[fortran, patch] PR25094 - check for private types in public interfaces

[Jerry DeLisle]

Daniel Franke wrote:
> As is, gfortran accepts PRIVATE procedures in PUBLIC interfaces which expect 
> dummy arguments of PRIVATE type. The patch is straightforward, the testcase 
> the reporter's.
> 
> 
> :ADDPATCH fortran:
> 
> gcc/fortran:
> 2007-07-02  Daniel Franke  <franke.daniel@gmail.com>
> 
> 	PR fortran/25094
> 	* resolve.c (resolve_fl_procedure): Added check for PRIVATE types 
> 	in PUBLIC interfaces.
> 
> gcc/testsuite:
> 2007-07-02  Daniel Franke  <franke.daniel@gmail.com>
> 
> 	PR fortran/25094
> 	* gfortran.dg/interface_13.f90: New test.
> 
> 
> Regression tested on i686-pc-linux-gnu. Ok for trunk?
> 
> Regards
> 	Daniel
> 
> 
> ------------------------------------------------------------------------
> 
> Index: fortran/resolve.c
> ===================================================================
> --- fortran/resolve.c	(revision 126214)
> +++ fortran/resolve.c	(working copy)
> @@ -6591,6 +6603,8 @@
>  	&& sym->ns->parent->proc_name->attr.flavor == FL_MODULE)
>        && gfc_check_access(sym->attr.access, sym->ns->default_access))
>      {
> +      gfc_interface *iface;
> +
>        for (arg = sym->formal; arg; arg = arg->next)
>  	{
>  	  if (arg->sym
> @@ -6608,6 +6622,29 @@
>  	      return FAILURE;
>  	    }
>  	}
> +
> +      /* PUBLIC interfaces may expose PRIVATE procedures that take takes
> +	 PRIVAT to the containing module.  */

"take" repeated twice in comment above and s/PRIVAT/PRIVATE/

> +      for (iface = sym->generic; iface; iface = iface->next)
> +	{
> +	  for (arg = iface->sym->formal; arg; arg = arg->next)
> +	    {
> +	      if (arg->sym
> +		  && arg->sym->ts.type == BT_DERIVED
> +		  && !arg->sym->ts.derived->attr.use_assoc
> +		  && !gfc_check_access (arg->sym->ts.derived->attr.access,
> +					arg->sym->ts.derived->ns->default_access))
> +		{
> +		  gfc_error_now ("Procedure '%s' in PUBLIC interface '%s' at %L takes "
> +				 "dummy arguments of '%s' which is PRIVATE",
> +				 iface->sym->name, sym->name, &iface->sym->declared_at,
> +				gfc_typename(&arg->sym->ts));
> +		  /* Stop this message from recurring.  */
> +		  arg->sym->ts.derived->attr.access = ACCESS_PUBLIC;

  I presume you get multiple error messages.  Would it be better to just break 
out of the loop?

> +		  return FAILURE;
> +		}
> +	     }
> +	}
>      }
>  
---- snip -----

Regards,

Jerry