add_subdirectory(details)
# ddim lib
proto_library(framework_proto SRCS framework.proto)

cc_library(ddim SRCS ddim.cc DEPS eigen3 boost)
cc_test(ddim_test SRCS ddim_test.cc DEPS ddim)
nv_test(dim_test SRCS dim_test.cu DEPS ddim)

if(WITH_GPU)
  nv_library(tensor SRCS tensor.cc tensor_util.cu DEPS ddim place paddle_memory device_context framework_proto)
else()
  cc_library(tensor SRCS tensor.cc tensor_util.cc DEPS ddim place paddle_memory device_context framework_proto)
endif()

cc_test(tensor_test SRCS tensor_test.cc DEPS tensor)
if(WITH_GPU)
  nv_test(tensor_util_test SRCS tensor_util_test.cc tensor_util_test.cu DEPS tensor)
else()
  cc_test(tensor_util_test SRCS tensor_util_test.cc DEPS tensor)
endif()

cc_test(eigen_test SRCS eigen_test.cc DEPS tensor)

nv_test(mixed_vector_test SRCS mixed_vector_test.cu DEPS place paddle_memory device_context init)
cc_library(lod_tensor SRCS lod_tensor.cc DEPS ddim place tensor framework_proto recordio)
cc_test(lod_tensor_test SRCS lod_tensor_test.cc DEPS lod_tensor paddle_memory)
nv_test(lod_tensor_gpu_test SRCS lod_tensor_test.cu DEPS lod_tensor init)

cc_library(reader SRCS reader.cc DEPS lod_tensor ddim)

cc_test(variable_test SRCS variable_test.cc)

cc_library(threadpool SRCS threadpool.cc DEPS enforce)
cc_test(threadpool_test SRCS threadpool_test.cc DEPS threadpool)

cc_library(scope SRCS scope.cc DEPS glog threadpool)
cc_test(scope_test SRCS scope_test.cc DEPS scope)

cc_library(data_device_transform SRCS data_device_transform.cc DEPS tensor)
nv_test(data_device_transform_test SRCS data_device_transform_test.cu
        DEPS operator op_registry init math_function)

if(WITH_GPU)
  nv_library(data_type_transform SRCS data_type_transform.cu DEPS tensor)
  nv_test(data_type_transform_test SRCS data_type_transform_test.cc data_type_transform_test.cu DEPS data_type_transform)
else()
  cc_library(data_type_transform SRCS data_type_transform.cc DEPS tensor)
  cc_test(data_type_transform_test SRCS data_type_transform_test.cc DEPS data_type_transform)
endif()

cc_library(data_layout_transform SRCS data_layout_transform.cc DEPS tensor math_function)
cc_test(data_layout_transform_test SRCS data_layout_transform_test.cc DEPS data_layout_transform)

cc_library(data_transform SRCS data_transform.cc DEPS math_function tensor
        framework_proto selected_rows data_device_transform data_type_transform data_layout_transform)

cc_library(attribute SRCS attribute.cc DEPS framework_proto boost)
cc_test(program_desc_test SRCS program_desc_test.cc DEPS proto_desc
device_context)
cc_library(op_proto_maker SRCS op_proto_maker.cc DEPS framework_proto attribute)
cc_test(op_proto_maker_test SRCS op_proto_maker_test.cc DEPS op_proto_maker)
cc_library(op_info SRCS op_info.cc DEPS attribute framework_proto)
cc_library(shape_inference SRCS shape_inference.cc DEPS ddim attribute device_context)
cc_library(operator SRCS operator.cc DEPS op_info device_context tensor scope glog
    shape_inference data_transform lod_tensor profiler)
cc_test(operator_test SRCS operator_test.cc DEPS operator op_registry init)
cc_library(proto_desc SRCS var_desc.cc op_desc.cc block_desc.cc program_desc.cc DEPS shape_inference op_info operator glog)

cc_library(op_registry SRCS op_registry.cc DEPS op_proto_maker op_info operator glog proto_desc)
nv_test(op_registry_test SRCS op_registry_test.cc DEPS op_registry)

py_proto_compile(framework_py_proto SRCS framework.proto)
# Generate an empty __init__.py to make framework_py_proto as a valid python module.
add_custom_target(framework_py_proto_init ALL COMMAND ${CMAKE_COMMAND} -E touch __init__.py)
add_dependencies(framework_py_proto framework_py_proto_init)
add_custom_command(TARGET framework_py_proto POST_BUILD
    COMMAND ${CMAKE_COMMAND} -E make_directory ${PADDLE_SOURCE_DIR}/python/paddle/fluid/proto
    COMMAND cp *.py ${PADDLE_SOURCE_DIR}/python/paddle/fluid/proto/
    COMMENT "Copy generated python proto into directory paddle/fluid/proto."
    WORKING_DIRECTORY ${CMAKE_CURRENT_BINARY_DIR})

cc_library(backward SRCS backward.cc DEPS net_op)
cc_test(backward_test SRCS backward_test.cc DEPS backward recurrent_op device_context fill_constant_op)
cc_library(lod_rank_table SRCS lod_rank_table.cc DEPS lod_tensor)

cc_library(feed_fetch_method SRCS feed_fetch_method.cc DEPS lod_tensor scope glog)

cc_library(executor SRCS executor.cc DEPS op_registry device_context scope
framework_proto backward glog lod_rank_table feed_fetch_method)

if(WITH_GPU)
  set(parallel_executor_cuda_deps nccl_all_reduce_op_handle)
else()
  set(parallel_executor_cuda_deps)
endif()
cc_library(parallel_executor SRCS parallel_executor.cc DEPS op_registry device_context scope
         backward glog lod_rank_table simple_threadpool scale_loss_grad_op_handle
        fetch_op_handle computation_op_handle ssa_graph ${parallel_executor_cuda_deps})

cc_library(prune SRCS prune.cc DEPS framework_proto)
cc_test(prune_test SRCS prune_test.cc DEPS op_info prune recurrent_op device_context)
cc_test(var_type_inference_test SRCS var_type_inference_test.cc DEPS op_registry
        proto_desc)
cc_library(selected_rows SRCS selected_rows.cc DEPS tensor)
cc_test(selected_rows_test SRCS selected_rows_test.cc DEPS selected_rows)

cc_library(init SRCS init.cc DEPS gflags device_context place stringpiece operator)
cc_test(init_test SRCS init_test.cc DEPS init)

cc_test(op_kernel_type_test SRCS op_kernel_type_test.cc DEPS place device_context framework_proto)
cc_test(cow_ptr_tests SRCS details/cow_ptr_test.cc)
      
cc_test(channel_test SRCS channel_test.cc)
cc_test(tuple_test SRCS tuple_test.cc )
cc_test(concurrency_test SRCS concurrency_test.cc DEPS go_op channel_close_op channel_create_op
        channel_send_op channel_recv_op sum_op select_op elementwise_add_op compare_op
        conditional_block_op while_op assign_op print_op executor proto_desc)