[Bug fortran/77678] ICE in fold_read_from_constant_string, at fold-const.c:13706

[tkoenig at gcc dot gnu.org]

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=77678

Thomas Koenig  changed:

   What|Removed |Added

 Status|NEW |RESOLVED
 CC||tkoenig at gcc dot gnu.org
 Resolution|--- |FIXED

--- Comment #9 from Thomas Koenig  ---
(In reply to kargl from comment #8)

> I think that this should be closed.

Yes.

[Bug fortran/77678] ICE in fold_read_from_constant_string, at fold-const.c:13706

[kargl at gcc dot gnu.org]

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=77678

--- Comment #8 from kargl at gcc dot gnu.org ---
(In reply to kargl from comment #6)
> (In reply to Richard Biener from comment #5)
> > ICE fixed.
> 
> Add -fcheck=all to your command line options.
> 
> With the ICE fixed, I think that this falls squarely in the
> WONTFIX or INVALID category.  It is the user's responsibility
> to check if the value of i is within the bounds of the string.
> 
> 
> % gfc7 -o z -O2 -fcheck=all -finit-integer=-123456 a.f90 && ./z
> At line 5 of file a.f90
> Fortran runtime error: Substring out of bounds: lower bound (-123456)
> is less than one

I think that this should be closed.

[Bug fortran/77678] ICE in fold_read_from_constant_string, at fold-const.c:13706

[marxin at gcc dot gnu.org]

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=77678

--- Comment #7 from Martin Liška  ---
Richi: Can the bug be marked as resolved?

[Bug fortran/77678] ICE in fold_read_from_constant_string, at fold-const.c:13706

[kargl at gcc dot gnu.org]

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=77678

kargl at gcc dot gnu.org changed:

   What|Removed |Added

 CC||kargl at gcc dot gnu.org

--- Comment #6 from kargl at gcc dot gnu.org ---
(In reply to Richard Biener from comment #5)
> ICE fixed.

Add -fcheck=all to your command line options.

With the ICE fixed, I think that this falls squarely in the
WONTFIX or INVALID category.  It is the user's responsibility
to check if the value of i is within the bounds of the string.


% gfc7 -o z -O2 -fcheck=all -finit-integer=-123456 a.f90 && ./z
At line 5 of file a.f90
Fortran runtime error: Substring out of bounds: lower bound (-123456)
is less than one

[Bug fortran/77678] ICE in fold_read_from_constant_string, at fold-const.c:13706

[rguenth at gcc dot gnu.org]

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=77678

Richard Biener  changed:

   What|Removed |Added

   Keywords|ice-on-invalid-code |

--- Comment #5 from Richard Biener  ---
ICE fixed.

[Bug fortran/77678] ICE in fold_read_from_constant_string, at fold-const.c:13706

[rguenth at gcc dot gnu.org]

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=77678

--- Comment #4 from Richard Biener  ---
Author: rguenth
Date: Thu Sep 22 12:15:38 2016
New Revision: 240351

URL: https://gcc.gnu.org/viewcvs?rev=240351&root=gcc&view=rev
Log:
2016-09-22  Richard Biener  

PR middle-end/77678
* expr.c (expand_expr_real_1): Guard array access against negative
offset.

Modified:
trunk/gcc/ChangeLog
trunk/gcc/expr.c

[Bug fortran/77678] ICE in fold_read_from_constant_string, at fold-const.c:13706

[rguenth at gcc dot gnu.org]

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=77678

Richard Biener  changed:

   What|Removed |Added

   Keywords||accepts-invalid

--- Comment #3 from Richard Biener  ---
I have a patch for the ICE.

Index: gcc/expr.c
===
--- gcc/expr.c  (revision 240342)
+++ gcc/expr.c  (working copy)
@@ -10274,7 +10257,8 @@ expand_expr_real_1 (tree exp, rtx target
fold_convert_loc (loc, sizetype,
  low_bound));

-   if (compare_tree_int (index1, TREE_STRING_LENGTH (init)) < 0)
+   if (tree_fits_uhwi_p (index1)
+   && compare_tree_int (index1, TREE_STRING_LENGTH (init)) <
0)
  {
tree type = TREE_TYPE (TREE_TYPE (init));
machine_mode mode = TYPE_MODE (type);

[Bug fortran/77678] ICE in fold_read_from_constant_string, at fold-const.c:13706

[marxin at gcc dot gnu.org]

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=77678

Martin Liška  changed:

   What|Removed |Added

   Keywords||ice-on-invalid-code
 Status|UNCONFIRMED |NEW
   Last reconfirmed||2016-09-22
 CC||marxin at gcc dot gnu.org
 Ever confirmed|0   |1

--- Comment #2 from Martin Liška  ---
Confirmed, all releases I have (4.5.0+) are affected.

For the second example, we ICE in expansion of:
  _1 = c[-12345678]{lb: 1 sz: 1};

where c:
  static character(kind=1) c[1:3] = "abc";

[Bug fortran/77678] ICE in fold_read_from_constant_string, at fold-const.c:13706

[gerhard.steinmetz.fort...@t-online.de]

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=77678

--- Comment #1 from Gerhard Steinmetz  
---

Slightly modified :


$ cat z2.f90
program p
   character(3) :: c = 'abc'
   character(1) :: z
   integer :: i
   ! i = 2
   z = c(i:i)
   print *, z
end


$ gfortran-7-20160918 -O2 -finit-integer=-1 z2.f90
$ gfortran-7-20160918 -O2 -finit-integer=-890 z2.f90
$ gfortran-7-20160918 -O2 -finit-integer=-12345678 z2.f90
z2.f90:6:0:

z = c(i:i)

internal compiler error: Segmentation fault
0xc21a5f crash_signal
../../gcc/toplev.c:336
0x9053b7 expand_expr_real_1(tree_node*, rtx_def*, machine_mode,
expand_modifier, rtx_def**, bool)
../../gcc/expr.c:10284
0x9077f8 expand_expr_real_1(tree_node*, rtx_def*, machine_mode,
expand_modifier, rtx_def**, bool)
../../gcc/expr.c:9725
0x912cf6 store_expr_with_bounds(tree_node*, rtx_def*, int, bool, bool,
tree_node*)
../../gcc/expr.c:5547
0x9147d7 expand_assignment(tree_node*, tree_node*, bool)
../../gcc/expr.c:5167
0x802cb6 expand_gimple_stmt_1
../../gcc/cfgexpand.c:3649
0x802cb6 expand_gimple_stmt
../../gcc/cfgexpand.c:3745
0x8050ce expand_gimple_basic_block
../../gcc/cfgexpand.c:5752
0x80b276 execute
../../gcc/cfgexpand.c:6363