Hi Tobias,

you are right in that I was barking up the wrong tree.
I was focussed too much on the testcase in the PR.

> I think that one is wrong. While CLASS_DATA (e) accesses e->ts.u.derived->components,
> which always works, your code assumes that there is only 'c' and not 'x%c' where
> 'c' is of type BT_CLASS and 'x' is of type BT_DERIVED.
>
> I wonder whether it works if you simply remove 'return true;'
> as gfc_add_class_array_ref sets 'e->rank = CLASS(e)->rank (and
> adds an AR_FULL ref, if needed). In the nonerror case, the
> 'return true' is obtained via:
>     if (e->rank != 0 && e->ts.type != BT_PROCEDURE)
>       return true;
> And, otherwise, it falls through to the error.
>
> OK if that works

Well, I tried and this does not work.

However, an additional plain check on e->rank != 0 also in the
CLASS cases fixes the original issue as well as your example:

> type t
>    class(*), allocatable :: c(:)
> end type t
> type(t) :: x
> x%c = [1,2,3,4]
> print *, size(x%c)
> print *, size(x%c(1)) ! { dg-error ... }
> end

And regtests ok. :-)

See attached updated patch.

Anything else I am missing?

Thanks for the constructive review!

Harald