This is the mail archive of the fortran@gcc.gnu.org mailing list for the GNU Fortran project.


Index Nav: [Date Index] [Subject Index] [Author Index] [Thread Index]
Message Nav: [Date Prev] [Date Next] [Thread Prev] [Thread Next]
Other format: [Raw text]

Re: [Patch, Fortran, Doc] Document type-bound operators


Paul Richard Thomas wrote:
Quite simple - it lays out for me of the structures associated with
extended derived types in a digestible form!  I'll need to track down
all the helper functions that you have written to handle them.  I'll
do that with the ChangeLog, though.

I am just on the point of putting fingers to keyboard to try out the
vtable/vindex.  I have figured out a way of dealing with generic
typebound procedures, which looks OK but does wind up with unavoidable
references to an internal error function if an extended class does not
contain some of the specific interfaces of class members.  However, I
cannot see how else to do it in an efficient way.

Good luck! But I'm not sure about GENERIC type-bound procedures -- my impression was that they don't need any special handling, and by translating them to the matching specific type-bound procedure it would work just as well with polymorphism.


Isn't dynamic dispatch for specific bindings all that needs to be implemented? I thought that the current implementation of GENERIC/operators would ideally need no change at all, but only the real type-bound procedures (which is currently handled by resolve_typebound_static or the like), where the binding is not known at compile-time (i.e. not NON_OVERRIDABLE type, polymorphic entity).

I'm looking forward to your results!
Daniel

--
Done:  Arc-Bar-Cav-Ran-Rog-Sam-Tou-Val-Wiz
To go: Hea-Kni-Mon-Pri


Index Nav: [Date Index] [Subject Index] [Author Index] [Thread Index]
Message Nav: [Date Prev] [Date Next] [Thread Prev] [Thread Next]