If you are looking to apply this to the selection criteria, you may be out of luck.  Your best bet is to do the filtering in your data source, where a couple subqueries or self-joins should do the trick.

If you are willing to bring down all the data, and simply suppress the unwanted data, there is a trick you can use.  It's inefficient, but it works.

Create a formula.  The formula should look something like:

DistinctCount({Code},{Code},"='E1'") +
DistinctCount({Code},{Code},"='C1'") +
DistinctCount({Code},{Code},"='K1'")

The DistinctCount, in this case, should return 0 if there are no records with that code, and 1 if there is.  (This is because it is counting the number of distinct codes that are equal to a given value, which will always be either 0 or 1.)

Then, simply group the report by customer, and suppress any group in which the above formula does not equal 3.  (That will be because at least one of the codes was not present in that group.)