[Bug fortran/55692] ICE on incorrect use of ASSOCIATED function

["janus at gcc dot gnu.org" <gcc-bugzilla at gcc dot gnu dot org>]

http://gcc.gnu.org/bugzilla/show_bug.cgi?id=55692

janus at gcc dot gnu.org changed:

           What    |Removed                     |Added
----------------------------------------------------------------------------
                 CC|                            |janus at gcc dot gnu.org

--- Comment #2 from janus at gcc dot gnu.org 2012-12-14 18:57:09 UTC ---
(In reply to comment #1)
> Confirmed on 4.4.6, 4.5.3, 4.6.3, 4.7.2, and trunk.

I see it also with 4.3.4.


Here is a preliminary patch which fixes it on trunk:


Index: gcc/fortran/check.c
===================================================================
--- gcc/fortran/check.c    (revision 194387)
+++ gcc/fortran/check.c    (working copy)
@@ -895,13 +895,11 @@ gfc_check_associated (gfc_expr *pointer, gfc_expr

   where = &pointer->where;

-  if (pointer->expr_type == EXPR_VARIABLE || pointer->expr_type ==
EXPR_FUNCTION)
-    attr1 = gfc_expr_attr (pointer);
-  else if (pointer->expr_type == EXPR_NULL)
+  if (pointer->expr_type == EXPR_NULL)
     goto null_arg;
-  else
-    gcc_assert (0); /* Pointer must be a variable or a function.  */

+  attr1 = gfc_expr_attr (pointer);
+
   if (!attr1.pointer && !attr1.proc_pointer)
     {
       gfc_error ("'%s' argument of '%s' intrinsic at %L must be a POINTER",



With this I get:

L = ASSOCIATED([P1,P2]) ! Error: intrinsic function ASSOCIATED is not ELEMENTAL
               1
Error: 'pointer' argument of 'associated' intrinsic at (1) must be a POINTER


... which, I think, is the proper way to reject the test case (and certainly
more polite than an ICE ;)