JDK-8214689 (http://hg.openjdk.java.net/valhalla/valhalla/rev/b890823b8dc6) introduced unloaded ciValueKlass for cases like:

      anewarray #3 // class "QV;"

... where the V class has not yet been loaded. However, ciObjectFactory stores unloaded classes using their unadorned (?) names -- i.e., any leading "L" or "Q" are stripped. This leads to confusion when both "Q" and "L" references are made to the same unloaded class:

    static Object[] test5(int n) {
        if (n == 0) {
            return null;
        } else if (n == 1) {
            MyValue5.box[] arr = new MyValue5.box[10];
            arr[1] = MyValue5.make(22345);
            return arr;
        } else {
            MyValue5[] arr = new MyValue5[10];
            arr[1] = MyValue5.make(12345);
            return arr;
        }
    }

The above method would cause C1 to assert.

A proper fix would be to revert the JDK-8214689 changes in unloaded classes handling. Instead, we should determine the Q-ness using ciBytecodeStream::is_klass_never_null()