Merge branch 'develop' of https://github.com/ROCmSoftwarePlatform/composable_kernel into wmma_op

library/src/tensor_operation_instance/gpu/gemm_add_fastgelu/CMakeLists.txt

+add_instance_library(device_gemm_add_fastgelu_instance
+   device_gemm_add_fastgelu_xdl_c_shuffle_f16_f16_f16_f16_km_kn_mn_mn_instance.cpp
+   device_gemm_add_fastgelu_xdl_c_shuffle_f16_f16_f16_f16_km_nk_mn_mn_instance.cpp
+   device_gemm_add_fastgelu_xdl_c_shuffle_f16_f16_f16_f16_mk_kn_mn_mn_instance.cpp
+   device_gemm_add_fastgelu_xdl_c_shuffle_f16_f16_f16_f16_mk_nk_mn_mn_instance.cpp
+)

library/src/tensor_operation_instance/gpu/gemm_fastgelu/CMakeLists.txt

+add_instance_library(device_gemm_fastgelu_instance
+   device_gemm_fastgelu_xdl_c_shuffle_f16_f16_f16_km_kn_mn_instance.cpp
+   device_gemm_fastgelu_xdl_c_shuffle_f16_f16_f16_km_nk_mn_instance.cpp
+   device_gemm_fastgelu_xdl_c_shuffle_f16_f16_f16_mk_kn_mn_instance.cpp
+   device_gemm_fastgelu_xdl_c_shuffle_f16_f16_f16_mk_nk_mn_instance.cpp
+)

profiler/CMakeLists.txt

@@ -26,6 +26,7 @@ set(PROFILER_SOURCE
     src/profile_groupnorm.cpp
     src/profile_layernorm.cpp
     src/profile_softmax.cpp
+    src/profile_batchnorm_fwd.cpp
 )
 
 add_executable(ckProfiler ${PROFILER_SOURCE})
@@ -57,5 +58,6 @@ target_link_libraries(ckProfiler PRIVATE device_conv2d_fwd_bias_relu_add_instanc
 target_link_libraries(ckProfiler PRIVATE device_normalization_instance)
 target_link_libraries(ckProfiler PRIVATE device_softmax_instance)
 target_link_libraries(ckProfiler PRIVATE device_reduce_instance)
+target_link_libraries(ckProfiler PRIVATE device_batchnorm_instance)
 
 rocm_install(TARGETS ckProfiler COMPONENT profiler)

test/CMakeLists.txt

@@ -53,6 +53,7 @@ add_subdirectory(softmax)
 add_subdirectory(normalization)
 add_subdirectory(data_type)
 add_subdirectory(elementwise_normalization)
+add_subdirectory(batchnorm_fwd)
 if(GPU_TARGETS MATCHES "gfx1100")
     add_subdirectory(wmma_op)
-endif()
\ No newline at end of file
+endif()