spark - in udf of dataframe pass reference another data frame

Question

In spark can we pass one dataframe df1 to udf of df2?

I am trying following but in udf i am getting null pointer exception while accessing df1 in udf.

example -

df2.select(df2.columns.map(c => testUDF(df1, c)(col(c)).alias(c + "_test")): _*)



def testUDF(df1: DataFrame, cn: String) =
   udf((c:String) => {
    val t = df1.where(col("fn") === cn)
    //somecode
})