This is the mail archive of the gcc-patches@gcc.gnu.org mailing list for the GCC 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] PR25099 - elemental subroutine argument conformance checking


On Sat, Apr 22, 2006 at 07:49:26AM +0200, Paul Thomas wrote:
> 
> >
> >Is the check on a->expr->rank actually needed here?  
> >Doesn't the preceding if statement guarantee a->expr->rank
> >is validate?
> >
> Seeing that it's you, does this do it for you?

Yes. This is fine.  The orginal patch appeared to be 
doing unnecessary checks of a->expr->rank.  To me,
the revised patch is easier to understand.

>    {
>      if (a->expr == NULL || a->expr->rank == 0)
>        continue;
> 
>     /* The last upper bound of an assumed size array argument must
>        be present.  */
>      if (resolve_assumed_size_actual (a->expr))
>        return FAILURE;
> 
>      /* Array actual arguments must conform.  */
>      if (e != NULL)
>        {
>          if (gfc_check_conformance ("elemental subroutine", a->expr, e)
>            == FAILURE)
>        return FAILURE;
>        }
>      else
>        e = a->expr;
>    }
> 
> PS If either this or the dependency checking tests are to be committed 
> to trunk before mid-May, somebody will have to review them today.  
> Alternatively, I could just commit tomorrow morning and see what falls 
> out of the tree over the next few weeks, before committing to 4.1.

You can commit this patch.  I won't have time to check the other
patch until tomorrow.  If Thomas or Tobi or ... can't get you a 
review, I'll review and possible commit for you in your absences.

-- 
Steve


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