[Bug fortran/108546] New: [11/12/13 Regression] ICE in expand_expr_real_1, at expr.cc:10910

[Bug fortran/108546] New: [11/12/13 Regression] ICE in expand_expr_real_1, at expr.cc:10910

[gscfq@t-online.de]

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=108546

            Bug ID: 108546
           Summary: [11/12/13 Regression] ICE in expand_expr_real_1, at
                    expr.cc:10910
           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 with r10, between 20191110 and 20191117 :


$ cat z1.f90
module m
   use iso_c_binding
contains
   subroutine s(x)
      type(c_ptr), optional :: x
      !$omp target is_device_ptr(x)
      if (c_associated(x)) stop
      !$omp end target
   end
end


$ cat z2.f90            # compiles, for reference only
module m
   use iso_c_binding
contains
   subroutine s(x)
      type(c_ptr) :: x
      !$omp target is_device_ptr(x)
      if (c_associated(x)) stop
      !$omp end target
   end
end


$ gfortran-13-20230122 -c z2.f90 -fopenmp
$ gfortran-10-20191110 -c z1.f90 -fopenmp
$
$ gfortran-13-20230122 -c z1.f90 -fopenmp
during RTL pass: expand
z1.f90:6:35:

    6 |       !$omp target is_device_ptr(x)
      |                                   ^
internal compiler error: in expand_expr_real_1, at expr.cc:10910
0xa570bf expand_expr_real_1(tree_node*, rtx_def*, machine_mode,
expand_modifier, rtx_def**, bool)
        ../../gcc/expr.cc:10904
0xa5e66e expand_expr
        ../../gcc/expr.h:310
0xa5e66e expand_operands(tree_node*, tree_node*, rtx_def*, rtx_def**,
rtx_def**, expand_modifier)
        ../../gcc/expr.cc:8582
0xa6a23a do_store_flag
        ../../gcc/expr.cc:13113
0xa6574e expand_expr_real_2(separate_ops*, rtx_def*, machine_mode,
expand_modifier)
        ../../gcc/expr.cc:10259
0x94fba2 expand_gimple_stmt_1
        ../../gcc/cfgexpand.cc:3983
0x94fba2 expand_gimple_stmt
        ../../gcc/cfgexpand.cc:4044
0x9548d7 expand_gimple_basic_block
        ../../gcc/cfgexpand.cc:6096
0x95739e execute
        ../../gcc/cfgexpand.cc:6831

[Bug fortran/108546] [11/12/13 Regression] ICE in expand_expr_real_1, at expr.cc:10910

[rguenth at gcc dot gnu.org]

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=108546

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

           What    |Removed                     |Added
----------------------------------------------------------------------------
   Target Milestone|---                         |11.4

[Bug fortran/108546] [11/12/13 Regression] ICE in expand_expr_real_1, at expr.cc:10910 since r10-4531-ga2c26c50310a3363

[marxin at gcc dot gnu.org]

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=108546

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

           What    |Removed                     |Added
----------------------------------------------------------------------------
            Summary|[11/12/13 Regression] ICE   |[11/12/13 Regression] ICE
                   |in expand_expr_real_1, at   |in expand_expr_real_1, at
                   |expr.cc:10910               |expr.cc:10910 since
                   |                            |r10-4531-ga2c26c50310a3363
     Ever confirmed|0                           |1
   Last reconfirmed|                            |2023-01-26
             Status|UNCONFIRMED                 |NEW
                 CC|                            |grosser at gcc dot gnu.org,
                   |                            |marxin at gcc dot gnu.org

--- Comment #1 from Martin Liška <marxin at gcc dot gnu.org> ---
Started with r10-4531-ga2c26c50310a3363.

[Bug middle-end/108546] [11/12/13 Regression] ICE in expand_expr_real_1, at expr.cc:10910 since r10-4531-ga2c26c50310a3363

[rguenth at gcc dot gnu.org]

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=108546

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

           What    |Removed                     |Added
----------------------------------------------------------------------------
          Component|fortran                     |middle-end
           Priority|P3                          |P2

--- Comment #2 from Richard Biener <rguenth at gcc dot gnu.org> ---
(gdb) p exp
$1 = <parm_decl 0x7ffff7ff3280 x>
(gdb) p context
$2 = <function_decl 0x7ffff6460f00 s>
(gdb) p current_function_decl
$3 = <function_decl 0x7ffff6065200 __m_MOD_s._omp_fn.0>

so we somehow forgot to remap 's'? Looks like a middle-end issue to me.

[Bug middle-end/108546] [11/12/13 Regression] ICE in expand_expr_real_1, at expr.cc:10910 since r10-4531-ga2c26c50310a3363

[cvs-commit at gcc dot gnu.org]

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=108546

--- Comment #3 from CVS Commits <cvs-commit at gcc dot gnu.org> ---
The master branch has been updated by Tobias Burnus <burnus@gcc.gnu.org>:

https://gcc.gnu.org/g:96ff97ff6574666a5509ae9fa596e7f2b6ad4f88

commit r13-6390-g96ff97ff6574666a5509ae9fa596e7f2b6ad4f88
Author: Tobias Burnus <tobias@codesourcery.com>
Date:   Wed Mar 1 13:53:09 2023 +0100

    OpenMP/Fortran: Fix handling of optional is_device_ptr + bind(C) [PR108546]

    For is_device_ptr, optional checks should only be done before calling
    libgomp, afterwards they are NULL either because of absent or, by
    chance, because it is unallocated or unassociated (for
pointers/allocatables).

    Additionally, it fixes an issue with explicit mapping for 'type(c_ptr)'.

            PR middle-end/108546

    gcc/fortran/ChangeLog:

            * trans-openmp.cc (gfc_trans_omp_clauses): Fix mapping of
            type(C_ptr) variables.

    gcc/ChangeLog:

            * omp-low.cc (lower_omp_target): Remove optional handling
            on the receiver side, i.e. inside target (data), for
            use_device_ptr.

    libgomp/ChangeLog:

            * testsuite/libgomp.fortran/is_device_ptr-3.f90: New test.
            * testsuite/libgomp.fortran/use_device_ptr-optional-4.f90: New
test.

[Bug middle-end/108546] [11/12/13 Regression] ICE in expand_expr_real_1, at expr.cc:10910 since r10-4531-ga2c26c50310a3363

[cvs-commit at gcc dot gnu.org]

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=108546

--- Comment #4 from CVS Commits <cvs-commit at gcc dot gnu.org> ---
The releases/gcc-12 branch has been updated by Tobias Burnus
<burnus@gcc.gnu.org>:

https://gcc.gnu.org/g:fa1e458d05a94d064e8daef88c2be300317d7e8b

commit r12-9235-gfa1e458d05a94d064e8daef88c2be300317d7e8b
Author: Tobias Burnus <tobias@codesourcery.com>
Date:   Wed Mar 1 13:53:09 2023 +0100

    OpenMP/Fortran: Fix handling of optional is_device_ptr + bind(C) [PR108546]

    For is_device_ptr, optional checks should only be done before calling
    libgomp, afterwards they are NULL either because of absent or, by
    chance, because it is unallocated or unassociated (for
pointers/allocatables).

    Additionally, it fixes an issue with explicit mapping for 'type(c_ptr)'.

            PR middle-end/108546

    gcc/fortran/ChangeLog:

            * trans-openmp.cc (gfc_trans_omp_clauses): Fix mapping of
            type(C_ptr) variables.

    gcc/ChangeLog:

            * omp-low.cc (lower_omp_target): Remove optional handling
            on the receiver side, i.e. inside target (data), for
            use_device_ptr.

    libgomp/ChangeLog:

            * testsuite/libgomp.fortran/is_device_ptr-3.f90: New test.
            * testsuite/libgomp.fortran/use_device_ptr-optional-4.f90: New
test.

    (cherry picked from commit 96ff97ff6574666a5509ae9fa596e7f2b6ad4f88)

[Bug middle-end/108546] [11/12/13 Regression] ICE in expand_expr_real_1, at expr.cc:10910 since r10-4531-ga2c26c50310a3363

[cvs-commit at gcc dot gnu.org]

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=108546

--- Comment #5 from CVS Commits <cvs-commit at gcc dot gnu.org> ---
The releases/gcc-11 branch has been updated by Tobias Burnus
<burnus@gcc.gnu.org>:

https://gcc.gnu.org/g:e7d995cf9bff2ca4a54b94c4a38455ebd1c2883e

commit r11-10565-ge7d995cf9bff2ca4a54b94c4a38455ebd1c2883e
Author: Tobias Burnus <tobias@codesourcery.com>
Date:   Wed Mar 1 13:53:09 2023 +0100

    OpenMP/Fortran: Fix handling of optional is_device_ptr + bind(C) [PR108546]

    For is_device_ptr, optional checks should only be done before calling
    libgomp, afterwards they are NULL either because of absent or, by
    chance, because it is unallocated or unassociated (for
pointers/allocatables).

    Additionally, it fixes an issue with explicit mapping for 'type(c_ptr)'.

            PR middle-end/108546

    gcc/fortran/ChangeLog:

            * trans-openmp.c (gfc_trans_omp_clauses): Fix mapping of
            type(C_ptr) variables.

    gcc/ChangeLog:

            * omp-low.c (lower_omp_target): Remove optional handling
            on the receiver side, i.e. inside target (data), for
            use_device_ptr.

    libgomp/ChangeLog:

            * testsuite/libgomp.fortran/is_device_ptr-3.f90: New test.
            * testsuite/libgomp.fortran/use_device_ptr-optional-4.f90: New
test.

    (cherry picked from commit 96ff97ff6574666a5509ae9fa596e7f2b6ad4f88)

[Bug middle-end/108546] [11/12/13 Regression] ICE in expand_expr_real_1, at expr.cc:10910 since r10-4531-ga2c26c50310a3363

[burnus at gcc dot gnu.org]

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=108546

Tobias Burnus <burnus at gcc dot gnu.org> changed:

           What    |Removed                     |Added
----------------------------------------------------------------------------
             Status|NEW                         |RESOLVED
         Resolution|---                         |FIXED
                 CC|                            |burnus at gcc dot gnu.org

--- Comment #6 from Tobias Burnus <burnus at gcc dot gnu.org> ---
FIXED on GCC 11/12/mainline(= GCC 13).

Thanks for this bugreport (and may others).