both modified:   tensorflow/c/eager/BUILD
        both modified:   tensorflow/compiler/jit/BUILD
        both modified:   tensorflow/compiler/tests/BUILD
        both modified:   tensorflow/compiler/xla/service/gpu/gpu_executable.cc
        both modified:   tensorflow/compiler/xla/service/gpu/ir_emitter_unnested.cc
        both modified:   tensorflow/compiler/xla/service/gpu/kernel_thunk.cc
        both modified:   tensorflow/compiler/xla/service/gpu/nvptx_compiler.cc
        both modified:   tensorflow/contrib/distributions/BUILD
        both modified:   tensorflow/contrib/linalg/BUILD
        both modified:   tensorflow/core/BUILD
        both modified:   tensorflow/core/common_runtime/gpu/gpu_bfc_allocator.cc
        both modified:   tensorflow/core/common_runtime/gpu/gpu_device.cc
        both modified:   tensorflow/core/common_runtime/gpu/gpu_device.h
        both modified:   tensorflow/core/common_runtime/gpu/gpu_id_manager.cc
        both modified:   tensorflow/core/common_runtime/gpu/gpu_id_manager.h
        both modified:   tensorflow/core/common_runtime/gpu/gpu_id_manager_test.cc
        both modified:   tensorflow/core/common_runtime/gpu/gpu_id_utils.h
        both modified:   tensorflow/core/common_runtime/gpu/gpu_util.cc
        both modified:   tensorflow/core/common_runtime/gpu/process_state.cc
        both modified:   tensorflow/core/distributed_runtime/BUILD
        both modified:   tensorflow/core/distributed_runtime/rpc/BUILD
        both modified:   tensorflow/core/kernels/batch_matmul_op_impl.h
        both modified:   tensorflow/core/kernels/batch_matmul_op_real.cc
        both modified:   tensorflow/core/kernels/conv_grad_filter_ops.cc
        both modified:   tensorflow/core/kernels/conv_grad_input_ops.cc
        both modified:   tensorflow/core/kernels/conv_ops.cc
        both modified:   tensorflow/core/kernels/crop_and_resize_op_gpu.cu.cc
        both modified:   tensorflow/core/kernels/depthwise_conv_grad_op.cc
        both modified:   tensorflow/core/kernels/depthwise_conv_op.cc
        both modified:   tensorflow/core/kernels/functional_ops.cc
        both modified:   tensorflow/core/kernels/matmul_op.cc
        both modified:   tensorflow/core/kernels/relu_op_gpu.cu.cc
        both modified:   tensorflow/core/kernels/segment_reduction_ops.h
        both modified:   tensorflow/core/platform/macros.h
        deleted by us:   tensorflow/core/util/cuda_kernel_helper.h
        both modified:   tensorflow/core/util/gpu_device_functions.h
        both modified:   tensorflow/core/util/port.cc
        both modified:   tensorflow/python/kernel_tests/linalg/BUILD
        both modified:   tensorflow/python/platform/test.py
        both modified:   tensorflow/tensorflow.bzl