[Bug fortran/108527] New: [13 Regression] ICE in compare_bound_int(): Bad expression

[Bug fortran/108527] New: [13 Regression] ICE in compare_bound_int(): Bad expression

[gscfq@t-online.de]

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=108527

            Bug ID: 108527
           Summary: [13 Regression] ICE in compare_bound_int(): Bad
                    expression
           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: ---

Started between 20221218 and 20230108,
was fixed, but testcases from pr103505 give an ICE again :


$ cat z1.f90
program p
   integer, parameter :: a((2.)) = [4, 8]
   integer(a(1:1)) :: b
end


$ cat z2.f90
program p
   integer, parameter :: a((2.)) = [4, 8]
   print *, [integer(a(1:1)) :: 2]
end


$ cat z3.f90
program p
   integer, parameter :: a((2.)) = [4, 8]
   integer :: b(a(1:1))
end


$ gfortran-13-20230122 -c z1.f90
f951: internal compiler error: compare_bound_int(): Bad expression
0x7c17e9 gfc_report_diagnostic
        ../../gcc/fortran/error.cc:883
0x7c3387 gfc_internal_error(char const*, ...)
        ../../gcc/fortran/error.cc:1503
0x834531 compare_bound_mpz_t
        ../../gcc/fortran/resolve.cc:4605
0x834a7c check_dimension
        ../../gcc/fortran/resolve.cc:4788
0x847ddd compare_spec_to_ref
        ../../gcc/fortran/resolve.cc:4853
0x847ddd resolve_array_ref
        ../../gcc/fortran/resolve.cc:5151
0x847ddd gfc_resolve_ref(gfc_expr*)
        ../../gcc/fortran/resolve.cc:5366
0x837bc7 resolve_variable
        ../../gcc/fortran/resolve.cc:5871
0x837bc7 gfc_resolve_expr(gfc_expr*)
        ../../gcc/fortran/resolve.cc:7197
0x7c77a4 gfc_reduce_init_expr(gfc_expr*)
        ../../gcc/fortran/expr.cc:3168
0x7ca700 gfc_match_init_expr(gfc_expr**)
        ../../gcc/fortran/expr.cc:3216
0x7aa94a gfc_match_kind_spec(gfc_typespec*, bool)
        ../../gcc/fortran/decl.cc:3273
0x7b12e9 gfc_match_decl_type_spec(gfc_typespec*, int)
        ../../gcc/fortran/decl.cc:4706
0x7b2cac gfc_match_data_decl()
        ../../gcc/fortran/decl.cc:6264
0x8215d3 match_word
        ../../gcc/fortran/parse.cc:67
0x8215d3 decode_statement
        ../../gcc/fortran/parse.cc:378
0x82301a next_free
        ../../gcc/fortran/parse.cc:1403
0x82301a next_statement
        ../../gcc/fortran/parse.cc:1635
0x824a5b parse_spec
        ../../gcc/fortran/parse.cc:4191
0x827d5c parse_progunit
        ../../gcc/fortran/parse.cc:6262

[Bug fortran/108527] [13 Regression] ICE in compare_bound_int(): Bad expression

[anlauf at gcc dot gnu.org]

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=108527

anlauf at gcc dot gnu.org changed:

           What    |Removed                     |Added
----------------------------------------------------------------------------
   Last reconfirmed|                            |2023-01-24
             Status|UNCONFIRMED                 |NEW
                 CC|                            |anlauf at gcc dot gnu.org
     Ever confirmed|0                           |1

--- Comment #1 from anlauf at gcc dot gnu.org ---
Oh well.

This fixes at least the first testcase:

diff --git a/gcc/fortran/resolve.cc b/gcc/fortran/resolve.cc
index 94213cd3cd4..65e689c6f10 100644
--- a/gcc/fortran/resolve.cc
+++ b/gcc/fortran/resolve.cc
@@ -4732,6 +4732,10 @@ check_dimension (int i, gfc_array_ref *ar,
gfc_array_spec *as)
 #define AR_START (ar->start[i] ? ar->start[i] : as->lower[i])
 #define AR_END (ar->end[i] ? ar->end[i] : as->upper[i])

+       /* Something has already gone wrong.  Allow sane recovery.  */
+       if (ar->stride[i] == NULL)
+         return false;
+
        compare_result comp_start_end = compare_bound (AR_START, AR_END);

        /* Check for zero stride, which is not allowed.  */

[Bug fortran/108527] [13 Regression] ICE in compare_bound_int(): Bad expression

[kargl at gcc dot gnu.org]

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=108527

kargl at gcc dot gnu.org changed:

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

--- Comment #2 from kargl at gcc dot gnu.org ---
This fixes all testcases.  These two helper functions can indicate
a problem occurs, so instead of asserting on (a->ts.type != BT_INTEGER)
return the CMP_UNKNOWN condition.  This does give a few run-on 
errors.


diff --git a/gcc/fortran/resolve.cc b/gcc/fortran/resolve.cc
index 94213cd3cd4..c5e1b87f8fe 100644
--- a/gcc/fortran/resolve.cc
+++ b/gcc/fortran/resolve.cc
@@ -4575,12 +4575,11 @@ compare_bound_int (gfc_expr *a, int b)
 {
   int i;

-  if (a == NULL || a->expr_type != EXPR_CONSTANT)
+  if (a == NULL
+      || a->expr_type != EXPR_CONSTANT
+      || a->ts.type != BT_INTEGER)
     return CMP_UNKNOWN;

-  if (a->ts.type != BT_INTEGER)
-    gfc_internal_error ("compare_bound_int(): Bad expression");
-
   i = mpz_cmp_si (a->value.integer, b);

   if (i < 0)
@@ -4598,12 +4597,11 @@ compare_bound_mpz_t (gfc_expr *a, mpz_t b)
 {
   int i;

-  if (a == NULL || a->expr_type != EXPR_CONSTANT)
+  if (a == NULL
+      || a->expr_type != EXPR_CONSTANT
+      || a->ts.type != BT_INTEGER)
     return CMP_UNKNOWN;

-  if (a->ts.type != BT_INTEGER)
-    gfc_internal_error ("compare_bound_int(): Bad expression");
-
   i = mpz_cmp (a->value.integer, b);

   if (i < 0)

[Bug fortran/108527] [13 Regression] ICE in compare_bound_int(): Bad expression

[anlauf at gcc dot gnu.org]

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=108527

--- Comment #3 from anlauf at gcc dot gnu.org ---
(In reply to kargl from comment #2)
> This fixes all testcases.  These two helper functions can indicate
> a problem occurs, so instead of asserting on (a->ts.type != BT_INTEGER)
> return the CMP_UNKNOWN condition.  This does give a few run-on 
> errors.

Indeed, this is better and more general.

[Bug fortran/108527] [13 Regression] ICE in compare_bound_int(): Bad expression

[anlauf at gcc dot gnu.org]

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=108527

--- Comment #4 from anlauf at gcc dot gnu.org ---
The patch in comment#2 regtests cleanly.

I can package it and submit if you agree.

[Bug fortran/108527] [13 Regression] ICE in compare_bound_int(): Bad expression

[sgk at troutmask dot apl.washington.edu]

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=108527

--- Comment #5 from Steve Kargl <sgk at troutmask dot apl.washington.edu> ---
On Tue, Jan 24, 2023 at 08:51:03PM +0000, anlauf at gcc dot gnu.org wrote:
> --- Comment #4 from anlauf at gcc dot gnu.org ---
> The patch in comment#2 regtests cleanly.
> 
> I can package it and submit if you agree.
> 

Fine with me.

[Bug fortran/108527] [13 Regression] ICE in compare_bound_int(): Bad expression

[anlauf at gcc dot gnu.org]

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=108527

anlauf at gcc dot gnu.org changed:

           What    |Removed                     |Added
----------------------------------------------------------------------------
   Target Milestone|---                         |13.0
           Assignee|unassigned at gcc dot gnu.org      |anlauf at gcc dot gnu.org
             Status|NEW                         |ASSIGNED

--- Comment #6 from anlauf at gcc dot gnu.org ---
Submitted: https://gcc.gnu.org/pipermail/fortran/2023-January/058826.html

[Bug fortran/108527] [13 Regression] ICE in compare_bound_int(): Bad expression

[rguenth at gcc dot gnu.org]

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=108527

Richard Biener <rguenth at gcc dot gnu.org> changed:

           What    |Removed                     |Added
----------------------------------------------------------------------------
           Priority|P3                          |P4

[Bug fortran/108527] [13 Regression] ICE in compare_bound_int(): Bad expression

[marxin at gcc dot gnu.org]

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=108527

Martin Liška <marxin at gcc dot gnu.org> changed:

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

--- Comment #7 from Martin Liška <marxin at gcc dot gnu.org> ---
Btw. started with r8-7594-g078c5aff5ed83e9c.

[Bug fortran/108527] [13 Regression] ICE in compare_bound_int(): Bad expression

[anlauf at gcc dot gnu.org]

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=108527

anlauf at gcc dot gnu.org changed:

           What    |Removed                     |Added
----------------------------------------------------------------------------
      Known to work|                            |7.5.0
      Known to fail|                            |10.4.1, 11.3.1, 12.2.1,
                   |                            |8.5.0, 9.5.0

--- Comment #8 from anlauf at gcc dot gnu.org ---
(In reply to Martin Liška from comment #7)
> Btw. started with r8-7594-g078c5aff5ed83e9c.

Right.  Adding known to work and known to fail versions.

[Bug fortran/108527] [13 Regression] ICE in compare_bound_int(): Bad expression

[cvs-commit at gcc dot gnu.org]

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=108527

--- Comment #9 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:22afa4947584c701633a79fd8750c9ceeaa96711

commit r13-5479-g22afa4947584c701633a79fd8750c9ceeaa96711
Author: Harald Anlauf <anlauf@gmx.de>
Date:   Tue Jan 24 22:36:25 2023 +0100

    Fortran: fix ICE in compare_bound_int [PR108527]

    gcc/fortran/ChangeLog:

            PR fortran/108527
            * resolve.cc (compare_bound_int): Expression to compare must be of
            type INTEGER.
            (compare_bound_mpz_t): Likewise.
            (check_dimension): Fix comment on checks applied to array section
            and clean up associated logic.

    gcc/testsuite/ChangeLog:

            PR fortran/108527
            * gfortran.dg/pr108527.f90: New test.

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

[Bug fortran/108527] [13 Regression] ICE in compare_bound_int(): Bad expression

[cvs-commit at gcc dot gnu.org]

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=108527

--- Comment #10 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:511928102dc9fa3f9c377e01f9bfb605b4995a61

commit r12-9106-g511928102dc9fa3f9c377e01f9bfb605b4995a61
Author: Harald Anlauf <anlauf@gmx.de>
Date:   Tue Jan 24 22:36:25 2023 +0100

    Fortran: fix ICE in compare_bound_int [PR108527]

    gcc/fortran/ChangeLog:

            PR fortran/108527
            * resolve.cc (compare_bound_int): Expression to compare must be of
            type INTEGER.
            (compare_bound_mpz_t): Likewise.
            (check_dimension): Fix comment on checks applied to array section
            and clean up associated logic.

    gcc/testsuite/ChangeLog:

            PR fortran/108527
            * gfortran.dg/pr108527.f90: New test.

    Co-authored-by: Steven G. Kargl <kargl@gcc.gnu.org>
    (cherry picked from commit 22afa4947584c701633a79fd8750c9ceeaa96711)

[Bug fortran/108527] [13 Regression] ICE in compare_bound_int(): Bad expression

[cvs-commit at gcc dot gnu.org]

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=108527

--- Comment #11 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:c01a495d96f7e9e07b16f62c3476c5242d66aeeb

commit r11-10514-gc01a495d96f7e9e07b16f62c3476c5242d66aeeb
Author: Harald Anlauf <anlauf@gmx.de>
Date:   Tue Jan 24 22:36:25 2023 +0100

    Fortran: fix ICE in compare_bound_int [PR108527]

    gcc/fortran/ChangeLog:

            PR fortran/108527
            * resolve.c (compare_bound_int): Expression to compare must be of
            type INTEGER.
            (compare_bound_mpz_t): Likewise.
            (check_dimension): Fix comment on checks applied to array section
            and clean up associated logic.

    gcc/testsuite/ChangeLog:

            PR fortran/108527
            * gfortran.dg/pr108527.f90: New test.

    Co-authored-by: Steven G. Kargl <kargl@gcc.gnu.org>
    (cherry picked from commit 22afa4947584c701633a79fd8750c9ceeaa96711)

[Bug fortran/108527] [13 Regression] ICE in compare_bound_int(): Bad expression

[cvs-commit at gcc dot gnu.org]

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=108527

--- Comment #12 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:2965a4f925461d7814972845fe480e03856fe3fe

commit r10-11206-g2965a4f925461d7814972845fe480e03856fe3fe
Author: Harald Anlauf <anlauf@gmx.de>
Date:   Tue Jan 24 22:36:25 2023 +0100

    Fortran: fix ICE in compare_bound_int [PR108527]

    gcc/fortran/ChangeLog:

            PR fortran/108527
            * resolve.c (compare_bound_int): Expression to compare must be of
            type INTEGER.
            (compare_bound_mpz_t): Likewise.
            (check_dimension): Fix comment on checks applied to array section
            and clean up associated logic.

    gcc/testsuite/ChangeLog:

            PR fortran/108527
            * gfortran.dg/pr108527.f90: New test.

    Co-authored-by: Steven G. Kargl <kargl@gcc.gnu.org>
    (cherry picked from commit 22afa4947584c701633a79fd8750c9ceeaa96711)

[Bug fortran/108527] [13 Regression] ICE in compare_bound_int(): Bad expression

[anlauf at gcc dot gnu.org]

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=108527

anlauf at gcc dot gnu.org changed:

           What    |Removed                     |Added
----------------------------------------------------------------------------
   Target Milestone|13.0                        |10.5
         Resolution|---                         |FIXED
             Status|ASSIGNED                    |RESOLVED

--- Comment #13 from anlauf at gcc dot gnu.org ---
Fixed on all open branches.  Closing.

Thanks for the report!