[Bug fortran/36153] ICE on size with kind parameter

[burnus at gcc dot gnu dot org]

------- Comment #1 from burnus at gcc dot gnu dot org  2008-05-06 14:43 -------
Confirmed. Note: Using -std=f95 the kind= is properly rejected. 

==16567== Invalid read of size 4
==16567==    at 0x4605FD: gfc_resolve_expr (resolve.c:2310)
==16567==    by 0x465625: resolve_code (resolve.c:6195)
==16567==    by 0x468D1B: gfc_resolve_blocks (resolve.c:5988)
==16567==    by 0x46560C: resolve_code (resolve.c:6187)
==16567==    by 0x466F40: resolve_codes (resolve.c:9121)

I think the following comparison does not make sense for KIND= (I have only
glanced at it):

      for (arg = expr->value.function.actual; arg; arg = arg->next)
        {
          if (inquiry && arg->next != NULL && arg->next->expr)
            {
              if (arg->next->expr->expr_type != EXPR_CONSTANT)
                break;

              if ((int)mpz_get_si (arg->next->expr->value.integer)
                        < arg->expr->rank)
                break;

I think this function still assumes that one uses only DIM= and not KIND=. The
failing line is the mpz_get_si, although that should work? Allowed options are:

   SIZE (ARRAY [, DIM, KIND])


-- 

burnus at gcc dot gnu dot org changed:

           What    |Removed                     |Added
----------------------------------------------------------------------------
             Status|UNCONFIRMED                 |NEW
     Ever Confirmed|0                           |1
           Keywords|                            |ice-on-valid-code
      Known to fail|                            |4.3.1 4.4.0
   Last reconfirmed|0000-00-00 00:00:00         |2008-05-06 14:43:06
               date|                            |


http://gcc.gnu.org/bugzilla/show_bug.cgi?id=36153