Bug 49430

I get:

test.f90:19:0: internal compiler error: tree check: expected tree that contains 'typed' structure, have '�' in fold_convert_loc, at fold-const.c:1859

==14082== Invalid read of size 1
==14082==    at 0x72F9B2: fold_convert_loc (fold-const.c:1859)
==14082==    by 0xA7D3F6: build_range_type_1 (tree.c:7170)
==14082==    by 0x5C8C69: gfc_get_character_type_len_for_eltype (trans-types.c:977)
==14082==    by 0x59EA63: gfc_conv_procedure_call (trans-expr.c:3538)
==14082==    by 0x59F271: gfc_conv_function_expr (trans-expr.c:4082)
==14082==    by 0x597D01: gfc_trans_assignment_1 (trans-expr.c:6112)

In trans-expr.c:3538:
(gdb) p *ts.u.cl
$2 = {length = 0x0, next = 0x0, length_from_typespec = 215 '\327',
      backend_decl = 0x1535340, passed_length = 0x2aaaacdd7f00,
      resolved = 6764539}

I wonder whether the
3538              type = gfc_get_character_type (ts.kind, ts.u.cl);
should be rather:
   type = gfc_get_character_type (ts.kind, ts->deferred ? NULL_TREE : ts.u.cl);

The same error occurs for the test case in PR 49112 comment 7:


module datetime_mod

    implicit none
    private

    type :: DateTime
    contains
        procedure :: getFormattedString
    end type

contains

    function getTimeString(dt, FMT) result(string)
        character(:), allocatable :: string
        class(DateTime), intent(IN) :: dt
        integer, optional, intent(IN) :: FMT

        string = dt%getFormattedString(2, FMT)
    end function 

    function getFormattedString(this, FILTER, FMT) &
        result(string)
        character(:), allocatable :: string
        class(DateTime), intent(IN) :: this
        integer, optional, intent(IN) :: FILTER, FMT

    end function

end module 


c7.f90: In function ‘gettimestring’:
c7.f90:18:0: internal compiler error: tree check: expected tree that contains ‘typed’ structure, have ‘’ in fold_convert_loc, at fold-const.c:1859

(In reply to comment #1)
> I wonder whether the
> 3538              type = gfc_get_character_type (ts.kind, ts.u.cl);
> should be rather:
>    type = gfc_get_character_type (ts.kind, ts->deferred ? NULL_TREE : ts.u.cl);

Well, with this:


Index: gcc/fortran/trans-expr.c
===================================================================
--- gcc/fortran/trans-expr.c	(revision 175256)
+++ gcc/fortran/trans-expr.c	(working copy)
@@ -3535,7 +3535,8 @@ gfc_conv_procedure_call (gfc_se * se, gfc_symbol *
       else if (ts.type == BT_CHARACTER)
 	{
 	  /* Pass the string length.  */
-	  type = gfc_get_character_type (ts.kind, ts.u.cl);
+	  type = gfc_get_character_type (ts.kind,
+					 ts.deferred ? NULL_TREE : ts.u.cl);
 	  type = build_pointer_type (type);
 
 	  /* Return an address to a char[0:len-1]* temporary for


I run into another segfault:


#0  0x0000000000575a6d in gfc_evaluate_now_loc (loc=2308, expr=0x0, pblock=0x7fffffffd560) at /home/jweil/gcc47/trunk/gcc/fortran/trans.c:126
#1  0x0000000000575b40 in gfc_evaluate_now (expr=0x0, pblock=0x7fffffffd560) at /home/jweil/gcc47/trunk/gcc/fortran/trans.c:139
#2  0x00000000005ca23b in gfc_trans_assignment_1 (expr1=0x1af3df0, expr2=0x1af3eb0, init_flag=0 '\000', dealloc=1 '\001')
    at /home/jweil/gcc47/trunk/gcc/fortran/trans-expr.c:6117

Slightly reduced/modified test case, giving a different error:


  abstract interface
    function messageProcedure()
      character(:), allocatable :: messageProcedure
    end function
  end interface

  type :: ctype
    procedure(messageProcedure), pointer, nopass :: getMessage
  end type

  type(ctype) :: this
  character :: message
  message = this%getMessage()

end


internal compiler error: vector VEC(tree,base) index domain error, in gfc_conv_procedure_call at fortran/trans-expr.c:3387

Comment #4 can be fixed with the following patch:


Index: gcc/fortran/trans-expr.c
===================================================================
--- gcc/fortran/trans-expr.c	(revision 175256)
+++ gcc/fortran/trans-expr.c	(working copy)
@@ -3381,7 +3381,7 @@ gfc_conv_procedure_call (gfc_se * se, gfc_symbol *
 	     we take the character length of the first argument for the result.
 	     For dummies, we have to look through the formal argument list for
 	     this function and use the character length found there.*/
-	  if (ts.deferred && (sym->attr.allocatable || sym->attr.pointer))
+	  if (ts.deferred)
 	    cl.backend_decl = gfc_create_var (gfc_charlen_type_node, "slen");
 	  else if (!sym->attr.dummy)
 	    cl.backend_decl = VEC_index (tree, stringargs, 0);
@@ -6112,7 +6112,7 @@ gfc_trans_assignment_1 (gfc_expr * expr1, gfc_expr
   gfc_conv_expr (&rse, expr2);
 
   /* Stabilize a string length for temporaries.  */
-  if (expr2->ts.type == BT_CHARACTER)
+  if (expr2->ts.type == BT_CHARACTER && !expr2->ts.deferred)
     string_length = gfc_evaluate_now (rse.string_length, &rse.pre);
   else
     string_length = NULL_TREE;

The combined patches of comment #3 and #5 regtest cleanly, but do not fully fix the original test case, cf. comment #3.

Author: burnus
Date: Sun May 13 10:52:32 2012
New Revision: 187436

URL: http://gcc.gnu.org/viewcvs?root=gcc&view=rev&rev=187436
Log:
2012-05-13  Alessandro Fanfarillo  <fanfarillo.gcc@gmail.com>
            Tobias Burnus  <burnus@net-b.de>

        PR fortran/52158
        PR fortran/45170
        PR fortran/49430
        * resolve.c (resolve_fl_derived0): Deferred character length 
        procedure components are supported.
        * trans-expr.c (gfc_conv_procedure_call): Handle TBP with 
        deferred-length results.
        (gfc_string_to_single_character): Add a new check to prevent
        NULL read.
        (gfc_conv_procedure_call): Remove unuseful checks on 
        symbol's attributes. Add new checks to prevent NULL read on
        string length. 

2012-05-13  Alessandro Fanfarillo  <fanfarillo.gcc@gmail.com>

        PR fortran/45170
        * gfortran.dg/deferred_type_param_3.f90: New.
        * gfortran.dg/deferred_type_proc_pointer_1.f90: New.
        * gfortran.dg/deferred_type_proc_pointer_2.f90: New.


Added:
    trunk/gcc/testsuite/gfortran.dg/deferred_type_param_3.f90
    trunk/gcc/testsuite/gfortran.dg/deferred_type_proc_pointer_1.f90
    trunk/gcc/testsuite/gfortran.dg/deferred_type_proc_pointer_2.f90
Modified:
    trunk/gcc/fortran/ChangeLog
    trunk/gcc/fortran/resolve.c
    trunk/gcc/fortran/trans-expr.c
    trunk/gcc/testsuite/ChangeLog

FIXED on the trunk (GCC 4.8).

Thanks for the bug report and sorry for taking nearly one year to get it fixed.

(gdb) call debug_rtx(insn)
(insn:TI 454 460 607 (set (reg:SI 2 r2 [orig:433 buf+2 ] [433])
        (zero_extend:SI (mem/u/c/i:QI (symbol_ref/u:SI ("*.LC0") [flags 0x2]) [0 S1 A8]))) cr_parse-reduced.ii:111 168 {*arm_zero_extendqisi2_v6}
     (nil))

oh, sorry please ignore my comment