Probably the indexes on c.udfkeyl or/and u.udfkeyl cant be used any more due to the application of the trim function.
1. Update the underlying values with trim
2. create a functional index on trim (udfkeyl)
By applying a function to your join conditions you will be preventing the database from being able to use indexes, so it's very likely the query will take longer, and if there's a lot of data, that could be a seriously big impact.
Are you sure your joining keys actually need trimming? That would seem like corrupt data to me to have even generated keys with erroneous additional characters in them.
If it's absolutely necessary, then you'd likely need to create function based indexes on those two.