Merge branch 'develop' into gridwise_gemm_double_buffer

7506342c · ltqin · 26b4fe97 · acbd7bd7 · 7506342c · 7506342c
Commit 7506342c authored Jan 12, 2022 by ltqin
20 changed files
--- a/composable_kernel/include/tensor_operation/threadwise_tensor_slice_transfer_v3r2.hpp
+++ b/composable_kernel/include/tensor_operation/threadwise_tensor_slice_transfer_v3r2.hpp
--- a/composable_kernel/include/tensor_operation/threadwise_tensor_slice_transfer_v3r3.hpp
+++ b/composable_kernel/include/tensor_operation/threadwise_tensor_slice_transfer_v3r3.hpp
--- a/composable_kernel/include/tensor_operation/threadwise_tensor_slice_transfer_v4r1.hpp
+++ b/composable_kernel/include/tensor_operation/threadwise_tensor_slice_transfer_v4r1.hpp
--- a/composable_kernel/include/tensor_operation/threadwise_tensor_slice_transfer_v2.hpp
+++ b/composable_kernel/include/tensor_operation/threadwise_tensor_slice_transfer_v2.hpp
--- a/composable_kernel/include/tensor_operation/threadwise_tensor_slice_transfer_v6r1.hpp
+++ b/composable_kernel/include/tensor_operation/threadwise_tensor_slice_transfer_v6r1.hpp
--- a/composable_kernel/include/tensor_operation/threadwise_tensor_slice_transfer_v6r2.hpp
+++ b/composable_kernel/include/tensor_operation/threadwise_tensor_slice_transfer_v6r2.hpp
--- a/composable_kernel/include/tensor_operation/threadwise_tensor_slice_transfer_v6r3.hpp
+++ b/composable_kernel/include/tensor_operation/threadwise_tensor_slice_transfer_v6r3.hpp
--- a/composable_kernel/include/utility/amd_buffer_addressing.hpp
+++ b/composable_kernel/include/utility/amd_buffer_addressing.hpp
--- a/composable_kernel/include/utility/common_header.hpp
+++ b/composable_kernel/include/utility/common_header.hpp
@@ -35,8 +35,8 @@
 #include "dynamic_buffer.hpp"
 #include "is_known_at_compile_time.hpp"
 #include "transpose_vectors.hpp"
-
 #include "inner_product.hpp"
+#include "element_wise_operation.hpp"

 // TODO: remove this
 #if CK_USE_AMD_INLINE_ASM

--- a/composable_kernel/include/utility/config.hpp
+++ b/composable_kernel/include/utility/config.hpp
--- a/composable_kernel/include/utility/utility.hpp
+++ b/composable_kernel/include/utility/utility.hpp
--- a/device_operation/device_conv2d_fwd_xdl_c_shuffle_bias_relu_add_nhwc_kyxc_nhwk_f16_instance.cpp
+++ b/device_operation/device_conv2d_fwd_xdl_c_shuffle_bias_relu_add_nhwc_kyxc_nhwk_f16_instance.cpp
--- a/device_operation/device_conv2d_fwd_xdl_c_shuffle_bias_relu_atomic_add_nhwc_kyxc_nhwk_f16_instance.cpp
+++ b/device_operation/device_conv2d_fwd_xdl_c_shuffle_bias_relu_atomic_add_nhwc_kyxc_nhwk_f16_instance.cpp
--- a/device_operation/device_conv2d_fwd_xdl_c_shuffle_bias_relu_nhwc_kyxc_nhwk_f16_instance.cpp
+++ b/device_operation/device_conv2d_fwd_xdl_c_shuffle_bias_relu_nhwc_kyxc_nhwk_f16_instance.cpp
--- a/device_operation/device_conv2d_fwd_xdl_c_shuffle_nhwc_kyxc_nhwk_f16_instance.cpp
+++ b/device_operation/device_conv2d_fwd_xdl_c_shuffle_nhwc_kyxc_nhwk_f16_instance.cpp
--- a/device_operation/device_conv2d_fwd_xdl_nhwc_kyxc_nhwk_f16_instance.cpp
+++ b/device_operation/device_conv2d_fwd_xdl_nhwc_kyxc_nhwk_f16_instance.cpp
--- a/device_operation/device_conv2d_fwd_xdl_nhwc_kyxc_nhwk_f32_instance.cpp
+++ b/device_operation/device_conv2d_fwd_xdl_nhwc_kyxc_nhwk_f32_instance.cpp
--- a/device_operation/device_conv_xdl_instance_f16_f16_f16_nhwc_kyxc_nhwk.cpp
+++ b/device_operation/device_conv_xdl_instance_f16_f16_f16_nhwc_kyxc_nhwk.cpp
--- a/device_operation/device_conv_xdl_instance_f32_f32_f32_nhwc_kyxc_nhwk.cpp
+++ b/device_operation/device_conv_xdl_instance_f32_f32_f32_nhwc_kyxc_nhwk.cpp
--- a/device_operation/device_gemm_xdl_instance_f16_f16_f16_km_kn_mn.cpp
+++ b/device_operation/device_gemm_xdl_instance_f16_f16_f16_km_kn_mn.cpp