1.程序报错：feed的值不能是一个tensor，只能是标量、字符串、列表、数组等，所以不能用tf.reshape, 应该使用np.reshape。

    with tf.Session() as sess:
        tf.global_variables_initializer().run()
        v_x = tf.reshape(mnist.validation.images, [mnist.validation.num_examples, 28, 28, 1])
        t_x = tf.reshape(mnist.test.images, [mnist.test.num_examples, 28, 28, 1])
        validate_feed = {x: v_x, y: mnist.validation.labels}
        test_feed = {x: t_x, y: mnist.test.labels}

TypeError: The value of a feed cannot be a tf.Tensor object. 
Acceptable feed values include Python scalars, strings, lists, numpy ndarrays, or TensorHandles.

2.pool_shape的第一维是None，这是为了便于调整batch大小，但是这样的话tf.reshape无法将输出的矩阵转换为向量。可以直接使用slim.flatten()函数进行转换，不需要读取shape的大小。

# 将输出矩阵拉伸成一个向量
    pool_shape = pool2.get_shape().as_list()    # 只有tensor能用get_shape,as_list将元组转换为列表
    nodes = pool_shape[1]*pool_shape[2]*pool_shape[3]
    reshaped = tf.reshape(pool2, [pool_shape[0], nodes])

TypeError: Failed to convert object of type <class 'list'> to Tensor.

    pool_shape = pool2.get_shape().as_list()   
    nodes = pool_shape[1]*pool_shape[2]*pool_shape[3]
    reshaped = tf.contrib.slim.flatten(pool2)