Merge remote-tracking branch 'origin/develop' into contraction

library/src/tensor_operation_instance/gpu/convnd_bwd_data/device_conv3d_bwd_data_xdl_ndhwc_kzyxc_ndhwk_f32_instance.cpp

-#include <stdlib.h>
-#include "config.hpp"
-#include "device_convnd_bwd_data_xdl_ndhwc_kzyxc_ndhwk.hpp"
-#include "element_wise_operation.hpp"
-#include "device_operation_instance.hpp"
+// SPDX-License-Identifier: MIT
+// Copyright (c) 2018-2022, Advanced Micro Devices, Inc. All rights reserved.
+
+#include <cstdlib>
+
+#include "ck/ck.hpp"
+#include "ck/tensor_operation/gpu/device/tensor_layout.hpp"
+#include "ck/tensor_operation/gpu/device/device_convnd_bwd_data_xdl_ndhwc_kzyxc_ndhwk.hpp"
+#include "ck/tensor_operation/gpu/element/element_wise_operation.hpp"
+#include "ck/library/tensor_operation_instance/device_operation_instance.hpp"
 
 namespace ck {
 namespace tensor_operation {

library/src/tensor_operation_instance/gpu/convnd_bwd_data/device_conv3d_bwd_data_xdl_ndhwc_kzyxc_ndhwk_int8_instance.cpp

-#include <stdlib.h>
-#include "config.hpp"
-#include "device_convnd_bwd_data_xdl_ndhwc_kzyxc_ndhwk.hpp"
-#include "element_wise_operation.hpp"
-#include "device_operation_instance.hpp"
+// SPDX-License-Identifier: MIT
+// Copyright (c) 2018-2022, Advanced Micro Devices, Inc. All rights reserved.
+
+#include <cstdlib>
+
+#include "ck/ck.hpp"
+#include "ck/tensor_operation/gpu/device/tensor_layout.hpp"
+#include "ck/tensor_operation/gpu/device/device_convnd_bwd_data_xdl_ndhwc_kzyxc_ndhwk.hpp"
+#include "ck/tensor_operation/gpu/element/element_wise_operation.hpp"
+#include "ck/library/tensor_operation_instance/device_operation_instance.hpp"
 
 namespace ck {
 namespace tensor_operation {
@@ -33,13 +38,11 @@ using device_conv3d_bwd_data_xdl_ndhwc_kzyxc_ndhwk_int8_instances =
         DeviceConvndBwdDataXdl_Input_N_Di_Hi_Wi_C_Weight_K_Z_Y_X_C_Output_N_Do_Ho_Wo_K<  DataType,  DataType,  DataType,    AccType, PassThrough, PassThrough, PassThrough,  ConvBwdDataDefault,   3,     256,   256,   128,     4,  16,   32,   32,    4,    2,     S<4, 64, 1>,     S<1, 0, 2>,     S<1, 0, 2>,              2,             16,             16,      true,     S<4, 64, 1>,     S<2, 0, 1>,     S<0, 2, 1>,             1,              2,             16,      true,               7,               1>,
         DeviceConvndBwdDataXdl_Input_N_Di_Hi_Wi_C_Weight_K_Z_Y_X_C_Output_N_Do_Ho_Wo_K<  DataType,  DataType,  DataType,    AccType, PassThrough, PassThrough, PassThrough,  ConvBwdDataDefault,   3,     256,   128,   256,     4,  16,   32,   32,    2,    4,     S<4, 64, 1>,     S<1, 0, 2>,     S<1, 0, 2>,              2,             16,             16,      true,     S<4, 64, 1>,     S<2, 0, 1>,     S<0, 2, 1>,             1,              4,             16,      true,               7,               1>,
         DeviceConvndBwdDataXdl_Input_N_Di_Hi_Wi_C_Weight_K_Z_Y_X_C_Output_N_Do_Ho_Wo_K<  DataType,  DataType,  DataType,    AccType, PassThrough, PassThrough, PassThrough,  ConvBwdDataDefault,   3,     128,   128,   128,     4,  16,   32,   32,    4,    2,     S<4, 32, 1>,     S<1, 0, 2>,     S<1, 0, 2>,              2,             16,             16,      true,     S<4, 32, 1>,     S<2, 0, 1>,     S<0, 2, 1>,             1,              2,             16,      true,               7,               1>,
-#if 1
         DeviceConvndBwdDataXdl_Input_N_Di_Hi_Wi_C_Weight_K_Z_Y_X_C_Output_N_Do_Ho_Wo_K<  DataType,  DataType,  DataType,    AccType, PassThrough, PassThrough, PassThrough,  ConvBwdDataDefault,   3,     256,   128,   128,     4,  16,   32,   32,    2,    2,     S<4, 64, 1>,     S<1, 0, 2>,     S<1, 0, 2>,              2,             16,             16,      true,     S<4, 64, 1>,     S<2, 0, 1>,     S<0, 2, 1>,             1,              2,             16,      true,               7,               1>,
         DeviceConvndBwdDataXdl_Input_N_Di_Hi_Wi_C_Weight_K_Z_Y_X_C_Output_N_Do_Ho_Wo_K<  DataType,  DataType,  DataType,    AccType, PassThrough, PassThrough, PassThrough,  ConvBwdDataDefault,   3,     128,    64,   128,     4,  16,   32,   32,    2,    2,     S<4, 32, 1>,     S<1, 0, 2>,     S<1, 0, 2>,              2,             16,             16,      true,     S<4, 32, 1>,     S<2, 0, 1>,     S<0, 2, 1>,             1,              4,             16,      true,               7,               1>,
         DeviceConvndBwdDataXdl_Input_N_Di_Hi_Wi_C_Weight_K_Z_Y_X_C_Output_N_Do_Ho_Wo_K<  DataType,  DataType,  DataType,    AccType, PassThrough, PassThrough, PassThrough,  ConvBwdDataDefault,   3,     256,    64,   128,     4,  16,   32,   32,    1,    2,     S<4, 64, 1>,     S<1, 0, 2>,     S<1, 0, 2>,              2,             16,             16,      true,     S<4, 64, 1>,     S<2, 0, 1>,     S<0, 2, 1>,             1,              2,             16,      true,               7,               1>,
         DeviceConvndBwdDataXdl_Input_N_Di_Hi_Wi_C_Weight_K_Z_Y_X_C_Output_N_Do_Ho_Wo_K<  DataType,  DataType,  DataType,    AccType, PassThrough, PassThrough, PassThrough,  ConvBwdDataDefault,   3,     128,    32,   128,     4,  16,   32,   32,    1,    2,     S<4, 32, 1>,     S<1, 0, 2>,     S<1, 0, 2>,              2,             16,             16,      true,     S<4, 32, 1>,     S<2, 0, 1>,     S<0, 2, 1>,             1,              4,             16,      true,               7,               1>,
         DeviceConvndBwdDataXdl_Input_N_Di_Hi_Wi_C_Weight_K_Z_Y_X_C_Output_N_Do_Ho_Wo_K<  DataType,  DataType,  DataType,    AccType, PassThrough, PassThrough, PassThrough,  ConvBwdDataDefault,   3,      64,    32,    64,     4,  16,   32,   32,    1,    2,     S<4, 16, 1>,     S<1, 0, 2>,     S<1, 0, 2>,              2,             16,             16,      true,     S<4, 16, 1>,     S<2, 0, 1>,     S<0, 2, 1>,             1,              4,             16,      true,               7,               1>,
-#endif
         DeviceConvndBwdDataXdl_Input_N_Di_Hi_Wi_C_Weight_K_Z_Y_X_C_Output_N_Do_Ho_Wo_K<  DataType,  DataType,  DataType,    AccType, PassThrough, PassThrough, PassThrough,  ConvBwdDataDefault,   3,     128,   128,    64,     4,  16,   32,   32,    2,    2,     S<4, 32, 1>,     S<1, 0, 2>,     S<1, 0, 2>,              2,             16,             16,      true,     S<4, 32, 1>,     S<2, 0, 1>,     S<0, 2, 1>,             1,              2,             16,      true,               7,               1>,
         DeviceConvndBwdDataXdl_Input_N_Di_Hi_Wi_C_Weight_K_Z_Y_X_C_Output_N_Do_Ho_Wo_K<  DataType,  DataType,  DataType,    AccType, PassThrough, PassThrough, PassThrough,  ConvBwdDataDefault,   3,      64,    64,    64,     4,  16,   32,   32,    2,    2,     S<4, 16, 1>,     S<1, 0, 2>,     S<1, 0, 2>,              2,             16,             16,      true,     S<4, 16, 1>,     S<2, 0, 1>,     S<0, 2, 1>,             1,              4,             16,      true,               7,               1>,
         DeviceConvndBwdDataXdl_Input_N_Di_Hi_Wi_C_Weight_K_Z_Y_X_C_Output_N_Do_Ho_Wo_K<  DataType,  DataType,  DataType,    AccType, PassThrough, PassThrough, PassThrough,  ConvBwdDataDefault,   3,     256,   128,    64,     4,  16,   32,   32,    2,    1,     S<4, 64, 1>,     S<1, 0, 2>,     S<1, 0, 2>,              2,             16,             16,      true,     S<4, 64, 1>,     S<2, 0, 1>,     S<0, 2, 1>,             1,              1,             16,      true,               7,               1>,

library/src/tensor_operation_instance/gpu/device_conv2d.cpp

-#include <stdlib.h>
-#include "config.hpp"
-#include "device_conv2d_fwd_xdl_c_shuffle_nhwc_kyxc_nhwk.hpp"
-#include "element_wise_operation.hpp"
-#include "device_operation_instance.hpp"
-#include "host_interface.hpp"
-
-namespace ck {
-namespace tensor_operation {
-namespace device {
-namespace device_conv2d_fwd_instance {
-using PassThrough = ck::tensor_operation::element_wise::PassThrough;
-void add_device_conv2d_fwd_xdl_c_shuffle_nhwc_kyxc_nhwk_f16_instances(
-    std::vector<DeviceConvFwdPtr<PassThrough, PassThrough, PassThrough>>& instances);
-void add_device_conv2d_fwd_xdl_nhwc_kyxc_nhwk_f32_instances(
-    std::vector<DeviceConvFwdPtr<PassThrough, PassThrough, PassThrough>>& instances);
-void add_device_conv2d_fwd_xdl_nhwc_kyxc_nhwk_bf16_instances(
-    std::vector<DeviceConvFwdPtr<PassThrough, PassThrough, PassThrough>>& instances);
-void add_device_conv2d_fwd_xdl_nhwc_kyxc_nhwk_f16_instances(
-    std::vector<DeviceConvFwdPtr<PassThrough, PassThrough, PassThrough>>& instances);
-void add_device_conv2d_fwd_xdl_nhwc_kyxc_nhwk_int8_instances(
-    std::vector<DeviceConvFwdPtr<PassThrough, PassThrough, PassThrough>>& instances);
-
-} // namespace device_conv2d_fwd_instance
-} // namespace device
-} // namespace tensor_operation
-} // namespace ck
-
-using PassThrough = ck::tensor_operation::element_wise::PassThrough;
-struct DeviceConvFwdPtr_t::DeviceConvFwdPtrImpl
-{
-    std::unique_ptr<DeviceConvFwdPtr_t::BaseArgument>
-    MakeArgumentPointer(void* in_ptr,
-                        void* wei_ptr,
-                        void* out_ptr,
-                        size_t N,
-                        size_t K,
-                        size_t C,
-                        std::vector<ck::index_t> input_spatial_lengths,
-                        std::vector<ck::index_t> filter_spatial_lengths,
-                        std::vector<ck::index_t> output_spatial_lengths,
-                        std::vector<ck::index_t> conv_filter_strides,
-                        std::vector<ck::index_t> conv_filter_dilations,
-                        std::vector<ck::index_t> input_left_pads,
-                        std::vector<ck::index_t> input_right_pads) const
-    {
-        return el->MakeArgumentPointer(in_ptr,
-                                       wei_ptr,
-                                       out_ptr,
-                                       N,
-                                       K,
-                                       C,
-                                       input_spatial_lengths,
-                                       filter_spatial_lengths,
-                                       output_spatial_lengths,
-                                       conv_filter_strides,
-                                       conv_filter_dilations,
-                                       input_left_pads,
-                                       input_right_pads,
-                                       PassThrough{},
-                                       PassThrough{},
-                                       PassThrough{});
-    }
-    std::unique_ptr<DeviceConvFwdPtr_t::BaseInvoker> MakeInvokerPointer() const
-    {
-        return el->MakeInvokerPointer();
-    }
-
-    std::string GetTypeString() { return el->GetTypeString(); }
-    bool IsSupportedArgument(const DeviceConvFwdPtr_t::BaseArgument* arg)
-    {
-        return el->IsSupportedArgument(arg);
-    }
-
-    ck::tensor_operation::device::DeviceConvFwdPtr<PassThrough, PassThrough, PassThrough> el;
-};
-
-DeviceConvFwdPtr_t::DeviceConvFwdPtr_t() : pImpl(nullptr) {}
-DeviceConvFwdPtr_t::~DeviceConvFwdPtr_t()                    = default;
-DeviceConvFwdPtr_t::DeviceConvFwdPtr_t(DeviceConvFwdPtr_t&&) = default;
-DeviceConvFwdPtr_t::DeviceConvFwdPtr_t(DeviceConvFwdPtr_t::DeviceConvFwdPtrImpl& other)
-    : pImpl(std::make_unique<DeviceConvFwdPtr_t::DeviceConvFwdPtrImpl>(std::move(other)))
-{
-}
-
-std::unique_ptr<DeviceConvFwdPtr_t::BaseArgument>
-DeviceConvFwdPtr_t::MakeArgumentPointer(void* in_ptr,
-                                        void* wei_ptr,
-                                        void* out_ptr,
-                                        size_t N,
-                                        size_t K,
-                                        size_t C,
-                                        std::vector<ck::index_t> input_spatial_lengths,
-                                        std::vector<ck::index_t> filter_spatial_lengths,
-                                        std::vector<ck::index_t> output_spatial_lengths,
-                                        std::vector<ck::index_t> conv_filter_strides,
-                                        std::vector<ck::index_t> conv_filter_dilations,
-                                        std::vector<ck::index_t> input_left_pads,
-                                        std::vector<ck::index_t> input_right_pads) const
-{
-    return pImpl->MakeArgumentPointer(in_ptr,
-                                      wei_ptr,
-                                      out_ptr,
-                                      N,
-                                      K,
-                                      C,
-                                      input_spatial_lengths,
-                                      filter_spatial_lengths,
-                                      output_spatial_lengths,
-                                      conv_filter_strides,
-                                      conv_filter_dilations,
-                                      input_left_pads,
-                                      input_right_pads);
-}
-
-std::unique_ptr<DeviceConvFwdPtr_t::BaseInvoker> DeviceConvFwdPtr_t::MakeInvokerPointer() const
-{
-    return pImpl->MakeInvokerPointer();
-}
-
-std::string DeviceConvFwdPtr_t::GetTypeString() { return pImpl->GetTypeString(); }
-bool DeviceConvFwdPtr_t::IsSupportedArgument(const DeviceConvFwdPtr_t::BaseArgument* arg_ptr)
-{
-    return pImpl->IsSupportedArgument(arg_ptr);
-}
-
-using namespace ck::tensor_operation::device::device_conv2d_fwd_instance;
-void add_device_conv2d_fwd_xdl_c_shuffle_nhwc_kyxc_nhwk_f16_instances_t(
-    std::vector<DeviceConvFwdPtr_t>& instances)
-{
-    std::vector<
-        ck::tensor_operation::device::DeviceConvFwdPtr<PassThrough, PassThrough, PassThrough>>
-        local_instances;
-    add_device_conv2d_fwd_xdl_c_shuffle_nhwc_kyxc_nhwk_f16_instances(local_instances);
-    for(auto& kinder : local_instances)
-    {
-        DeviceConvFwdPtr_t::DeviceConvFwdPtrImpl tmp{std::move(kinder)};
-        instances.emplace_back(tmp);
-    }
-    return;
-}
-
-void add_device_conv2d_fwd_xdl_nhwc_kyxc_nhwk_f32_instances_t(
-    std::vector<DeviceConvFwdPtr_t>& instances)
-{
-    std::vector<
-        ck::tensor_operation::device::DeviceConvFwdPtr<PassThrough, PassThrough, PassThrough>>
-        local_instances;
-    add_device_conv2d_fwd_xdl_nhwc_kyxc_nhwk_f32_instances(local_instances);
-    for(auto& kinder : local_instances)
-    {
-        DeviceConvFwdPtr_t::DeviceConvFwdPtrImpl tmp{std::move(kinder)};
-        instances.emplace_back(tmp); // Perhaps we can do better
-    }
-    return;
-}
-
-void add_device_conv2d_fwd_xdl_nhwc_kyxc_nhwk_bf16_instances_t(
-    std::vector<DeviceConvFwdPtr_t>& instances)
-{
-    std::vector<
-        ck::tensor_operation::device::DeviceConvFwdPtr<PassThrough, PassThrough, PassThrough>>
-        local_instances;
-    add_device_conv2d_fwd_xdl_nhwc_kyxc_nhwk_bf16_instances(local_instances);
-    for(auto& kinder : local_instances)
-    {
-        DeviceConvFwdPtr_t::DeviceConvFwdPtrImpl tmp{std::move(kinder)};
-        instances.emplace_back(tmp); // Perhaps we can do better
-    }
-    return;
-}
-
-void add_device_conv2d_fwd_xdl_nhwc_kyxc_nhwk_f16_instances_t(
-    std::vector<DeviceConvFwdPtr_t>& instances)
-{
-    std::vector<
-        ck::tensor_operation::device::DeviceConvFwdPtr<PassThrough, PassThrough, PassThrough>>
-        local_instances;
-    add_device_conv2d_fwd_xdl_nhwc_kyxc_nhwk_f16_instances(local_instances);
-    for(auto& kinder : local_instances)
-    {
-        DeviceConvFwdPtr_t::DeviceConvFwdPtrImpl tmp{std::move(kinder)};
-        instances.emplace_back(tmp); // Perhaps we can do better
-    }
-    return;
-}
-
-void add_device_conv2d_fwd_xdl_nhwc_kyxc_nhwk_int8_instances_t(
-    std::vector<DeviceConvFwdPtr_t>& instances)
-{
-    std::vector<
-        ck::tensor_operation::device::DeviceConvFwdPtr<PassThrough, PassThrough, PassThrough>>
-        local_instances;
-    add_device_conv2d_fwd_xdl_nhwc_kyxc_nhwk_int8_instances(local_instances);
-    for(auto& kinder : local_instances)
-    {
-        DeviceConvFwdPtr_t::DeviceConvFwdPtrImpl tmp{std::move(kinder)};
-        instances.emplace_back(tmp);
-    }
-    return;
-}

library/src/tensor_operation_instance/gpu/gemm/device_gemm_dl_f16_f16_f16_km_kn_mn_instance.cpp

-#include <stdlib.h>
-#include "config.hpp"
-#include "device_gemm_dl.hpp"
-#include "element_wise_operation.hpp"
-#include "device_operation_instance.hpp"
+// SPDX-License-Identifier: MIT
+// Copyright (c) 2018-2022, Advanced Micro Devices, Inc. All rights reserved.
+
+#include <cstdlib>
+
+#include "ck/ck.hpp"
+#include "ck/tensor_operation/gpu/device/tensor_layout.hpp"
+#include "ck/tensor_operation/gpu/device/gemm_specialization.hpp"
+#include "ck/tensor_operation/gpu/device/device_gemm_dl.hpp"
+#include "ck/library/tensor_operation_instance/device_operation_instance.hpp"
 
 namespace ck {
 namespace tensor_operation {

library/src/tensor_operation_instance/gpu/gemm/device_gemm_dl_f16_f16_f16_km_nk_mn_instance.cpp

-#include <stdlib.h>
-#include "config.hpp"
-#include "device_gemm_dl.hpp"
-#include "element_wise_operation.hpp"
-#include "device_operation_instance.hpp"
+// SPDX-License-Identifier: MIT
+// Copyright (c) 2018-2022, Advanced Micro Devices, Inc. All rights reserved.
+
+#include <cstdlib>
+
+#include "ck/ck.hpp"
+#include "ck/tensor_operation/gpu/device/tensor_layout.hpp"
+#include "ck/tensor_operation/gpu/device/gemm_specialization.hpp"
+#include "ck/tensor_operation/gpu/device/device_gemm_dl.hpp"
+#include "ck/library/tensor_operation_instance/device_operation_instance.hpp"
 
 namespace ck {
 namespace tensor_operation {

library/src/tensor_operation_instance/gpu/gemm/device_gemm_dl_f16_f16_f16_mk_kn_mn_instance.cpp

-#include <stdlib.h>
-#include "config.hpp"
-#include "device_gemm_dl.hpp"
-#include "element_wise_operation.hpp"
-#include "device_operation_instance.hpp"
+// SPDX-License-Identifier: MIT
+// Copyright (c) 2018-2022, Advanced Micro Devices, Inc. All rights reserved.
+
+#include <cstdlib>
+
+#include "ck/ck.hpp"
+#include "ck/tensor_operation/gpu/device/tensor_layout.hpp"
+#include "ck/tensor_operation/gpu/device/gemm_specialization.hpp"
+#include "ck/tensor_operation/gpu/device/device_gemm_dl.hpp"
+#include "ck/library/tensor_operation_instance/device_operation_instance.hpp"
 
 namespace ck {
 namespace tensor_operation {

library/src/tensor_operation_instance/gpu/gemm/device_gemm_dl_f16_f16_f16_mk_nk_mn_instance.cpp

-#include <stdlib.h>
-#include "config.hpp"
-#include "device_gemm_dl.hpp"
-#include "element_wise_operation.hpp"
-#include "device_operation_instance.hpp"
+// SPDX-License-Identifier: MIT
+// Copyright (c) 2018-2022, Advanced Micro Devices, Inc. All rights reserved.
+
+#include <cstdlib>
+
+#include "ck/ck.hpp"
+#include "ck/tensor_operation/gpu/device/tensor_layout.hpp"
+#include "ck/tensor_operation/gpu/device/gemm_specialization.hpp"
+#include "ck/tensor_operation/gpu/device/device_gemm_dl.hpp"
+#include "ck/library/tensor_operation_instance/device_operation_instance.hpp"
 
 namespace ck {
 namespace tensor_operation {

library/src/tensor_operation_instance/gpu/gemm/device_gemm_dl_f32_f32_f32_km_kn_mn_instance.cpp

-#include <stdlib.h>
-#include "config.hpp"
-#include "device_gemm_dl.hpp"
-#include "element_wise_operation.hpp"
-#include "device_operation_instance.hpp"
+// SPDX-License-Identifier: MIT
+// Copyright (c) 2018-2022, Advanced Micro Devices, Inc. All rights reserved.
+
+#include <cstdlib>
+
+#include "ck/ck.hpp"
+#include "ck/tensor_operation/gpu/device/tensor_layout.hpp"
+#include "ck/tensor_operation/gpu/device/gemm_specialization.hpp"
+#include "ck/tensor_operation/gpu/device/device_gemm_dl.hpp"
+#include "ck/library/tensor_operation_instance/device_operation_instance.hpp"
 
 namespace ck {
 namespace tensor_operation {

library/src/tensor_operation_instance/gpu/gemm/device_gemm_dl_f32_f32_f32_km_nk_mn_instance.cpp

-#include <stdlib.h>
-#include "config.hpp"
-#include "device_gemm_dl.hpp"
-#include "element_wise_operation.hpp"
-#include "device_operation_instance.hpp"
+// SPDX-License-Identifier: MIT
+// Copyright (c) 2018-2022, Advanced Micro Devices, Inc. All rights reserved.
+
+#include <cstdlib>
+
+#include "ck/ck.hpp"
+#include "ck/tensor_operation/gpu/device/tensor_layout.hpp"
+#include "ck/tensor_operation/gpu/device/gemm_specialization.hpp"
+#include "ck/tensor_operation/gpu/device/device_gemm_dl.hpp"
+#include "ck/library/tensor_operation_instance/device_operation_instance.hpp"
 
 namespace ck {
 namespace tensor_operation {

library/src/tensor_operation_instance/gpu/gemm/device_gemm_dl_f32_f32_f32_mk_kn_mn_instance.cpp

-#include <stdlib.h>
-#include "config.hpp"
-#include "device_gemm_dl.hpp"
-#include "element_wise_operation.hpp"
-#include "device_operation_instance.hpp"
+// SPDX-License-Identifier: MIT
+// Copyright (c) 2018-2022, Advanced Micro Devices, Inc. All rights reserved.
+
+#include <cstdlib>
+
+#include "ck/ck.hpp"
+#include "ck/tensor_operation/gpu/device/tensor_layout.hpp"
+#include "ck/tensor_operation/gpu/device/gemm_specialization.hpp"
+#include "ck/tensor_operation/gpu/device/device_gemm_dl.hpp"
+#include "ck/library/tensor_operation_instance/device_operation_instance.hpp"
 
 namespace ck {
 namespace tensor_operation {

library/src/tensor_operation_instance/gpu/gemm/device_gemm_dl_f32_f32_f32_mk_nk_mn_instance.cpp

-#include <stdlib.h>
-#include "config.hpp"
-#include "device_gemm_dl.hpp"
-#include "element_wise_operation.hpp"
-#include "device_operation_instance.hpp"
+// SPDX-License-Identifier: MIT
+// Copyright (c) 2018-2022, Advanced Micro Devices, Inc. All rights reserved.
+
+#include <cstdlib>
+
+#include "ck/ck.hpp"
+#include "ck/tensor_operation/gpu/device/tensor_layout.hpp"
+#include "ck/tensor_operation/gpu/device/gemm_specialization.hpp"
+#include "ck/tensor_operation/gpu/device/device_gemm_dl.hpp"
+#include "ck/library/tensor_operation_instance/device_operation_instance.hpp"
 
 namespace ck {
 namespace tensor_operation {

library/src/tensor_operation_instance/gpu/gemm/device_gemm_dl_i8_i8_i8_km_kn_mn_instance.cpp

-#include <stdlib.h>
-#include "config.hpp"
-#include "device_gemm_dl.hpp"
-#include "element_wise_operation.hpp"
-#include "device_operation_instance.hpp"
+// SPDX-License-Identifier: MIT
+// Copyright (c) 2018-2022, Advanced Micro Devices, Inc. All rights reserved.
+
+#include <cstdlib>
+
+#include "ck/ck.hpp"
+#include "ck/tensor_operation/gpu/device/tensor_layout.hpp"
+#include "ck/tensor_operation/gpu/device/gemm_specialization.hpp"
+#include "ck/tensor_operation/gpu/device/device_gemm_dl.hpp"
+#include "ck/library/tensor_operation_instance/device_operation_instance.hpp"
 
 namespace ck {
 namespace tensor_operation {

library/src/tensor_operation_instance/gpu/gemm/device_gemm_dl_i8_i8_i8_km_nk_mn_instance.cpp

-#include <stdlib.h>
-#include "config.hpp"
-#include "device_gemm_dl.hpp"
-#include "element_wise_operation.hpp"
-#include "device_operation_instance.hpp"
+// SPDX-License-Identifier: MIT
+// Copyright (c) 2018-2022, Advanced Micro Devices, Inc. All rights reserved.
+
+#include <cstdlib>
+
+#include "ck/ck.hpp"
+#include "ck/tensor_operation/gpu/device/tensor_layout.hpp"
+#include "ck/tensor_operation/gpu/device/gemm_specialization.hpp"
+#include "ck/tensor_operation/gpu/device/device_gemm_dl.hpp"
+#include "ck/library/tensor_operation_instance/device_operation_instance.hpp"
 
 namespace ck {
 namespace tensor_operation {

library/src/tensor_operation_instance/gpu/gemm/device_gemm_dl_i8_i8_i8_mk_kn_mn_instance.cpp

-#include <stdlib.h>
-#include "config.hpp"
-#include "device_gemm_dl.hpp"
-#include "element_wise_operation.hpp"
-#include "device_operation_instance.hpp"
+// SPDX-License-Identifier: MIT
+// Copyright (c) 2018-2022, Advanced Micro Devices, Inc. All rights reserved.
+
+#include <cstdlib>
+
+#include "ck/ck.hpp"
+#include "ck/tensor_operation/gpu/device/tensor_layout.hpp"
+#include "ck/tensor_operation/gpu/device/gemm_specialization.hpp"
+#include "ck/tensor_operation/gpu/device/device_gemm_dl.hpp"
+#include "ck/library/tensor_operation_instance/device_operation_instance.hpp"
 
 namespace ck {
 namespace tensor_operation {

library/src/tensor_operation_instance/gpu/gemm/device_gemm_dl_i8_i8_i8_mk_nk_mn_instance.cpp

-#include <stdlib.h>
-#include "config.hpp"
-#include "device_gemm_dl.hpp"
-#include "element_wise_operation.hpp"
-#include "device_operation_instance.hpp"
+// SPDX-License-Identifier: MIT
+// Copyright (c) 2018-2022, Advanced Micro Devices, Inc. All rights reserved.
+
+#include <cstdlib>
+
+#include "ck/ck.hpp"
+#include "ck/tensor_operation/gpu/device/tensor_layout.hpp"
+#include "ck/tensor_operation/gpu/device/gemm_specialization.hpp"
+#include "ck/tensor_operation/gpu/device/device_gemm_dl.hpp"
+#include "ck/library/tensor_operation_instance/device_operation_instance.hpp"
 
 namespace ck {
 namespace tensor_operation {

library/src/tensor_operation_instance/gpu/gemm/device_gemm_xdl_c_shuffle_2_stage_f16_f16_f16_mk_nk_mn_instance.cpp

-#include <stdlib.h>
-#include "config.hpp"
-#include "device_gemm_xdl_cshuffle.hpp"
-#include "element_wise_operation.hpp"
-#include "device_operation_instance.hpp"
+// SPDX-License-Identifier: MIT
+// Copyright (c) 2018-2022, Advanced Micro Devices, Inc. All rights reserved.
+
+#include <cstdlib>
+
+#include "ck/ck.hpp"
+#include "ck/tensor_operation/gpu/device/tensor_layout.hpp"
+#include "ck/tensor_operation/gpu/device/gemm_specialization.hpp"
+#include "ck/tensor_operation/gpu/device/device_gemm_xdl_cshuffle.hpp"
+#include "ck/library/tensor_operation_instance/device_operation_instance.hpp"
 
 namespace ck {
 namespace tensor_operation {

library/src/tensor_operation_instance/gpu/gemm/device_gemm_xdl_c_shuffle_bf16_bf16_bf16_km_kn_mn_instance.cpp

-#include <stdlib.h>
-#include "config.hpp"
-#include "device_gemm_xdl_cshuffle.hpp"
-#include "element_wise_operation.hpp"
-#include "device_operation_instance.hpp"
+// SPDX-License-Identifier: MIT
+// Copyright (c) 2018-2022, Advanced Micro Devices, Inc. All rights reserved.
+
+#include <cstdlib>
+
+#include "ck/ck.hpp"
+#include "ck/tensor_operation/gpu/device/tensor_layout.hpp"
+#include "ck/tensor_operation/gpu/device/gemm_specialization.hpp"
+#include "ck/tensor_operation/gpu/device/device_gemm_xdl_cshuffle.hpp"
+#include "ck/library/tensor_operation_instance/device_operation_instance.hpp"
 
 namespace ck {
 namespace tensor_operation {

library/src/tensor_operation_instance/gpu/gemm/device_gemm_xdl_c_shuffle_bf16_bf16_bf16_km_nk_mn_instance.cpp

-#include <stdlib.h>
-#include "config.hpp"
-#include "device_gemm_xdl_cshuffle.hpp"
-#include "element_wise_operation.hpp"
-#include "device_operation_instance.hpp"
+// SPDX-License-Identifier: MIT
+// Copyright (c) 2018-2022, Advanced Micro Devices, Inc. All rights reserved.
+
+#include <cstdlib>
+
+#include "ck/ck.hpp"
+#include "ck/tensor_operation/gpu/device/tensor_layout.hpp"
+#include "ck/tensor_operation/gpu/device/gemm_specialization.hpp"
+#include "ck/tensor_operation/gpu/device/device_gemm_xdl_cshuffle.hpp"
+#include "ck/library/tensor_operation_instance/device_operation_instance.hpp"
 
 namespace ck {
 namespace tensor_operation {

library/src/tensor_operation_instance/gpu/gemm/device_gemm_xdl_c_shuffle_bf16_bf16_bf16_mk_kn_mn_instance.cpp

-#include <stdlib.h>
-#include "config.hpp"
-#include "device_gemm_xdl_cshuffle.hpp"
-#include "element_wise_operation.hpp"
-#include "device_operation_instance.hpp"
+// SPDX-License-Identifier: MIT
+// Copyright (c) 2018-2022, Advanced Micro Devices, Inc. All rights reserved.
+
+#include <cstdlib>
+
+#include "ck/ck.hpp"
+#include "ck/tensor_operation/gpu/device/tensor_layout.hpp"
+#include "ck/tensor_operation/gpu/device/gemm_specialization.hpp"
+#include "ck/tensor_operation/gpu/device/device_gemm_xdl_cshuffle.hpp"
+#include "ck/library/tensor_operation_instance/device_operation_instance.hpp"
 
 namespace ck {
 namespace tensor_operation {

library/src/tensor_operation_instance/gpu/gemm/device_gemm_xdl_c_shuffle_bf16_bf16_bf16_mk_nk_mn_instance.cpp

-#include <stdlib.h>
-#include "config.hpp"
-#include "device_gemm_xdl_cshuffle.hpp"
-#include "element_wise_operation.hpp"
-#include "device_operation_instance.hpp"
+// SPDX-License-Identifier: MIT
+// Copyright (c) 2018-2022, Advanced Micro Devices, Inc. All rights reserved.
+
+#include <cstdlib>
+
+#include "ck/ck.hpp"
+#include "ck/tensor_operation/gpu/device/tensor_layout.hpp"
+#include "ck/tensor_operation/gpu/device/gemm_specialization.hpp"
+#include "ck/tensor_operation/gpu/device/device_gemm_xdl_cshuffle.hpp"
+#include "ck/library/tensor_operation_instance/device_operation_instance.hpp"
 
 namespace ck {
 namespace tensor_operation {