[Bug fortran/106209] New: ICE in add_init_expr_to_sym, at fortran/decl.cc:2132

[Bug fortran/106209] New: ICE in add_init_expr_to_sym, at fortran/decl.cc:2132

[gscfq@t-online.de]

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=106209

            Bug ID: 106209
           Summary: ICE in add_init_expr_to_sym, at fortran/decl.cc:2132
           Product: gcc
           Version: 13.0
            Status: UNCONFIRMED
          Severity: normal
          Priority: P3
         Component: fortran
          Assignee: unassigned at gcc dot gnu.org
          Reporter: gscfq@t-online.de
  Target Milestone: ---

Affects versions down to at least r5 :


$ cat z1.f90
program p
   integer, parameter :: a(:) = 0
   integer, parameter :: b(*) = a
end


$ cat z2.f90
program p
   integer, parameter :: a(:) = 0
   integer, parameter :: b(*) = [a]
end


$ gfortran-13-20220703 -c z1.f90
f951: internal compiler error: in add_init_expr_to_sym, at fortran/decl.cc:2132
0x6e225e add_init_expr_to_sym
        ../../gcc/fortran/decl.cc:2132
0x6ebe06 variable_decl
        ../../gcc/fortran/decl.cc:3120
0x6ebe06 gfc_match_data_decl()
        ../../gcc/fortran/decl.cc:6319
0x757723 match_word
        ../../gcc/fortran/parse.cc:67
0x757723 decode_statement
        ../../gcc/fortran/parse.cc:378
0x75916a next_free
        ../../gcc/fortran/parse.cc:1397
0x75916a next_statement
        ../../gcc/fortran/parse.cc:1629
0x75a6fb parse_spec
        ../../gcc/fortran/parse.cc:4168
0x75d89c parse_progunit
        ../../gcc/fortran/parse.cc:6210
0x75ef61 gfc_parse_file()
        ../../gcc/fortran/parse.cc:6755
0x7acb3f gfc_be_parse_file
        ../../gcc/fortran/f95-lang.cc:229

[Bug fortran/106209] ICE in add_init_expr_to_sym, at fortran/decl.cc:2132

[kargl at gcc dot gnu.org]

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=106209

kargl at gcc dot gnu.org changed:

           What    |Removed                     |Added
----------------------------------------------------------------------------
   Last reconfirmed|                            |2022-07-05
             Status|UNCONFIRMED                 |NEW
     Ever confirmed|0                           |1
                 CC|                            |kargl at gcc dot gnu.org
           Priority|P3                          |P4

--- Comment #1 from kargl at gcc dot gnu.org ---
Instead of an assert(), simply return false to give gfortran a chance to emit
an error.

% gfcx -c a.f90
a.f90:3:4:

    3 |    integer, parameter :: b(*) = a
      |    1
Error: Unclassifiable statement at (1)
a.f90:2:29:

    2 |    integer, parameter :: a(:) = 0
      |                             1
Error: Parameter array 'a' at (1) cannot be automatic or of deferred shape

diff --git a/gcc/fortran/decl.cc b/gcc/fortran/decl.cc
index bd586e75008..cec178fc80b 100644
--- a/gcc/fortran/decl.cc
+++ b/gcc/fortran/decl.cc
@@ -2129,7 +2129,9 @@ add_init_expr_to_sym (const char *name, gfc_expr **initp,
locus *var_locus)
          /* The shape may be NULL for EXPR_ARRAY, set it.  */
          if (init->shape == NULL)
            {
-             gcc_assert (init->expr_type == EXPR_ARRAY);
+             if (init->expr_type != EXPR_ARRAY)
+               return false;
+
              init->shape = gfc_get_shape (1);
              if (!gfc_array_size (init, &init->shape[0]))
                  gfc_internal_error ("gfc_array_size failed");

[Bug fortran/106209] ICE in add_init_expr_to_sym, at fortran/decl.cc:2132

[anlauf at gcc dot gnu.org]

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=106209

anlauf at gcc dot gnu.org changed:

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

--- Comment #2 from anlauf at gcc dot gnu.org ---
(In reply to kargl from comment #1)
> Instead of an assert(), simply return false to give gfortran a chance to
> emit an error.

Steve, your patch does fix z1, but not z2, which hits the

  gfc_internal_error ("gfc_array_size failed");

a few lines later...

The following adjusted patch fixes the latter, too, regtests cleanly,
and gives nicer error messages :)

diff --git a/gcc/fortran/decl.cc b/gcc/fortran/decl.cc
index 339f8b15035..b6400514731 100644
--- a/gcc/fortran/decl.cc
+++ b/gcc/fortran/decl.cc
@@ -2129,10 +2129,21 @@ add_init_expr_to_sym (const char *name, gfc_expr
**initp, locus *var_
locus)
          /* The shape may be NULL for EXPR_ARRAY, set it.  */
          if (init->shape == NULL)
            {
-             gcc_assert (init->expr_type == EXPR_ARRAY);
+             if (init->expr_type != EXPR_ARRAY)
+               {
+                 gfc_error ("Bad shape of initializer at %L", &init->where);
+                 return false;
+               }
+
              init->shape = gfc_get_shape (1);
              if (!gfc_array_size (init, &init->shape[0]))
-                 gfc_internal_error ("gfc_array_size failed");
+               {
+                 gfc_error ("Cannot determine shape of initializer at %L",
+                            &init->where);
+                 free (init->shape);
+                 init->shape = NULL;
+                 return false;
+               }
            }

          for (dim = 0; dim < sym->as->rank; ++dim)

[Bug fortran/106209] ICE in add_init_expr_to_sym, at fortran/decl.cc:2132

[sgk at troutmask dot apl.washington.edu]

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=106209

--- Comment #3 from Steve Kargl <sgk at troutmask dot apl.washington.edu> ---
On Thu, Jul 14, 2022 at 07:35:21PM +0000, anlauf at gcc dot gnu.org wrote:
> 
> --- Comment #2 from anlauf at gcc dot gnu.org ---
> (In reply to kargl from comment #1)
> > Instead of an assert(), simply return false to give gfortran a chance to
> > emit an error.
> 
> Steve, your patch does fix z1, but not z2, which hits the
> 
>   gfc_internal_error ("gfc_array_size failed");

Sorry about that.  Your patch looks good to me.

[Bug fortran/106209] ICE in add_init_expr_to_sym, at fortran/decl.cc:2132

[cvs-commit at gcc dot gnu.org]

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=106209

--- Comment #4 from CVS Commits <cvs-commit at gcc dot gnu.org> ---
The master branch has been updated by Harald Anlauf <anlauf@gcc.gnu.org>:

https://gcc.gnu.org/g:748f8a8b145dde59c7b63aa68b5a59515b7efc49

commit r13-1698-g748f8a8b145dde59c7b63aa68b5a59515b7efc49
Author: Harald Anlauf <anlauf@gmx.de>
Date:   Thu Jul 14 22:24:55 2022 +0200

    Fortran: error recovery for bad initializers of implied-shape arrays
[PR106209]

    gcc/fortran/ChangeLog:

            PR fortran/106209
            * decl.cc (add_init_expr_to_sym): Handle bad initializers for
            implied-shape arrays.

    gcc/testsuite/ChangeLog:

            PR fortran/106209
            * gfortran.dg/pr106209.f90: New test.

    Co-authored-by: Steven G. Kargl <kargl@gcc.gnu.org>

[Bug fortran/106209] ICE in add_init_expr_to_sym, at fortran/decl.cc:2132

[anlauf at gcc dot gnu.org]

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=106209

anlauf at gcc dot gnu.org changed:

           What    |Removed                     |Added
----------------------------------------------------------------------------
             Status|NEW                         |RESOLVED
         Resolution|---                         |FIXED

--- Comment #5 from anlauf at gcc dot gnu.org ---
Fixed for gcc-13.

Thanks for the report!

[Bug fortran/106209] ICE in add_init_expr_to_sym, at fortran/decl.cc:2132

[cvs-commit at gcc dot gnu.org]

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=106209

--- Comment #6 from CVS Commits <cvs-commit at gcc dot gnu.org> ---
The releases/gcc-12 branch has been updated by Harald Anlauf
<anlauf@gcc.gnu.org>:

https://gcc.gnu.org/g:6dc4f984dd14ffad34e0540f2030f553579adefd

commit r12-9079-g6dc4f984dd14ffad34e0540f2030f553579adefd
Author: Harald Anlauf <anlauf@gmx.de>
Date:   Thu Jul 14 22:24:55 2022 +0200

    Fortran: error recovery for bad initializers of implied-shape arrays
[PR106209]

    gcc/fortran/ChangeLog:

            PR fortran/106209
            * decl.cc (add_init_expr_to_sym): Handle bad initializers for
            implied-shape arrays.

    gcc/testsuite/ChangeLog:

            PR fortran/106209
            * gfortran.dg/pr106209.f90: New test.

    Co-authored-by: Steven G. Kargl <kargl@gcc.gnu.org>
    (cherry picked from commit 748f8a8b145dde59c7b63aa68b5a59515b7efc49)

[Bug fortran/106209] ICE in add_init_expr_to_sym, at fortran/decl.cc:2132

[cvs-commit at gcc dot gnu.org]

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=106209

--- Comment #7 from CVS Commits <cvs-commit at gcc dot gnu.org> ---
The releases/gcc-11 branch has been updated by Harald Anlauf
<anlauf@gcc.gnu.org>:

https://gcc.gnu.org/g:8e58d94ac56127ebca3a893284455032a707d948

commit r11-10505-g8e58d94ac56127ebca3a893284455032a707d948
Author: Harald Anlauf <anlauf@gmx.de>
Date:   Thu Jul 14 22:24:55 2022 +0200

    Fortran: error recovery for bad initializers of implied-shape arrays
[PR106209]

    gcc/fortran/ChangeLog:

            PR fortran/106209
            * decl.c (add_init_expr_to_sym): Handle bad initializers for
            implied-shape arrays.

    gcc/testsuite/ChangeLog:

            PR fortran/106209
            * gfortran.dg/pr106209.f90: New test.

    Co-authored-by: Steven G. Kargl <kargl@gcc.gnu.org>
    (cherry picked from commit 748f8a8b145dde59c7b63aa68b5a59515b7efc49)

[Bug fortran/106209] ICE in add_init_expr_to_sym, at fortran/decl.cc:2132

[cvs-commit at gcc dot gnu.org]

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=106209

--- Comment #8 from CVS Commits <cvs-commit at gcc dot gnu.org> ---
The releases/gcc-10 branch has been updated by Harald Anlauf
<anlauf@gcc.gnu.org>:

https://gcc.gnu.org/g:b523b58690c84b04cc9695d2d652611beb6f28ca

commit r10-11197-gb523b58690c84b04cc9695d2d652611beb6f28ca
Author: Harald Anlauf <anlauf@gmx.de>
Date:   Thu Jul 14 22:24:55 2022 +0200

    Fortran: error recovery for bad initializers of implied-shape arrays
[PR106209]

    gcc/fortran/ChangeLog:

            PR fortran/106209
            * decl.c (add_init_expr_to_sym): Handle bad initializers for
            implied-shape arrays.

    gcc/testsuite/ChangeLog:

            PR fortran/106209
            * gfortran.dg/pr106209.f90: New test.

    Co-authored-by: Steven G. Kargl <kargl@gcc.gnu.org>
    (cherry picked from commit 748f8a8b145dde59c7b63aa68b5a59515b7efc49)