[RFC] PR fortran/20839

[Steve Kargl]

On Sun, Nov 13, 2005 at 04:30:39PM +0100, Steven Bosscher wrote:
>      case ST_IMPLIED_ENDDO:
> +      /* If the do-stmt of this DO construct has a do-construct-name,
> +	 the corresponding end-do must be an end-do-stmt (with a matching
> +	 name, but in that case we must have seen ST_ENDDO first).
> +         We only complain about this in pedantic mode.  */
> +      if (pedantic && gfc_current_block () != NULL)
> +	gfc_error_now
> +	  ("Named DO construct at %L has no matching named ENDDO statement",

"Named DO construct at %L missing a named ENDDO statement",

or

"Named END DO statement not found for named DO construct at %L"

> +	   &gfc_current_block()->declared_at);
>        break;
>  
>      default:

-- 
Steve