[Patch Modified] re: Enable Hollerith constant and character arrayin format tag (take 3).

[Steven Bosscher]

On Friday 01 July 2005 12:13, Feng Wang wrote:
> Handle allocatable array assiged Hollerith constant and more test.
>
> Please remove the diff on trans-io.c in the patch:
> http://gcc.gnu.org/ml/fortran/2005-06/msg00475.html
> And use this diff attached.

Wrt. http://gcc.gnu.org/ml/gcc-patches/2005-06/msg02163.html I only
have one remark:

> !       /* Numeric can be converted to any other numeric. And Hollerith can be
> !        converted to any other type.  */
> !       if ((gfc_numeric_ts (&lvalue->ts) && gfc_numeric_ts (&rvalue->ts))
> !       || rvalue->ts.type == BT_HOLLERITH)
>         return SUCCESS;

Please use proper indentation.  The rest looks reasonable to me. Could
someone double-check the io.c changes please?


For http://gcc.gnu.org/ml/gcc-patches/2005-07/msg00020.html there are
a few more things that look just wrong:

> +       gcc_assert (GFC_DESCRIPTOR_TYPE_P (type));
> +       size = gfc_conv_array_stride (array, rank);
> +       tmp = fold_build2 (MINUS_EXPR, gfc_array_index_type,
> +               gfc_conv_array_ubound (array, rank),
> +               gfc_conv_array_lbound (array, rank));
> +       tmp = fold_build2 (PLUS_EXPR, gfc_array_index_type, tmp,
> +               gfc_index_one_node);
> +       size = fold_build2 (MULT_EXPR, gfc_array_index_type, tmp, size);   

This can't be right.  gfc_conv_array_stride returns a COMPONENT_REF,
 so the last line here tries to multiply gfc_array_index_type and a
COMPONENT_REF.  Very Wrong thing to do.
  

> *************** set_string (stmtblock_t * block, stmtblo
> *** 400,406 ****
>       }
>     else
>       {
> !       gfc_conv_expr (&se, e);
>         gfc_conv_string_parameter (&se);
>         gfc_add_modify_expr (&se.pre, io, fold_convert (TREE_TYPE (io), se.expr));
>         gfc_add_modify_expr (&se.pre, len, se.string_length);
> --- 462,474 ----
>       }
>     else
>       {
> !       /* General character.  */
> !       if (e->ts.type == BT_CHARACTER && e->rank == 0)
> !       gfc_conv_expr (&se, e);
> !       /* Array assigned Hollerith constant or character array.  */
> !       else if (e->symtree && (e->symtree->n.sym->as->rank > 0))
> !       gfc_convert_array_to_string (&se, e);

You should add an extra "else gcc_unreachable ()" clause.

> ! 
>         gfc_conv_string_parameter (&se);
>         gfc_add_modify_expr (&se.pre, io, fold_convert (TREE_TYPE (io), se.expr));
>         gfc_add_modify_expr (&se.pre, len, se.string_length);

Thanks for working on this.

Gr.
Steven