include "tensor_layout.hpp"
-#include "tensor_descriptor.hpp"
-#include "tensor_descriptor_helper.hpp"
-#include "gridwise_gemm_xdlops_v3r3.hpp"
+
+#include "ck/utility/common_header.hpp"
+#include "ck/tensor_description/tensor_descriptor.hpp"
+#include "ck/tensor_description/tensor_descriptor_helper.hpp"
+#include "ck/tensor_operation/gpu/device/tensor_layout.hpp"
+#include "ck/tensor_operation/gpu/device/device_gemm_bias_activation_add.hpp"
+#include "ck/tensor_operation/gpu/device/gemm_specialization.hpp"
+#include "ck/tensor_operation/gpu/grid/gridwise_gemm_xdlops_v3r3.hpp"
+#include "ck/device_utility/device_prop.hpp"
+#include "ck/device_utility/kernel_launch.hpp"

 namespace ck {
 namespace tensor_operation {
@@ -573,4 +575,3 @@ struct DeviceGemmXdl_C_Shuffle_Bias_Activation_Add
 } // namespace device
 } // namespace tensor_operation
 } // namespace ck
-#endif
--- a/include/ck/tensor_operation/gpu/device/device_gemm_xdl_cshuffle.hpp
+++ b/include/ck/tensor_operation/gpu/device/device_gemm_xdl_cshuffle.hpp
@@ -1,15 +1,17 @@
 #pragma once
+
 #include <iostream>
 #include <sstream>
-#include "device.hpp"
-#include "device_gemm.hpp"
-#include "common_header.hpp"
-#include "tensor_layout.hpp"
-#include "tensor_descriptor.hpp"
-#include "tensor_descriptor_helper.hpp"
-#include "gridwise_gemm_xdl_cshuffle_v1.hpp"
-#include "tensor_operation/gpu/device/gemm_specialization.hpp"
-#include "device_prop.hpp"
+
+#include "ck/utility/common_header.hpp"
+#include "ck/tensor_description/tensor_descriptor.hpp"
+#include "ck/tensor_description/tensor_descriptor_helper.hpp"
+#include "ck/tensor_operation/gpu/device/tensor_layout.hpp"
+#include "ck/tensor_operation/gpu/device/device_gemm.hpp"
+#include "ck/tensor_operation/gpu/device/gemm_specialization.hpp"
+#include "ck/tensor_operation/gpu/grid/gridwise_gemm_xdl_cshuffle_v1.hpp"
+#include "ck/device_utility/device_prop.hpp"
+#include "ck/device_utility/kernel_launch.hpp"

 namespace ck {
 namespace tensor_operation {
--- a/include/ck/tensor_operation/gpu/device/device_gemm_xdl_splitk.hpp
+++ b/include/ck/tensor_operation/gpu/device/device_gemm_xdl_splitk.hpp
@@ -1,22 +1,17 @@
-#ifndef DEVICE_GEMM_SPLITK_XDL_HPP
-#define DEVICE_GEMM_SPLITK_XDL_HPP
+#pragma once

 #include <iostream>
 #include <sstream>
-#include "device.hpp"
-#include "device_base.hpp"
-#include "device_gemm.hpp"
-#include "common_header.hpp"
-#include "tensor_layout.hpp"
-#include "tensor_descriptor.hpp"
-#include "tensor_descriptor_helper.hpp"
-#include "gridwise_gemm_xdlops_v2r4.hpp"
-#include "gemm_specialization.hpp"
-#include "device_prop.hpp"

-#ifndef CK_RUN_KERNEL_AND_TIME
-#define CK_RUN_KERNEL_AND_TIME 1
-#endif
+#include "ck/utility/common_header.hpp"
+#include "ck/tensor_description/tensor_descriptor.hpp"
+#include "ck/tensor_description/tensor_descriptor_helper.hpp"
+#include "ck/tensor_operation/gpu/device/tensor_layout.hpp"
+#include "ck/tensor_operation/gpu/device/device_gemm.hpp"
+#include "ck/tensor_operation/gpu/device/gemm_specialization.hpp"
+#include "ck/tensor_operation/gpu/grid/gridwise_gemm_xdlops_v2r4.hpp"
+#include "ck/device_utility/device_prop.hpp"
+#include "ck/device_utility/kernel_launch.hpp"

 namespace ck {
 namespace tensor_operation {
@@ -639,4 +634,3 @@ struct DeviceGemmXdlSplitK
 } // namespace device
 } // namespace tensor_operation
 } // namespace ck
-#endif
--- a/include/ck/tensor_operation/gpu/device/device_gemm_xdl_splitk_c_shuffle.hpp
+++ b/include/ck/tensor_operation/gpu/device/device_gemm_xdl_splitk_c_shuffle.hpp
@@ -1,21 +1,17 @@
-#ifndef DEVICE_GEMM_XDL_SPLITK_C_SHUFFLE_HPP
-#define DEVICE_GEMM_XDL_SPLITK_C_SHUFFLE_HPP
+#pragma once

 #include <iostream>
 #include <sstream>
-#include "device.hpp"
-#include "device_base.hpp"
-#include "device_gemm.hpp"
-#include "common_header.hpp"
-#include "tensor_layout.hpp"
-#include "tensor_descriptor.hpp"
-#include "tensor_descriptor_helper.hpp"
-#include "gridwise_gemm_xdlops_v2r4r2.hpp"
-#include "gemm_specialization.hpp"

-#ifndef CK_RUN_KERNEL_AND_TIME
-#define CK_RUN_KERNEL_AND_TIME 1
-#endif
+#include "ck/utility/common_header.hpp"
+#include "ck/tensor_description/tensor_descriptor.hpp"
+#include "ck/tensor_description/tensor_descriptor_helper.hpp"
+#include "ck/tensor_operation/gpu/device/tensor_layout.hpp"
+#include "ck/tensor_operation/gpu/device/device_gemm.hpp"
+#include "ck/tensor_operation/gpu/device/gemm_specialization.hpp"
+#include "ck/tensor_operation/gpu/grid/gridwise_gemm_xdlops_v2r4r2.hpp"
+#include "ck/device_utility/device_prop.hpp"
+#include "ck/device_utility/kernel_launch.hpp"

 namespace ck {
 namespace tensor_operation {
@@ -641,4 +637,3 @@ struct DeviceGemmXdlSplitKCShuffle
 } // namespace device
 } // namespace tensor_operation
 } // namespace ck
-#endif
--- a/include/ck/tensor_operation/gpu/device/device_grouped_gemm_xdl.hpp
+++ b/include/ck/tensor_operation/gpu/device/device_grouped_gemm_xdl.hpp
@@ -1,17 +1,17 @@
-#ifndef DEVICE_GROUPED_GEMM_XDL_HPP
-#define DEVICE_GROUPED_GEMM_XDL_HPP
+#pragma once

 #include <iostream>
 #include <sstream>
-#include "device.hpp"
-#include "device_base.hpp"
-#include "device_gemm.hpp"
-#include "common_header.hpp"
-#include "tensor_layout.hpp"
-#include "tensor_descriptor.hpp"
-#include "tensor_descriptor_helper.hpp"
-#include "gridwise_gemm_xdlops_v2r3.hpp"
-#include "gemm_specialization.hpp"
+
+#include "ck/utility/common_header.hpp"
+#include "ck/tensor_description/tensor_descriptor.hpp"
+#include "ck/tensor_description/tensor_descriptor_helper.hpp"
+#include "ck/tensor_operation/gpu/device/tensor_layout.hpp"
+#include "ck/tensor_operation/gpu/device/device_gemm.hpp"
+#include "ck/tensor_operation/gpu/device/gemm_specialization.hpp"
+#include "ck/tensor_operation/gpu/grid/gridwise_gemm_xdlops_v2r3.hpp"
+#include "ck/device_utility/device_prop.hpp"
+#include "ck/device_utility/kernel_launch.hpp"

 namespace ck {
 namespace tensor_operation {
@@ -638,4 +638,3 @@ struct DeviceGroupedGemmXdl
 } // namespace device
 } // namespace tensor_operation
 } // namespace ck
-#endif
--- a/include/ck/tensor_operation/gpu/device/device_pool2d_fwd.hpp
+++ b/include/ck/tensor_operation/gpu/device/device_pool2d_fwd.hpp
@@ -1,10 +1,10 @@
-#ifndef DEVICE_POOL2D_FWD_HPP
-#define DEVICE_POOL2D_FWD_HPP
+#pragma once

 #include <iostream>
 #include <array>
-#include "device_base.hpp"
-#include "reduction_enums.hpp"
+
+#include "ck/tensor_operation/gpu/device/device_base.hpp"
+#include "ck/utility/reduction_enums.hpp"

 namespace ck {
 namespace tensor_operation {
@@ -35,4 +35,3 @@ using DevicePool2dFwdPtr = std::unique_ptr<DevicePool2dFwd<ReduceOpId>>;
 } // namespace device
 } // namespace tensor_operation
 } // namespace ck
-#endif
--- a/include/ck/tensor_operation/gpu/device/device_pool2d_fwd_nhwc_nhwc.hpp
+++ b/include/ck/tensor_operation/gpu/device/device_pool2d_fwd_nhwc_nhwc.hpp
@@ -1,13 +1,15 @@
-#ifndef DEVICE_POOL2D_FWD_NHWC_NHWC_HPP
-#define DEVICE_POOL2D_FWD_NHWC_NHWC_HPP
+#pragma once

 #include <iostream>
 #include <sstream>
-#include "device_pool2d_fwd.hpp"
-#include "tensor_descriptor.hpp"
-#include "tensor_descriptor_helper.hpp"
-#include "reduction_operator_mapping.hpp"
-#include "gridwise_2d_reduction_threadwise.hpp"
+
+#include "ck/tensor_description/tensor_descriptor.hpp"
+#include "ck/tensor_description/tensor_descriptor_helper.hpp"
+#include "ck/tensor_operation/gpu/device/reduction_operator_mapping.hpp"
+#include "ck/tensor_operation/gpu/device/device_pool2d_fwd.hpp"
+#include "ck/tensor_operation/gpu/grid/gridwise_2d_reduction_threadwise.hpp"
+#include "ck/device_utility/device_prop.hpp"
+#include "ck/device_utility/kernel_launch.hpp"

 namespace ck {
 namespace tensor_operation {
@@ -315,9 +317,8 @@ struct DevicePool2dFwd_Input_N_Hi_Wi_C_Output_N_Ho_Wo_C : public DevicePool2dFwd

        return str.str();
    }
-}; // namespace device
+};

 } // namespace device
 } // namespace tensor_operation
 } // namespace ck
-#endif
--- a/include/ck/tensor_operation/gpu/device/device_reduce.hpp
+++ b/include/ck/tensor_operation/gpu/device/device_reduce.hpp
@@ -1,13 +1,12 @@
-#ifndef DEVICE_REDUCE_HPP
-#define DEVICE_REDUCE_HPP
+#pragma once

 #include <vector>
 #include <memory>
 #include <iostream>

-#include "common_header.hpp"
-#include "device_base.hpp"
-#include "reduction_enums.hpp"
+#include "ck/utility/common_header.hpp"
+#include "ck/utility/reduction_enums.hpp"
+#include "ck/tensor_operation/gpu/device/device_base.hpp"

 namespace ck {
 namespace tensor_operation {
@@ -41,4 +40,3 @@ using DeviceReducePtr =
 } // namespace device
 } // namespace tensor_operation
 } // namespace ck
-#endif
--- a/include/ck/tensor_operation/gpu/device/device_reduce_common.hpp
+++ b/include/ck/tensor_operation/gpu/device/device_reduce_common.hpp
@@ -1,12 +1,11 @@
-#ifndef DEVICE_REDUCE_COMMON_HPP
-#define DEVICE_REDUCE_COMMON_HPP
+#pragma once

 #include <vector>
 #include <cassert>

-#include "common_header.hpp"
-#include "reduction_enums.hpp"
-#include "reduction_operator.hpp"
+#include "ck/utility/common_header.hpp"
+#include "ck/utility/reduction_enums.hpp"
+#include "ck/utility/reduction_operator.hpp"

 namespace ck {
 namespace tensor_operation {
@@ -85,6 +84,4 @@ std::vector<index_t> shuffle_tensor_dimensions(const std::vector<index_t>& origL

 } // namespace device
 } // namespace tensor_operation
-
 } // namespace ck
-#endif
--- a/include/ck/tensor_operation/gpu/device/device_reduce_multiblock.hpp
+++ b/include/ck/tensor_operation/gpu/device/device_reduce_multiblock.hpp
@@ -1,15 +1,18 @@
-#ifndef DEVICE_REDUCE_MULTIBLOCK_HPP
-#define DEVICE_REDUCE_MULTIBLOCK_HPP
+#pragma once

 #include <iostream>
 #include <sstream>
-#include "device.hpp"
-#include "device_base.hpp"
-#include "device_reduce.hpp"
-#include "device_reduce_common.hpp"
-#include "gridwise_2d_reduction_multiblock.hpp"
-#include "gridwise_set_buffer_value.hpp"
-#include "reduction_operator.hpp"
+
+#include "ck/utility/common_header.hpp"
+#include "ck/utility/reduction_operator.hpp"
+#include "ck/tensor_description/tensor_descriptor.hpp"
+#include "ck/tensor_description/tensor_descriptor_helper.hpp"
+#include "ck/tensor_operation/gpu/device/device_reduce.hpp"
+#include "ck/tensor_operation/gpu/device/device_reduce_common.hpp"
+#include "ck/tensor_operation/gpu/grid/gridwise_2d_reduction_multiblock.hpp"
+#include "ck/tensor_operation/gpu/grid/gridwise_set_buffer_value.hpp"
+#include "ck/device_utility/device_prop.hpp"
+#include "ck/device_utility/kernel_launch.hpp"

 namespace ck {
 namespace tensor_operation {
@@ -505,4 +508,3 @@ struct DeviceReduceMultiBlock : public DeviceReduce<InElementwiseOperation, AccE
 } // namespace device
 } // namespace tensor_operation
 } // namespace ck
-#endif
--- a/include/ck/tensor_operation/gpu/device/device_reduce_threadwise.hpp
+++ b/include/ck/tensor_operation/gpu/device/device_reduce_threadwise.hpp
@@ -1,13 +1,14 @@
-#ifndef DEVICE_REDUCE_THREADWISE_HPP
-#define DEVICE_REDUCE_THREADWISE_HPP
+#pragma once

 #include <iostream>
 #include <sstream>
-#include "device.hpp"
-#include "device_reduce.hpp"
-#include "device_reduce_common.hpp"
-#include "gridwise_2d_reduction_multiblock.hpp"
-#include "gridwise_2d_reduction_threadwise.hpp"
+
+#include "ck/device_utility/device_prop.hpp"
+#include "ck/device_utility/kernel_launch.hpp"
+#include "ck/tensor_operation/gpu/device/device_reduce.hpp"
+#include "ck/tensor_operation/gpu/device/device_reduce_common.hpp"
+#include "ck/tensor_operation/gpu/grid/gridwise_2d_reduction_multiblock.hpp"
+#include "ck/tensor_operation/gpu/grid/gridwise_2d_reduction_threadwise.hpp"

 namespace ck {
 namespace tensor_operation {
@@ -370,4 +371,3 @@ struct DeviceReduceThreadWise : public DeviceReduce<InElementwiseOperation, AccE
 } // namespace device
 } // namespace tensor_operation
 } // namespace ck
-#endif
--- a/include/ck/tensor_operation/gpu/device/device_softmax.hpp
+++ b/include/ck/tensor_operation/gpu/device/device_softmax.hpp
@@ -1,16 +1,17 @@
-#ifndef DEVICE_SOFTMAX_HPP
-#define DEVICE_SOFTMAX_HPP
+#pragma once

 #include <iostream>
 #include <sstream>
-#include "device.hpp"
-#include "device_base.hpp"
-#include "device_reduce.hpp"
-#include "device_reduce_multiblock.hpp"
-#include "device_reduce_common.hpp"
-#include "gridwise_softmax.hpp"
-#include "gridwise_set_buffer_value.hpp"
-#include "reduction_operator.hpp"
+
+#include "ck/utility/reduction_operator.hpp"
+#include "ck/tensor_operation/gpu/device/device_base.hpp"
+#include "ck/tensor_operation/gpu/device/device_reduce.hpp"
+#include "ck/tensor_operation/gpu/device/device_reduce_multiblock.hpp"
+#include "ck/tensor_operation/gpu/device/device_reduce_common.hpp"
+#include "ck/tensor_operation/gpu/grid/gridwise_softmax.hpp"
+#include "ck/tensor_operation/gpu/grid/gridwise_set_buffer_value.hpp"
+#include "ck/device_utility/device_prop.hpp"
+#include "ck/device_utility/kernel_launch.hpp"

 namespace ck {
 namespace tensor_operation {
@@ -200,4 +201,3 @@ struct DeviceSoftmax : public BaseOperator
 } // namespace device
 } // namespace tensor_operation
 } // namespace ck
-#endif // DEVICE_SOFTMAX_HPP
--- a/include/ck/tensor_operation/gpu/device/device_unary_elementwise.hpp
+++ b/include/ck/tensor_operation/gpu/device/device_unary_elementwise.hpp
@@ -1,10 +1,12 @@
 #pragma once
+
 #include <iostream>
 #include <vector>

-#include "device.hpp"
-#include "device_base.hpp"
-#include "gridwise_unary_elementwise_1d.hpp"
+#include "ck/device_utility/device_prop.hpp"
+#include "ck/device_utility/kernel_launch.hpp"
+#include "ck/tensor_operation/gpu/device/device_base.hpp"
+#include "ck/tensor_operation/gpu/grid/gridwise_unary_elementwise_1d.hpp"

 namespace ck {
 namespace tensor_operation {
--- a/include/ck/tensor_operation/gpu/device/gemm_specialization.hpp
+++ b/include/ck/tensor_operation/gpu/device/gemm_specialization.hpp
@@ -1,5 +1,4 @@
-#ifndef GEMM_SPECIALIZATION
-#define GEMM_SPECIALIZATION
+#pragma once

 namespace ck {
 namespace tensor_operation {
@@ -20,4 +19,3 @@ enum struct GemmSpecialization
 } // namespace device
 } // namespace tensor_operation
 } // namespace ck
-#endif
--- a/include/ck/tensor_operation/gpu/device/reduction_operator_mapping.hpp
+++ b/include/ck/tensor_operation/gpu/device/reduction_operator_mapping.hpp
@@ -1,34 +1,9 @@
-/*******************************************************************************
- *
- * MIT License
- *
- * Copyright (c) 2020 Advanced Micro Devices, Inc.
- *
- * Permission is hereby granted, free of charge, to any person obtaining a copy
- * of this software and associated documentation files (the "Software"), to deal
- * in the Software without restriction, including without limitation the rights
- * to use, copy, modify, merge, publish, distribute, sublicense, and/or sell
- * copies of the Software, and to permit persons to whom the Software is
- * furnished to do so, subject to the following conditions:
- *
- * The above copyright notice and this permission notice shall be included in all
- * copies or substantial portions of the Software.
- *
- * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
- * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
- * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE
- * AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER
- * LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM,
- * OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN THE
- * SOFTWARE.
- *
- *******************************************************************************/
-#ifndef CK_REDUCTION_OPERATOR_MAPPING_HPP
-#define CK_REDUCTION_OPERATOR_MAPPING_HPP
+#pragma once

-#include "reduction_operator.hpp"
-#include "reduction_enums.hpp"
-#include "element_wise_operation.hpp"
+#include "ck/utility/reduction_operator.hpp"
+#include "ck/utility/reduction_enums.hpp"
+#include "ck/tensor_operation/gpu/element/element_wise_operation.hpp"
+// FIXME: can it be replaced with ck::Tuple?
 #include <tuple>

 namespace ck {
@@ -205,6 +180,4 @@ struct reduce_unary_operator<ReduceTensorOp::NORM2, false, true>
    };
 };

-} // end of namespace ck
-
-#endif
+} // namespace ck
--- a/include/ck/tensor_operation/gpu/element/binary_element_wise_operation.hpp
+++ b/include/ck/tensor_operation/gpu/element/binary_element_wise_operation.hpp
@@ -1,31 +1,6 @@
-/*******************************************************************************
- *
- * MIT License
- *
- * Copyright (c) 2022 Advanced Micro Devices, Inc.
- *
- * Permission is hereby granted, free of charge, to any person obtaining a copy
- * of this software and associated documentation files (the "Software"), to deal
- * in the Software without restriction, including without limitation the rights
- * to use, copy, modify, merge, publish, distribute, sublicense, and/or sell
- * copies of the Software, and to permit persons to whom the Software is
- * furnished to do so, subject to the following conditions:
- *
- * The above copyright notice and this permission notice shall be included in all
- * copies or substantial portions of the Software.
- *
- * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
- * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
- * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE
- * AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER
- * LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM,
- * OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN THE
- * SOFTWARE.
- *
- *******************************************************************************/
 #pragma once

-#include "data_type.hpp"
+#include "ck/utility/data_type.hpp"

 namespace ck {
 namespace tensor_operation {
--- a/include/ck/tensor_operation/gpu/element/element_wise_operation.hpp
+++ b/include/ck/tensor_operation/gpu/element/element_wise_operation.hpp
@@ -1,9 +1,9 @@
 #pragma once

-#include "data_type.hpp"
-#include "math_v2.hpp"
-#include "unary_element_wise_operation.hpp"
-#include "binary_element_wise_operation.hpp"
+#include "ck/utility/data_type.hpp"
+#include "ck/utility/math_v2.hpp"
+#include "ck/tensor_operation/gpu/element/unary_element_wise_operation.hpp"
+#include "ck/tensor_operation/gpu/element/binary_element_wise_operation.hpp"

 namespace ck {
 namespace tensor_operation {
--- a/include/ck/tensor_operation/gpu/element/unary_element_wise_operation.hpp
+++ b/include/ck/tensor_operation/gpu/element/unary_element_wise_operation.hpp
@@ -1,7 +1,7 @@
 #pragma once

-#include "data_type.hpp"
-#include "math_v2.hpp"
+#include "ck/utility/data_type.hpp"
+#include "ck/utility/math_v2.hpp"

 namespace ck {
 namespace tensor_operation {
--- a/include/ck/tensor_operation/gpu/grid/block_to_ctile_map.hpp
+++ b/include/ck/tensor_operation/gpu/grid/block_to_ctile_map.hpp
@@ -1,10 +1,9 @@
-#ifndef UTILITY_BLOCK_TO_CTILE_MAP
-#define UTILITY_BLOCK_TO_CTILE_MAP
+#pragma once

-#include "utility/math.hpp"
-#include "utility/number.hpp"
-#include "tensor_description/tensor_adaptor.hpp"
-#include "tensor_description/multi_index_transform_helper.hpp"
+#include "ck/utility/math.hpp"
+#include "ck/utility/number.hpp"
+#include "ck/tensor_description/tensor_adaptor.hpp"
+#include "ck/tensor_description/multi_index_transform_helper.hpp"

 namespace ck {

@@ -485,5 +484,3 @@ __host__ __device__ bool DefaultValidCTileIndex(const CTileIdx& c_tile_idx,
 }

 } // namespace ck
-
-#endif // UTILITY_BLOCK_TO_CTILE_MAP
--- a/include/ck/tensor_operation/gpu/grid/gridwise_2d_reduction_multiblock.hpp
+++ b/include/ck/tensor_operation/gpu/grid/gridwise_2d_reduction_multiblock.hpp
@@ -1,39 +1,12 @@
-/*******************************************************************************
- *
- * MIT License
- *
- * Copyright (c) 2020 Advanced Micro Devices, Inc.
- *
- * Permission is hereby granted, free of charge, to any person obtaining a copy
- * of this software and associated documentation files (the "Software"), to deal
- * in the Software without restriction, including without limitation the rights
- * to use, copy, modify, merge, publish, distribute, sublicense, and/or sell
- * copies of the Software, and to permit persons to whom the Software is
- * furnished to do so, subject to the following conditions:
- *
- * The above copyright notice and this permission notice shall be included in all
- * copies or substantial portions of the Software.
- *
- * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
- * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
- * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE
- * AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER
- * LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM,
- * OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN THE
- * SOFTWARE.
- *
- *******************************************************************************/
-#ifndef CK_GRIDWISE_2D_REDUCTION_MULTIBLOCK_HPP
-#define CK_GRIDWISE_2D_REDUCTION_MULTIBLOCK_HPP
+#pragma once

-#include "reduction_common.hpp"
-#include "reduction_operator.hpp"
-#include "reduction_functions_accumulate.hpp"
-#include "reduction_functions_blockwise.hpp"
-#include "reduction_functions_threadwise.hpp"
-
-#include "threadwise_tensor_slice_transfer.hpp"
-#include "element_wise_operation.hpp"
+#include "ck/utility/reduction_common.hpp"
+#include "ck/utility/reduction_operator.hpp"
+#include "ck/utility/reduction_functions_accumulate.hpp"
+#include "ck/tensor_operation/gpu/block/reduction_functions_blockwise.hpp"
+#include "ck/tensor_operation/gpu/thread/reduction_functions_threadwise.hpp"
+#include "ck/tensor_operation/gpu/thread/threadwise_tensor_slice_transfer.hpp"
+#include "ck/tensor_operation/gpu/element/element_wise_operation.hpp"

 namespace ck {

@@ -635,4 +608,3 @@ struct GridwiseReduction_mk_to_m_multiblock
 };

 } // namespace ck
-#endif
--- a/include/ck/tensor_operation/gpu/grid/gridwise_2d_reduction_threadwise.hpp
+++ b/include/ck/tensor_operation/gpu/grid/gridwise_2d_reduction_threadwise.hpp
@@ -1,38 +1,12 @@
-/*******************************************************************************
- *
- * MIT License
- *
- * Copyright (c) 2021 Advanced Micro Devices, Inc.
- *
- * Permission is hereby granted, free of charge, to any person obtaining a copy
- * of this software and associated documentation files (the "Software"), to deal
- * in the Software without restriction, including without limitation the rights
- * to use, copy, modify, merge, publish, distribute, sublicense, and/or sell
- * copies of the Software, and to permit persons to whom the Software is
- * furnished to do so, subject to the following conditions:
- *
- * The above copyright notice and this permission notice shall be included in all
- * copies or substantial portions of the Software.
- *
- * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
- * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
- * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE
- * AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER
- * LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM,
- * OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN THE
- * SOFTWARE.
- *
- *******************************************************************************/
-#ifndef CK_GRIDWISE_2D_REDUCTION_THREADWISE_HPP
-#define CK_GRIDWISE_2D_REDUCTION_THREADWISE_HPP
+#pragma once

-#include "data_type.hpp"
-#include "reduction_common.hpp"
-#include "reduction_operator.hpp"
-#include "reduction_functions_accumulate.hpp"
-#include "reduction_functions_threadwise.hpp"
-#include "threadwise_tensor_slice_transfer.hpp"
-#include "element_wise_operation.hpp"
+#include "ck/utility/data_type.hpp"
+#include "ck/utility/reduction_common.hpp"
+#include "ck/utility/reduction_operator.hpp"
+#include "ck/utility/reduction_functions_accumulate.hpp"
+#include "ck/tensor_operation/gpu/thread/reduction_functions_threadwise.hpp"
+#include "ck/tensor_operation/gpu/thread/threadwise_tensor_slice_transfer.hpp"
+#include "ck/tensor_operation/gpu/element/element_wise_operation.hpp"

 namespace ck {

@@ -495,4 +469,3 @@ struct GridwiseReduction_mk_to_m_threadwise
 };

 } // namespace ck
-#endif
--- a/include/ck/tensor_operation/gpu/grid/gridwise_5ary_Elementwise_1d.hpp
+++ b/include/ck/tensor_operation/gpu/grid/gridwise_5ary_Elementwise_1d.hpp
@@ -1,9 +1,9 @@
 #pragma once

-#include "cluster_descriptor.hpp"
-#include "data_type.hpp"
-#include "element_wise_operation.hpp"
-#include "threadwise_tensor_slice_transfer.hpp"
+#include "ck/tensor_description/cluster_descriptor.hpp"
+#include "ck/utility/data_type.hpp"
+#include "ck/tensor_operation/gpu/element/element_wise_operation.hpp"
+#include "ck/tensor_operation/gpu/thread/threadwise_tensor_slice_transfer.hpp"

 namespace ck {

--- a/include/ck/tensor_operation/gpu/grid/gridwise_binary_elementwise_1d.hpp
+++ b/include/ck/tensor_operation/gpu/grid/gridwise_binary_elementwise_1d.hpp
@@ -1,9 +1,9 @@
 #pragma once

-#include "cluster_descriptor.hpp"
-#include "data_type.hpp"
-#include "element_wise_operation.hpp"
-#include "threadwise_tensor_slice_transfer.hpp"
+#include "ck/utility/data_type.hpp"
+#include "ck/tensor_description/cluster_descriptor.hpp"
+#include "ck/tensor_operation/gpu/element/element_wise_operation.hpp"
+#include "ck/tensor_operation/gpu/thread/threadwise_tensor_slice_transfer.hpp"

 namespace ck {

--- a/include/ck/tensor_operation/gpu/grid/gridwise_gemm_bias_add_reduce_xdl_cshuffle_v1.hpp
+++ b/include/ck/tensor_operation/gpu/grid/gridwise_gemm_bias_add_reduce_xdl_cshuffle_v1.hpp
@@ -1,14 +1,17 @@
 #pragma once
-#include "multi_index_transform_helper.hpp"
-#include "tensor_descriptor.hpp"
-#include "tensor_descriptor_helper.hpp"
-#include "tensor_operation/gpu/grid/block_to_ctile_map.hpp"
-#include "blockwise_gemm_xdlops.hpp"
-#include "thread_group_tensor_slice_transfer_v4r1.hpp"
-#include "thread_group_tensor_slice_transfer_v6r1.hpp"
-#include "threadwise_tensor_slice_transfer.hpp"
-#include "gridwise_gemm_pipeline_v1.hpp"
-#include "reduction_functions_threadwise.hpp"
+
+#include "ck/utility/common_header.hpp"
+#include "ck/tensor_description/multi_index_transform_helper.hpp"
+#include "ck/tensor_description/tensor_descriptor.hpp"
+#include "ck/tensor_description/tensor_descriptor_helper.hpp"
+#include "ck/tensor_operation/gpu/grid/block_to_ctile_map.hpp"
+#include "ck/tensor_operation/gpu/grid/gridwise_gemm_pipeline_v1.hpp"
+#include "ck/tensor_operation/gpu/block/blockwise_gemm_xdlops.hpp"
+#include "ck/tensor_operation/gpu/block/thread_group_tensor_slice_transfer_v4r1.hpp"
+#include "ck/tensor_operation/gpu/block/thread_group_tensor_slice_transfer_v6r1.hpp"
+#include "ck/tensor_operation/gpu/thread/threadwise_tensor_slice_transfer.hpp"
+#include "ck/tensor_operation/gpu/thread/reduction_functions_threadwise.hpp"
+#include "ck/tensor_operation/gpu/element/element_wise_operation.hpp"

 namespace ck {

--- a/include/ck/tensor_operation/gpu/grid/gridwise_gemm_dl_v1r3.hpp
+++ b/include/ck/tensor_operation/gpu/grid/gridwise_gemm_dl_v1r3.hpp
@@ -1,15 +1,16 @@
 #pragma once

-#include "common_header.hpp"
-#include "multi_index_transform_helper.hpp"
-#include "tensor_descriptor.hpp"
-#include "tensor_descriptor_helper.hpp"
-#include "tensor_operation/gpu/grid/block_to_ctile_map.hpp"
-#include "blockwise_gemm_dl_v2r3.hpp"
-#include "blockwise_tensor_slice_transfer_v5r1.hpp"
-#include "threadwise_tensor_slice_transfer.hpp"
-#include "threadwise_tensor_slice_set.hpp"
-#include "element_wise_operation.hpp"
+#include "ck/utility/common_header.hpp"
+#include "ck/tensor_description/multi_index_transform_helper.hpp"
+#include "ck/tensor_description/tensor_descriptor.hpp"
+#include "ck/tensor_description/tensor_descriptor_helper.hpp"
+#include "ck/tensor_operation/gpu/grid/block_to_ctile_map.hpp"
+#include "ck/tensor_operation/gpu/grid/gridwise_gemm_pipeline_v1.hpp"
+#include "ck/tensor_operation/gpu/block/blockwise_gemm_dl_v2r3.hpp"
+#include "ck/tensor_operation/gpu/block/blockwise_tensor_slice_transfer_v5r1.hpp"
+#include "ck/tensor_operation/gpu/thread/threadwise_tensor_slice_transfer.hpp"
+#include "ck/tensor_operation/gpu/thread/threadwise_tensor_slice_set.hpp"
+#include "ck/tensor_operation/gpu/element/element_wise_operation.hpp"

 namespace ck {

--- a/include/ck/tensor_operation/gpu/grid/gridwise_gemm_multiple_d_xdl_cshuffle.hpp
+++ b/include/ck/tensor_operation/gpu/grid/gridwise_gemm_multiple_d_xdl_cshuffle.hpp
@@ -1,15 +1,16 @@
 #pragma once

-#include "common_header.hpp"
-#include "multi_index_transform_helper.hpp"
-#include "tensor_descriptor.hpp"
-#include "tensor_descriptor_helper.hpp"
-#include "tensor_operation/gpu/grid/block_to_ctile_map.hpp"
-#include "blockwise_gemm_xdlops.hpp"
-#include "thread_group_tensor_slice_transfer_v4r1.hpp"
-#include "thread_group_tensor_slice_transfer_v7.hpp"
-#include "threadwise_tensor_slice_transfer.hpp"
-#include "gridwise_gemm_pipeline_v1.hpp"
+#include "ck/utility/common_header.hpp"
+#include "ck/tensor_description/multi_index_transform_helper.hpp"
+#include "ck/tensor_description/tensor_descriptor.hpp"
+#include "ck/tensor_description/tensor_descriptor_helper.hpp"
+#include "ck/tensor_operation/gpu/grid/block_to_ctile_map.hpp"
+#include "ck/tensor_operation/gpu/grid/gridwise_gemm_pipeline_v1.hpp"
+#include "ck/tensor_operation/gpu/block/blockwise_gemm_xdlops.hpp"
+#include "ck/tensor_operation/gpu/block/thread_group_tensor_slice_transfer_v4r1.hpp"
+#include "ck/tensor_operation/gpu/block/thread_group_tensor_slice_transfer_v7.hpp"
+#include "ck/tensor_operation/gpu/thread/threadwise_tensor_slice_transfer.hpp"
+#include "ck/tensor_operation/gpu/element/element_wise_operation.hpp"

 namespace ck {

--- a/include/ck/tensor_operation/gpu/grid/gridwise_gemm_pipeline_v1.hpp
+++ b/include/ck/tensor_operation/gpu/grid/gridwise_gemm_pipeline_v1.hpp
@@ -1,6 +1,7 @@
 #pragma once
-#include "common_header.hpp"
-#include "tensor_operation/gpu/block/blockwise_gemm_xdlops.hpp"
+
+#include "ck/utility/common_header.hpp"
+#include "ck/tensor_operation/gpu/block/blockwise_gemm_xdlops.hpp"

 namespace ck {

--- a/include/ck/tensor_operation/gpu/grid/gridwise_gemm_reduce_xdl_cshuffle_v1.hpp
+++ b/include/ck/tensor_operation/gpu/grid/gridwise_gemm_reduce_xdl_cshuffle_v1.hpp
@@ -1,15 +1,17 @@
 #pragma once
-#include "common_header.hpp"
-#include "multi_index_transform_helper.hpp"
-#include "tensor_descriptor.hpp"
-#include "tensor_descriptor_helper.hpp"
-#include "tensor_operation/gpu/grid/block_to_ctile_map.hpp"
-#include "blockwise_gemm_xdlops.hpp"
-#include "thread_group_tensor_slice_transfer_v4r1.hpp"
-#include "thread_group_tensor_slice_transfer_v6r1.hpp"
-#include "threadwise_tensor_slice_transfer.hpp"
-#include "gridwise_gemm_pipeline_v1.hpp"
-#include "reduction_functions_threadwise.hpp"
+
+#include "ck/utility/common_header.hpp"
+#include "ck/tensor_description/multi_index_transform_helper.hpp"
+#include "ck/tensor_description/tensor_descriptor.hpp"
+#include "ck/tensor_description/tensor_descriptor_helper.hpp"
+#include "ck/tensor_operation/gpu/grid/block_to_ctile_map.hpp"
+#include "ck/tensor_operation/gpu/grid/gridwise_gemm_pipeline_v1.hpp"
+#include "ck/tensor_operation/gpu/block/blockwise_gemm_xdlops.hpp"
+#include "ck/tensor_operation/gpu/block/thread_group_tensor_slice_transfer_v4r1.hpp"
+#include "ck/tensor_operation/gpu/block/thread_group_tensor_slice_transfer_v6r1.hpp"
+#include "ck/tensor_operation/gpu/thread/threadwise_tensor_slice_transfer.hpp"
+#include "ck/tensor_operation/gpu/thread/reduction_functions_threadwise.hpp"
+#include "ck/tensor_operation/gpu/element/element_wise_operation.hpp"

 namespace ck {

--- a/include/ck/tensor_operation/gpu/grid/gridwise_gemm_xdl_cshuffle_v1.hpp
+++ b/include/ck/tensor_operation/gpu/grid/gridwise_gemm_xdl_cshuffle_v1.hpp
@@ -1,14 +1,16 @@
 #pragma once
-#include "common_header.hpp"
-#include "multi_index_transform_helper.hpp"
-#include "tensor_descriptor.hpp"
-#include "tensor_descriptor_helper.hpp"
-#include "tensor_operation/gpu/grid/block_to_ctile_map.hpp"
-#include "blockwise_gemm_xdlops.hpp"
-#include "thread_group_tensor_slice_transfer_v4r1.hpp"
-#include "thread_group_tensor_slice_transfer_v6r1.hpp"
-#include "threadwise_tensor_slice_transfer.hpp"
-#include "gridwise_gemm_pipeline_v1.hpp"
+
+#include "ck/utility/common_header.hpp"
+#include "ck/tensor_description/multi_index_transform_helper.hpp"
+#include "ck/tensor_description/tensor_descriptor.hpp"
+#include "ck/tensor_description/tensor_descriptor_helper.hpp"
+#include "ck/tensor_operation/gpu/grid/block_to_ctile_map.hpp"
+#include "ck/tensor_operation/gpu/grid/gridwise_gemm_pipeline_v1.hpp"
+#include "ck/tensor_operation/gpu/block/blockwise_gemm_xdlops.hpp"
+#include "ck/tensor_operation/gpu/block/thread_group_tensor_slice_transfer_v4r1.hpp"
+#include "ck/tensor_operation/gpu/block/thread_group_tensor_slice_transfer_v6r1.hpp"
+#include "ck/tensor_operation/gpu/thread/threadwise_tensor_slice_transfer.hpp"
+#include "ck/tensor_operation/gpu/element/element_wise_operation.hpp"

 namespace ck {

--- a/include/ck/tensor_operation/gpu/grid/gridwise_gemm_xdlops_bwd_weight.hpp
+++ b/include/ck/tensor_operation/gpu/grid/gridwise_gemm_xdlops_bwd_weight.hpp
@@ -1,15 +1,16 @@
 #pragma once

-#include "common_header.hpp"
-#include "multi_index_transform_helper.hpp"
-#include "tensor_descriptor.hpp"
-#include "tensor_descriptor_helper.hpp"
-#include "tensor_operation/gpu/grid/block_to_ctile_map.hpp"
-#include "blockwise_gemm_xdlops.hpp"
-#include "thread_group_tensor_slice_transfer_v4r1.hpp"
-#include "thread_group_tensor_slice_transfer_v6r1.hpp"
-#include "threadwise_tensor_slice_transfer.hpp"
-#include "gridwise_gemm_pipeline_v1.hpp"
+#include "ck/utility/common_header.hpp"
+#include "ck/tensor_description/multi_index_transform_helper.hpp"
+#include "ck/tensor_description/tensor_descriptor.hpp"
+#include "ck/tensor_description/tensor_descriptor_helper.hpp"
+#include "ck/tensor_operation/gpu/grid/block_to_ctile_map.hpp"
+#include "ck/tensor_operation/gpu/grid/gridwise_gemm_pipeline_v1.hpp"
+#include "ck/tensor_operation/gpu/block/blockwise_gemm_xdlops.hpp"
+#include "ck/tensor_operation/gpu/block/thread_group_tensor_slice_transfer_v4r1.hpp"
+#include "ck/tensor_operation/gpu/block/thread_group_tensor_slice_transfer_v6r1.hpp"
+#include "ck/tensor_operation/gpu/thread/threadwise_tensor_slice_transfer.hpp"
+#include "ck/tensor_operation/gpu/element/element_wise_operation.hpp"

 namespace ck {

--- a/include/ck/tensor_operation/gpu/grid/gridwise_gemm_xdlops_v2r3.hpp
+++ b/include/ck/tensor_operation/gpu/grid/gridwise_gemm_xdlops_v2r3.hpp
@@ -1,14 +1,15 @@
 #pragma once

-#include "common_header.hpp"
-#include "multi_index_transform_helper.hpp"
-#include "tensor_descriptor.hpp"
-#include "tensor_descriptor_helper.hpp"
-#include "tensor_operation/gpu/grid/block_to_ctile_map.hpp"
-#include "blockwise_gemm_xdlops.hpp"
-#include "thread_group_tensor_slice_transfer_v4r1.hpp"
-#include "threadwise_tensor_slice_transfer.hpp"
-#include "gridwise_gemm_pipeline_v1.hpp"
+#include "ck/utility/common_header.hpp"
+#include "ck/tensor_description/multi_index_transform_helper.hpp"
+#include "ck/tensor_description/tensor_descriptor.hpp"
+#include "ck/tensor_description/tensor_descriptor_helper.hpp"
+#include "ck/tensor_operation/gpu/grid/block_to_ctile_map.hpp"
+#include "ck/tensor_operation/gpu/grid/gridwise_gemm_pipeline_v1.hpp"
+#include "ck/tensor_operation/gpu/block/blockwise_gemm_xdlops.hpp"
+#include "ck/tensor_operation/gpu/block/thread_group_tensor_slice_transfer_v4r1.hpp"
+#include "ck/tensor_operation/gpu/thread/threadwise_tensor_slice_transfer.hpp"
+#include "ck/tensor_operation/gpu/element/element_wise_operation.hpp"

 namespace ck {

--- a/include/ck/tensor_operation/gpu/grid/gridwise_gemm_xdlops_v2r4.hpp
+++ b/include/ck/tensor_operation/gpu/grid/gridwise_gemm_xdlops_v2r4.hpp
@@ -1,14 +1,15 @@
-#ifndef CK_GRIDWISE_GEMM_XDLOPS_V2R4_HPP
-#define CK_GRIDWISE_GEMM_XDLOPS_V2R4_HPP
+#pragma once

-#include "common_header.hpp"
-#include "multi_index_transform_helper.hpp"
-#include "tensor_descriptor.hpp"
-#include "tensor_descriptor_helper.hpp"
-#include "tensor_operation/gpu/grid/block_to_ctile_map.hpp"
-#include "blockwise_gemm_xdlops.hpp"
-#include "thread_group_tensor_slice_transfer_v4r1.hpp"
-#include "threadwise_tensor_slice_transfer.hpp"
+#include "ck/utility/common_header.hpp"
+#include "ck/tensor_description/multi_index_transform_helper.hpp"
+#include "ck/tensor_description/tensor_descriptor.hpp"
+#include "ck/tensor_description/tensor_descriptor_helper.hpp"
+#include "ck/tensor_operation/gpu/grid/block_to_ctile_map.hpp"
+#include "ck/tensor_operation/gpu/grid/gridwise_gemm_pipeline_v1.hpp"
+#include "ck/tensor_operation/gpu/block/blockwise_gemm_xdlops.hpp"
+#include "ck/tensor_operation/gpu/block/thread_group_tensor_slice_transfer_v4r1.hpp"
+#include "ck/tensor_operation/gpu/thread/threadwise_tensor_slice_transfer.hpp"
+#include "ck/tensor_operation/gpu/element/element_wise_operation.hpp"

 namespace ck {

@@ -607,7 +608,6 @@ struct GridwiseGemm_bk0mk1_bk0nk1_mn_xdlops_v2r4
                              c_grid_buf);
        }
    }
-}; // namespace ck
+};

 } // namespace ck
-#endif
--- a/include/ck/tensor_operation/gpu/grid/gridwise_gemm_xdlops_v2r4r2.hpp
+++ b/include/ck/tensor_operation/gpu/grid/gridwise_gemm_xdlops_v2r4r2.hpp
@@ -1,15 +1,16 @@
-#ifndef CK_GRIDWISE_GEMM_XDLOPS_V2R4R2_HPP
-#define CK_GRIDWISE_GEMM_XDLOPS_V2R4R2_HPP
+#pragma once

-#include "common_header.hpp"
-#include "multi_index_transform_helper.hpp"
-#include "tensor_descriptor.hpp"
-#include "tensor_descriptor_helper.hpp"
-#include "tensor_operation/gpu/grid/block_to_ctile_map.hpp"
-#include "blockwise_gemm_xdlops.hpp"
-#include "thread_group_tensor_slice_transfer_v4r1.hpp"
-#include "thread_group_tensor_slice_transfer_v6r1.hpp"
-#include "threadwise_tensor_slice_transfer.hpp"
+#include "ck/utility/common_header.hpp"
+#include "ck/tensor_description/multi_index_transform_helper.hpp"
+#include "ck/tensor_description/tensor_descriptor.hpp"
+#include "ck/tensor_description/tensor_descriptor_helper.hpp"
+#include "ck/tensor_operation/gpu/grid/block_to_ctile_map.hpp"
+#include "ck/tensor_operation/gpu/grid/gridwise_gemm_pipeline_v1.hpp"
+#include "ck/tensor_operation/gpu/block/blockwise_gemm_xdlops.hpp"
+#include "ck/tensor_operation/gpu/block/thread_group_tensor_slice_transfer_v4r1.hpp"
+#include "ck/tensor_operation/gpu/block/thread_group_tensor_slice_transfer_v6r1.hpp"
+#include "ck/tensor_operation/gpu/thread/threadwise_tensor_slice_transfer.hpp"
+#include "ck/tensor_operation/gpu/element/element_wise_operation.hpp"

 namespace ck {

@@ -717,7 +718,6 @@ struct GridwiseGemm_bk0mk1_bk0nk1_mn_xdlops_v2r4r2
            });
        }
    }
-}; // namespace ck
+};

 } // namespace ck
-#endif
--- a/include/ck/tensor_operation/gpu/grid/gridwise_gemm_xdlops_v3r1.hpp
+++ b/include/ck/tensor_operation/gpu/grid/gridwise_gemm_xdlops_v3r1.hpp
@@ -1,15 +1,17 @@
 #pragma once
-#include "common_header.hpp"
-#include "multi_index_transform_helper.hpp"
-#include "tensor_descriptor.hpp"
-#include "tensor_descriptor_helper.hpp"
-#include "tensor_operation/gpu/grid/block_to_ctile_map.hpp"
-#include "blockwise_gemm_xdlops.hpp"
-#include "thread_group_tensor_slice_transfer_v4r1.hpp"
-#include "thread_group_tensor_slice_transfer_v6r1.hpp"
-#include "threadwise_tensor_slice_transfer.hpp"
-#include "gridwise_gemm_pipeline_v1.hpp"
-#include "tensor_space_filling_curve.hpp"
+
+#include "ck/utility/common_header.hpp"
+#include "ck/tensor_description/multi_index_transform_helper.hpp"
+#include "ck/tensor_description/tensor_space_filling_curve.hpp"
+#include "ck/tensor_description/tensor_descriptor.hpp"
+#include "ck/tensor_description/tensor_descriptor_helper.hpp"
+#include "ck/tensor_operation/gpu/grid/block_to_ctile_map.hpp"
+#include "ck/tensor_operation/gpu/grid/gridwise_gemm_pipeline_v1.hpp"
+#include "ck/tensor_operation/gpu/block/blockwise_gemm_xdlops.hpp"
+#include "ck/tensor_operation/gpu/block/thread_group_tensor_slice_transfer_v4r1.hpp"
+#include "ck/tensor_operation/gpu/block/thread_group_tensor_slice_transfer_v6r1.hpp"
+#include "ck/tensor_operation/gpu/thread/threadwise_tensor_slice_transfer.hpp"
+#include "ck/tensor_operation/gpu/element/element_wise_operation.hpp"

 namespace ck {

--- a/include/ck/tensor_operation/gpu/grid/gridwise_gemm_xdlops_v3r2.hpp
+++ b/include/ck/tensor_operation/gpu/grid/gridwise_gemm_xdlops_v3r2.hpp
@@ -1,16 +1,16 @@
-#ifndef CK_GRIDWISE_GEMM_XDLOPS_V3R2_HPP
-#define CK_GRIDWISE_GEMM_XDLOPS_V3R2_HPP
+#pragma once

-#include "common_header.hpp"
-#include "multi_index_transform_helper.hpp"
-#include "tensor_descriptor.hpp"
-#include "tensor_descriptor_helper.hpp"
-#include "tensor_operation/gpu/grid/block_to_ctile_map.hpp"
-#include "blockwise_gemm_xdlops.hpp"
-#include "thread_group_tensor_slice_transfer_v4r1.hpp"
-#include "thread_group_tensor_slice_transfer_v6r2.hpp"
-#include "threadwise_tensor_slice_transfer.hpp"
-#include "gridwise_gemm_pipeline_v1.hpp"
+#include "ck/utility/common_header.hpp"
+#include "ck/tensor_description/multi_index_transform_helper.hpp"
+#include "ck/tensor_description/tensor_descriptor.hpp"
+#include "ck/tensor_description/tensor_descriptor_helper.hpp"
+#include "ck/tensor_operation/gpu/grid/block_to_ctile_map.hpp"
+#include "ck/tensor_operation/gpu/grid/gridwise_gemm_pipeline_v1.hpp"
+#include "ck/tensor_operation/gpu/block/blockwise_gemm_xdlops.hpp"
+#include "ck/tensor_operation/gpu/block/thread_group_tensor_slice_transfer_v4r1.hpp"
+#include "ck/tensor_operation/gpu/block/thread_group_tensor_slice_transfer_v6r2.hpp"
+#include "ck/tensor_operation/gpu/thread/threadwise_tensor_slice_transfer.hpp"
+#include "ck/tensor_operation/gpu/element/element_wise_operation.hpp"

 namespace ck {

@@ -755,4 +755,3 @@ struct GridwiseGemm_k0mk1_k0nk1_mn_xdlops_v3r2
 };

 } // namespace ck
-#endif
--- a/include/ck/tensor_operation/gpu/grid/gridwise_gemm_xdlops_v3r3.hpp
+++ b/include/ck/tensor_operation/gpu/grid/gridwise_gemm_xdlops_v3r3.hpp
@@ -1,14 +1,16 @@
 #pragma once
-#include "common_header.hpp"
-#include "multi_index_transform_helper.hpp"
-#include "tensor_descriptor.hpp"
-#include "tensor_descriptor_helper.hpp"
-#include "tensor_operation/gpu/grid/block_to_ctile_map.hpp"
-#include "blockwise_gemm_xdlops.hpp"
-#include "thread_group_tensor_slice_transfer_v4r1.hpp"
-#include "thread_group_tensor_slice_transfer_v6r3.hpp"
-#include "threadwise_tensor_slice_transfer.hpp"
-#include "gridwise_gemm_pipeline_v1.hpp"
+
+#include "ck/utility/common_header.hpp"
+#include "ck/tensor_description/multi_index_transform_helper.hpp"
+#include "ck/tensor_description/tensor_descriptor.hpp"
+#include "ck/tensor_description/tensor_descriptor_helper.hpp"
+#include "ck/tensor_operation/gpu/grid/block_to_ctile_map.hpp"
+#include "ck/tensor_operation/gpu/grid/gridwise_gemm_pipeline_v1.hpp"
+#include "ck/tensor_operation/gpu/block/blockwise_gemm_xdlops.hpp"
+#include "ck/tensor_operation/gpu/block/thread_group_tensor_slice_transfer_v4r1.hpp"
+#include "ck/tensor_operation/gpu/block/thread_group_tensor_slice_transfer_v6r3.hpp"
+#include "ck/tensor_operation/gpu/thread/threadwise_tensor_slice_transfer.hpp"
+#include "ck/tensor_operation/gpu/element/element_wise_operation.hpp"

 namespace ck {

--- a/include/ck/tensor_operation/gpu/grid/gridwise_set_buffer_value.hpp
+++ b/include/ck/tensor_operation/gpu/grid/gridwise_set_buffer_value.hpp
@@ -1,32 +1,6 @@
-/*******************************************************************************
- *
- * MIT License
- *
- * Copyright (c) 2020 Advanced Micro Devices, Inc.
- *
- * Permission is hereby granted, free of charge, to any person obtaining a copy
- * of this software and associated documentation files (the "Software"), to deal
- * in the Software without restriction, including without limitation the rights
- * to use, copy, modify, merge, publish, distribute, sublicense, and/or sell
- * copies of the Software, and to permit persons to whom the Software is
- * furnished to do so, subject to the following conditions:
- *
- * The above copyright notice and this permission notice shall be included in all
- * copies or substantial portions of the Software.
- *
- * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
- * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
- * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE
- * AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER
- * LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM,
- * OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN THE
- * SOFTWARE.
- *
- *******************************************************************************/
-#ifndef CK_GRIDWISE_SET_BUFFER_VALUE_HPP
-#define CK_GRIDWISE_SET_BUFFER_VALUE_HPP
+#pragma once

-#include "threadwise_tensor_slice_transfer.hpp"
+#include "ck/tensor_operation/gpu/thread/threadwise_tensor_slice_transfer.hpp"

 namespace ck {

@@ -77,4 +51,3 @@ __global__ void kernel_buffer_set_value(const Grid1dBufferDescType grid_1d_buffe
 };

 } // namespace ck
-#endif
--- a/include/ck/tensor_operation/gpu/grid/gridwise_softmax.hpp
+++ b/include/ck/tensor_operation/gpu/grid/gridwise_softmax.hpp
@@ -1,39 +1,13 @@
-/*******************************************************************************
- *
- * MIT License
- *
- * Copyright (c) 2022 Advanced Micro Devices, Inc.
- *
- * Permission is hereby granted, free of charge, to any person obtaining a copy
- * of this software and associated documentation files (the "Software"), to deal
- * in the Software without restriction, including without limitation the rights
- * to use, copy, modify, merge, publish, distribute, sublicense, and/or sell
- * copies of the Software, and to permit persons to whom the Software is
- * furnished to do so, subject to the following conditions:
- *
- * The above copyright notice and this permission notice shall be included in all
- * copies or substantial portions of the Software.
- *
- * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
- * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
- * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE
- * AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER
- * LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM,
- * OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN THE
- * SOFTWARE.
- *
- *******************************************************************************/
-#ifndef GRIDWISE_SOFTMAX_HPP
-#define GRIDWISE_SOFTMAX_HPP
+#pragma once

-#include "reduction_common.hpp"
-#include "reduction_operator.hpp"
-#include "reduction_functions_accumulate.hpp"
-#include "reduction_functions_blockwise.hpp"
-#include "reduction_functions_threadwise.hpp"
-
-#include "threadwise_tensor_slice_transfer.hpp"
-#include "element_wise_operation.hpp"
+#include "ck/utility/data_type.hpp"
+#include "ck/utility/reduction_common.hpp"
+#include "ck/utility/reduction_operator.hpp"
+#include "ck/utility/reduction_functions_accumulate.hpp"
+#include "ck/tensor_operation/gpu/block/reduction_functions_blockwise.hpp"
+#include "ck/tensor_operation/gpu/thread/reduction_functions_threadwise.hpp"
+#include "ck/tensor_operation/gpu/thread/threadwise_tensor_slice_transfer.hpp"
+#include "ck/tensor_operation/gpu/element/element_wise_operation.hpp"

 namespace ck {

@@ -404,4 +378,3 @@ struct GridwiseSoftmax_mk_to_mk
 };

 } // namespace ck
-#endif // GRIDWISE_SOFTMAX_HPP
--- a/include/ck/tensor_operation/gpu/grid/gridwise_unary_elementwise_1d.hpp
+++ b/include/ck/tensor_operation/gpu/grid/gridwise_unary_elementwise_1d.hpp
@@ -1,9 +1,9 @@
 #pragma once

-#include "cluster_descriptor.hpp"
-#include "data_type.hpp"
-#include "element_wise_operation.hpp"
-#include "threadwise_tensor_slice_transfer.hpp"
+#include "ck/utility/data_type.hpp"
+#include "ck/tensor_description/cluster_descriptor.hpp"
+#include "ck/tensor_operation/gpu/thread/threadwise_tensor_slice_transfer.hpp"
+#include "ck/tensor_operation/gpu/element/element_wise_operation.hpp"

 namespace ck {

--- a/include/ck/tensor_operation/gpu/thread/reduction_functions_threadwise.hpp
+++ b/include/ck/tensor_operation/gpu/thread/reduction_functions_threadwise.hpp
@@ -1,32 +1,6 @@
-/*******************************************************************************
- *
- * MIT License
- *
- * Copyright (c) 2020 Advanced Micro Devices, Inc.
- *
- * Permission is hereby granted, free of charge, to any person obtaining a copy
- * of this software and associated documentation files (the "Software"), to deal
- * in the Software without restriction, including without limitation the rights
- * to use, copy, modify, merge, publish, distribute, sublicense, and/or sell
- * copies of the Software, and to permit persons to whom the Software is
- * furnished to do so, subject to the following conditions:
- *
- * The above copyright notice and this permission notice shall be included in all
- * copies or substantial portions of the Software.
- *
- * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
- * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
- * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE
- * AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER
- * LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM,
- * OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN THE
- * SOFTWARE.
- *
- *******************************************************************************/
-#ifndef CK_REDUCTION_FUNCTIONS_THREADWISE_HPP
-#define CK_REDUCTION_FUNCTIONS_THREADWISE_HPP
+#pragma once

-#include "reduction_functions_accumulate.hpp"
+#include "ck/utility/reduction_functions_accumulate.hpp"

 namespace ck {

@@ -117,6 +91,4 @@ struct ThreadwiseReductionWithIndex
    };
 };

-}; // end of namespace ck
-
-#endif
+} // namespace ck
--- a/include/ck/tensor_operation/gpu/thread/threadwise_contraction_dl.hpp
+++ b/include/ck/tensor_operation/gpu/thread/threadwise_contraction_dl.hpp
@@ -1,6 +1,7 @@
 #pragma once
-#include "common_header.hpp"
-#include "math.hpp"
+
+#include "ck/utility/common_header.hpp"
+#include "ck/utility/math.hpp"

 namespace ck {

--- a/include/ck/tensor_operation/gpu/thread/threadwise_tensor_slice_set.hpp
+++ b/include/ck/tensor_operation/gpu/thread/threadwise_tensor_slice_set.hpp
@@ -1,9 +1,8 @@
-#ifndef CK_THREADWISE_TENSOR_SET_HPP
-#define CK_THREADWISE_TENSOR_SET_HPP
+#pragma once

-#include "common_header.hpp"
-#include "tensor_descriptor.hpp"
-#include "tensor_descriptor_helper.hpp"
+#include "ck/utility/common_header.hpp"
+#include "ck/tensor_description/tensor_descriptor.hpp"
+#include "ck/tensor_description/tensor_descriptor_helper.hpp"

 namespace ck {

@@ -56,4 +55,3 @@ struct ThreadwiseTensorSliceSet_v1
 };

 } // namespace ck
-#endif
--- a/include/ck/tensor_operation/gpu/thread/threadwise_tensor_slice_transfer.hpp
+++ b/include/ck/tensor_operation/gpu/thread/threadwise_tensor_slice_transfer.hpp
@@ -1,10 +1,9 @@
-#ifndef CK_THREADWISE_TENSOR_SLICE_TRANSFER_HPP
-#define CK_THREADWISE_TENSOR_SLICE_TRANSFER_HPP
+#pragma once

-#include "common_header.hpp"
-#include "tensor_descriptor.hpp"
-#include "tensor_descriptor_helper.hpp"
-#include "tensor_space_filling_curve.hpp"
+#include "ck/utility/common_header.hpp"
+#include "ck/tensor_description/tensor_space_filling_curve.hpp"
+#include "ck/tensor_description/tensor_descriptor.hpp"
+#include "ck/tensor_description/tensor_descriptor_helper.hpp"

 namespace ck {

@@ -1168,4 +1167,3 @@ struct ThreadwiseTensorSliceTransfer_v4
 };

 } // namespace ck
-#endif
--- a/include/ck/tensor_operation/gpu/thread/threadwise_tensor_slice_transfer_v3r1.hpp
+++ b/include/ck/tensor_operation/gpu/thread/threadwise_tensor_slice_transfer_v3r1.hpp
@@ -1,10 +1,9 @@
-#ifndef CK_THREADWISE_TENSOR_SLICE_TRANSFER_V3R1_HPP
-#define CK_THREADWISE_TENSOR_SLICE_TRANSFER_V3R1_HPP
+#pragma once

-#include "common_header.hpp"
-#include "tensor_descriptor.hpp"
-#include "tensor_descriptor_helper.hpp"
-#include "static_tensor.hpp"
+#include "ck/utility/common_header.hpp"
+#include "ck/tensor_description/tensor_descriptor.hpp"
+#include "ck/tensor_description/tensor_descriptor_helper.hpp"
+#include "ck/tensor/static_tensor.hpp"

 namespace ck {

@@ -789,4 +788,3 @@ struct ThreadwiseTensorSliceTransfer_v3r1
 };

 } // namespace ck
-#endif
--- a/include/ck/tensor_operation/gpu/thread/threadwise_tensor_slice_transfer_v4r1.hpp
+++ b/include/ck/tensor_operation/gpu/thread/threadwise_tensor_slice_transfer_v4r1.hpp
@@ -1,9 +1,8 @@
-#ifndef CK_THREADWISE_TENSOR_SLICE_TRANSFER_V4R1_HPP
-#define CK_THREADWISE_TENSOR_SLICE_TRANSFER_V4R1_HPP
+#pragma once

-#include "common_header.hpp"
-#include "tensor_descriptor.hpp"
-#include "tensor_descriptor_helper.hpp"
+#include "ck/utility/common_header.hpp"
+#include "ck/tensor_description/tensor_descriptor.hpp"
+#include "ck/tensor_description/tensor_descriptor_helper.hpp"

 namespace ck {
 // Assume:
@@ -171,4 +170,3 @@ struct ThreadwiseTensorSliceTransfer_v4r1
 };

 } // namespace ck
-#endif
--- a/include/ck/tensor_operation/gpu/thread/threadwise_tensor_slice_transfer_v5r1.hpp
+++ b/include/ck/tensor_operation/gpu/thread/threadwise_tensor_slice_transfer_v5r1.hpp
@@ -1,8 +1,9 @@
 #pragma once

-#include "common_header.hpp"
-#include "tensor_descriptor.hpp"
-#include "tensor_descriptor_helper.hpp"
+#include "ck/utility/common_header.hpp"
+#include "ck/tensor_description/tensor_descriptor.hpp"
+#include "ck/tensor_description/tensor_descriptor_helper.hpp"
+#include "ck/tensor_description/tensor_space_filling_curve.hpp"

 namespace ck {

--- a/include/ck/tensor_operation/gpu/thread/threadwise_tensor_slice_transfer_v6r1.hpp
+++ b/include/ck/tensor_operation/gpu/thread/threadwise_tensor_slice_transfer_v6r1.hpp
@@ -1,10 +1,9 @@
-#ifndef CK_THREADWISE_TENSOR_SLICE_TRANSFER_V6R1_HPP
-#define CK_THREADWISE_TENSOR_SLICE_TRANSFER_V6R1_HPP
+#pragma once

-#include "common_header.hpp"
-#include "tensor_descriptor.hpp"
-#include "tensor_descriptor_helper.hpp"
-#include "tensor_space_filling_curve.hpp"
+#include "ck/utility/common_header.hpp"
+#include "ck/tensor_description/tensor_descriptor.hpp"
+#include "ck/tensor_description/tensor_descriptor_helper.hpp"
+#include "ck/tensor_description/tensor_space_filling_curve.hpp"

 namespace ck {

@@ -206,7 +205,6 @@ struct ThreadwiseTensorSliceTransfer_v6r1
    SrcCoord src_coord_;
    DstCoord dst_coord_;
    const ElementwiseOperation element_op_;
-}; // namespace ck
+};

 } // namespace ck
-#endif
--- a/include/ck/tensor_operation/gpu/thread/threadwise_tensor_slice_transfer_v6r2.hpp
+++ b/include/ck/tensor_operation/gpu/thread/threadwise_tensor_slice_transfer_v6r2.hpp
@@ -1,10 +1,9 @@
-#ifndef CK_THREADWISE_TENSOR_SLICE_TRANSFER_V6R2_HPP
-#define CK_THREADWISE_TENSOR_SLICE_TRANSFER_V6R2_HPP
+#pragma once

-#include "common_header.hpp"
-#include "tensor_descriptor.hpp"
-#include "tensor_descriptor_helper.hpp"
-#include "tensor_space_filling_curve.hpp"
+#include "ck/utility/common_header.hpp"
+#include "ck/tensor_description/tensor_descriptor.hpp"
+#include "ck/tensor_description/tensor_descriptor_helper.hpp"
+#include "ck/tensor_description/tensor_space_filling_curve.hpp"

 namespace ck {

@@ -256,4 +255,3 @@ struct ThreadwiseTensorSliceTransfer_v6r2
 };

 } // namespace ck
-#endif
--- a/include/ck/tensor_operation/gpu/thread/threadwise_tensor_slice_transfer_v6r3.hpp
+++ b/include/ck/tensor_operation/gpu/thread/threadwise_tensor_slice_transfer_v6r3.hpp
@@ -1,10 +1,9 @@
-#ifndef CK_THREADWISE_TENSOR_SLICE_TRANSFER_V6R3_HPP
-#define CK_THREADWISE_TENSOR_SLICE_TRANSFER_V6R3_HPP
+#pragma once

-#include "common_header.hpp"
-#include "tensor_descriptor.hpp"
-#include "tensor_descriptor_helper.hpp"
-#include "tensor_space_filling_curve.hpp"
+#include "ck/utility/common_header.hpp"
+#include "ck/tensor_description/tensor_descriptor.hpp"
+#include "ck/tensor_description/tensor_descriptor_helper.hpp"
+#include "ck/tensor_description/tensor_space_filling_curve.hpp"

 namespace ck {

@@ -306,4 +305,3 @@ struct ThreadwiseTensorSliceTransfer_v6r3
 };

 } // namespace ck
-#endif
--- a/include/ck/tensor_operation/gpu/thread/threadwise_tensor_slice_transfer_v7.hpp
+++ b/include/ck/tensor_operation/gpu/thread/threadwise_tensor_slice_transfer_v7.hpp
@@ -1,9 +1,9 @@
 #pragma once

-#include "common_header.hpp"
-#include "tensor_descriptor.hpp"
-#include "tensor_descriptor_helper.hpp"
-#include "tensor_space_filling_curve.hpp"
+#include "ck/utility/common_header.hpp"
+#include "ck/tensor_description/tensor_descriptor.hpp"
+#include "ck/tensor_description/tensor_descriptor_helper.hpp"
+#include "ck/tensor_description/tensor_space_filling_curve.hpp"

 namespace ck {

--- a/include/ck/tensor_operation/gpu/warp/xdlops_gemm.hpp
+++ b/include/ck/tensor_operation/gpu/warp/xdlops_gemm.hpp
@@ -1,9 +1,8 @@
-#ifndef CK_XDLOPS_GEMM_HPP
-#define CK_XDLOPS_GEMM_HPP
+#pragma once

-#include "common_header.hpp"
-#include "math.hpp"
-#include "amd_xdlops.hpp"
+#include "ck/utility/common_header.hpp"
+#include "ck/utility/math.hpp"
+#include "ck/utility/amd_xdlops.hpp"

 namespace ck {

@@ -786,4 +785,3 @@ struct XdlopsGemm
 };

 } // namespace ck
-#endif