Re: [PATCH] Fix for PR fortran/21375

[Tobias Schlüter <tobias dot schlueter at physik dot uni-muenchen dot de>]

Steve Kargl wrote:
> OK for mainline?  Ok for 4.0.1 with Mark's approval?

This is ok with one minor nit:

> *** 1899,1905 ****
>   
>     gfor_fndecl_deallocate =
>       gfc_build_library_function_decl (get_identifier (PREFIX("deallocate")),
> ! 				     void_type_node, 1, ppvoid_type_node);
>   
>     gfor_fndecl_stop_numeric =
>       gfc_build_library_function_decl (get_identifier (PREFIX("stop_numeric")),
> --- 1899,1906 ----
>   
>     gfor_fndecl_deallocate =
>       gfc_build_library_function_decl (get_identifier (PREFIX("deallocate")),
> ! 				     void_type_node, 2, ppvoid_type_node,
> ! 				     gfc_int4_type_node);
                                     ^^^^^^^^^^^^^^^^^^
> +   /* Set up the optional STAT= */
> +   if (code->expr)
> +     {
> +       /* Variable used with the library call.  */
> +       tree gfc_int4_type_node = gfc_get_int_type (4);
> +       stat = gfc_create_var (gfc_int4_type_node, "stat");
> +       pstat = gfc_build_addr_expr (NULL, stat);
> + 
> +       /* Running total of possible deallocation failures.  */
> +       astat = gfc_create_var (gfc_int4_type_node, "astat");
> +       apstat = gfc_build_addr_expr (NULL, astat);
> + 
> +       /* Initialize astat to 0.  */
> +       gfc_add_modify_expr (&block, astat, build_int_cst (TREE_TYPE (astat), 0));
> +     }
> +   else
> +     {
> +       pstat = apstat = integer_zero_node;
                           ^^^^^^^^^^^^^^^^^

Those should be pointers to int4.  See the top of
gfc_build_io_library_fndecls() for how to create that type.  We should
probably add a global variable gfc_pint4_type_node for this along the other
type nodes, and move that code out of gfc_build_io_library_fndecls, but I'm
fine if you put it in a local variable.

- Tobi