Method run_op

src/gpu/cl/kernels/ClDepthConcatenateKernel.cpp:126–149

}

void ClDepthConcatenateKernel::run_op(ITensorPack &tensors, const Window &window, ::cl::CommandQueue &queue)
{
    ARM_COMPUTE_ERROR_ON_UNCONFIGURED_KERNEL(this);
    ARM_COMPUTE_ERROR_ON_INVALID_SUBWINDOW(ICLKernel::window(), window);

    const auto src =
        utils::cast::polymorphic_downcast<const ICLTensor *>(tensors.get_const_tensor(TensorType::ACL_SRC));
    auto dst = utils::cast::polymorphic_downcast<ICLTensor *>(tensors.get_tensor(TensorType::ACL_DST));

    Window slice = window.first_slice_window_3D();

    const int offset_to_first_elements_in_bytes = _depth_offset * dst->info()->strides_in_bytes()[2];

    unsigned int idx = 2 * num_arguments_per_3D_tensor(); // Skip the src and dst parameters
    _kernel.setArg<cl_int>(idx, offset_to_first_elements_in_bytes);

    do
    {
        unsigned int idx = 0;
        add_3D_tensor_argument(idx, src, slice);
        add_3D_tensor_argument(idx, dst, slice);
        enqueue(queue, *this, slice, lws_hint());
    } while (window.slide_window_slice_3D(slice));
}
} // namespace kernels
} // namespace opencl
} // namespace arm_compute