我有两个表要使用r markdown连接到sql块中。列太多，我不想在select语句中指定所有列。在 Impala /Hive或Spark壳里我可以 select * ，但在sql块中 select * 抛出错误。下面是一个最小的工作示例：

```{sql connection=sc, eval=TRUE}
with tmp1 as (
   select 'mark' as name, 1 as mark
   union
   select 'tim' as name, 2 as mark
),
tmp2 as (
   select 'mark' as name, 'a' as class
   union
   select 'tim' as name, 'b' as class
)
select
  *
from
  tmp1
inner join
  tmp2
  on tmp1.name = tmp2.name
    ```

我得到了错误

Error: org.apache.spark.sql.AnalysisException: Reference 'name' is ambiguous, could be: a.name, b.name.;

为什么这样不行？