Is that normal to have the same declaration of type being considered as to different type by the error :
"The type of the actual argument differs from the type of the dummy argument".
I attached a testfile.
Unless the specific rules in Article 188.8.131.52, "Determination of derived types", of the Fortran 2003 standard are satisfied, repeating a type definition word for word may still cause the compiler to regard the two types so declared as distinct.
When I saw this "feature" the first time, I was surprised, too. However, it does make it easier for the compiler to keep track of user-defined types.
The IMPORT ... statement is provided to alleviate the problem that you ran into.
Exactly what i was looking for.