Re: Fortran assumed-shape array problems with MPI



On May 30, 10:03 pm, "Michael Metcalf" <michaelmetc...@xxxxxxxxxxxxxx>
wrote:
"Jane Smith" <claire.s....@xxxxxxxxx> wrote in message

news:1180574928.670034.260430@xxxxxxxxxxxxxxxxxxxxxxxxxxxxxx

Do we need an interface to use assumed-shape array ? (I thought
that it is a common feature in fortran 90). What interface do we need?

Yes, an explicit one. Try putting the called procedure into a module that is
USEd by the caller, or make it an internal procedure to the caller, or put
them both in the same module (the simplest ways).

Regards,

Mike Metcalf

Thank you very much :)

.



Relevant Pages

  • Re: explicit interface of a module function
    ... >> I am trying to write a module function with optional argument. ... >> the user guide for F95, such a module function must have a explicit ... > interface, that does not mean there has to be an interface body ... This is another example of Fortran standardese natural language ...
    (comp.lang.fortran)
  • Re: Fortran 2003 questions
    ... external procedure or a dummy procedure." ... in the Fortran 95 or 2003 standard, even though all compilers I tried ... explicit interface if the procedure is an internal procedure, ...
    (comp.lang.fortran)
  • Re: Fortran assumed-shape array problems with MPI
    ... Do we need an interface to use assumed-shape array? ... that it is a common feature in fortran 90). ... USEd by the caller, or make it an internal procedure to the caller, or put ...
    (comp.lang.fortran)
  • Re: aspnet_wp.exe and COM+
    ... For releasing the interface i am not sure what you mean by, ... Such components do run "in-process", that means they run in the security context of the caller, so here the caller is allowed to call LogonUser(which I suppose you do in your Impersonate method) and the CreateEventEmailFolder Method doesn't need distributed transaction services either. ...
    (microsoft.public.dotnet.languages.csharp)
  • Re: [UPDATE PATCH] push rounding up of relative request to schedule_timeout()
    ... >>sleeps probably need to use a different interface. ... > Adding 1 is easy enough for the caller and even easier to explain in the ... > instructions (i.e. this call sleeps for X jiffies edges). ... Lots of drivers basically do something like ...
    (Linux-Kernel)