[Bug fortran/104826] New: [11/12 Regression] ICE in gimple_range_global, at value-query.cc:424

[Bug fortran/104826] New: [11/12 Regression] ICE in gimple_range_global, at value-query.cc:424

[gscfq@t-online.de]

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=104826

            Bug ID: 104826
           Summary: [11/12 Regression] ICE in gimple_range_global, at
                    value-query.cc:424
           Product: gcc
           Version: 12.0
            Status: UNCONFIRMED
          Severity: normal
          Priority: P3
         Component: fortran
          Assignee: unassigned at gcc dot gnu.org
          Reporter: gscfq@t-online.de
  Target Milestone: ---

Changed between 20210606 and 20210613 :


$ cat z1.f90
program p
   character(:), allocatable :: x
   save
   !$omp target
   x = 'abc'
   !$omp end target
end


$ cat z2.f90   # no save
program p
   character(:), allocatable :: x
   !$omp target
   x = 'abc'
   !$omp end target
end


$ gfortran-12-20210606 -c z1.f90 -fopenmp
$ gfortran-12-20220213 -c z2.f90 -fopenmp
$
$ gfortran-12-20220213 -c z1.f90 -fopenmp
during GIMPLE pass: walloca
z1.f90:5:12:

    5 |    x = 'abc'
      |            ^
internal compiler error: Segmentation fault
0xccad7f crash_signal
        ../../gcc/toplev.cc:322
0xf5c9ec gimple_range_global(tree_node*)
        ../../gcc/value-query.cc:424
0x1864d80 ranger_cache::get_global_range(irange&, tree_node*) const
        ../../gcc/gimple-range-cache.cc:923
0x1860f8e gimple_ranger::export_global_ranges()
        ../../gcc/gimple-range.cc:474
0x189c0e1 pass_walloca::execute(function*)
        ../../gcc/gimple-ssa-warn-alloca.cc:381

[Bug fortran/104826] [11/12 Regression] ICE in gimple_range_global, at value-query.cc:424

[rguenth at gcc dot gnu.org]

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=104826

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

           What    |Removed                     |Added
----------------------------------------------------------------------------
   Last reconfirmed|                            |2022-03-08
     Ever confirmed|0                           |1
             Status|UNCONFIRMED                 |NEW
   Target Milestone|---                         |11.3
           Priority|P3                          |P2
           Keywords|                            |ice-on-valid-code
                 CC|                            |jakub at gcc dot gnu.org

--- Comment #1 from Richard Biener <rguenth at gcc dot gnu.org> ---
I think we have a duplicate of this (but it was fixed already?!)

(gdb) p name
$1 = <ssa_name 0x7ffff651f990 2>
(gdb) p debug_tree (name)
 <ssa_name 0x7ffff651f990
    type <integer_type 0x7ffff652f738 integer(kind=8) public DI
        size <integer_cst 0x7ffff6517c00 constant 64>
        unit-size <integer_cst 0x7ffff6517c18 constant 8>
        align:64 warn_if_not_align:0 symtab:0 alias-set -1 canonical-type
0x7ffff652f738 precision:64 min <integer_cst 0x7ffff6517e88
-9223372036854775808> max <integer_cst 0x7ffff6517ea0 9223372036854775807>
        pointer_to_this <pointer_type 0x7ffff6564738>>

    def_stmt 
    version:2>
$2 = void
(gdb) bt
#0  0x0000000000dd69e4 in is_a_helper<gphi*>::test<gimple> (gs=<gimple 0x0>)
    at /home/rguenther/src/gcc3/gcc/gimple.h:1243
#1  0x0000000000dd8f50 in is_a<gphi*, gimple> (p=<gimple 0x0>)
    at /home/rguenther/src/gcc3/gcc/is-a.h:232
#2  0x0000000001967e67 in gimple_range_global (
    name=<ssa_name 0x7ffff651f990 2>)
    at /home/rguenther/src/gcc3/gcc/value-query.cc:424
#3  0x00000000029eadda in ranger_cache::get_global_range (this=0x413c508, 
    r=..., name=<ssa_name 0x7ffff651f990 2>)
    at /home/rguenther/src/gcc3/gcc/gimple-range-cache.cc:923
#4  0x00000000029e778f in gimple_ranger::export_global_ranges (this=0x413c4e0)
    at /home/rguenther/src/gcc3/gcc/gimple-range.cc:474
#5  0x0000000002a40694 in pass_walloca::execute (this=0x411c290, 
    fun=0x7ffff670c000)
    at /home/rguenther/src/gcc3/gcc/gimple-ssa-warn-alloca.cc:381
#6  0x00000000013b4749 in execute_one_pass (
    pass=<opt_pass* 0x411c290 "walloca"(23)>)
    at /home/rguenther/src/gcc3/gcc/passes.cc:2637

[Bug fortran/104826] [11/12 Regression] ICE in gimple_range_global, at value-query.cc:424

[jakub at gcc dot gnu.org]

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=104826

--- Comment #2 from Jakub Jelinek <jakub at gcc dot gnu.org> ---
Started with my r12-1319-gd4d38135b3137f1d3148138340e67bdcd7ea8216
Very similar to PR102314 which isn't fixed yet.

[Bug fortran/104826] [11/12 Regression] ICE in gimple_range_global, at value-query.cc:424

[rguenth at gcc dot gnu.org]

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=104826

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

           What    |Removed                     |Added
----------------------------------------------------------------------------
   Target Milestone|11.3                        |11.4

--- Comment #3 from Richard Biener <rguenth at gcc dot gnu.org> ---
GCC 11.3 is being released, retargeting bugs to GCC 11.4.

[Bug fortran/104826] [11/12/13 Regression] ICE in gimple_range_global, at value-query.cc:424 – with deferred-length character variable

[burnus at gcc dot gnu.org]

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=104826

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

           What    |Removed                     |Added
----------------------------------------------------------------------------
                 CC|                            |burnus at gcc dot gnu.org
            Summary|[11/12/13 Regression] ICE   |[11/12/13 Regression] ICE
                   |in gimple_range_global, at  |in gimple_range_global, at
                   |value-query.cc:424          |value-query.cc:424 – with
                   |                            |deferred-length character
                   |                            |variable

--- Comment #4 from Tobias Burnus <burnus at gcc dot gnu.org> ---
The code as written is invalid as 'x' is originally unallocated on entry to the
target region but then gets allocated on assignment in the target region.

However, adding before '!$omp target' the line
   allocate(character(len=3) :: x)
does not help.

[Bug fortran/104826] [11/12/13/14 Regression] ICE in gimple_range_global, at value-query.cc:424 – with deferred-length character variable

[jakub at gcc dot gnu.org]

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=104826

Jakub Jelinek <jakub at gcc dot gnu.org> changed:

           What    |Removed                     |Added
----------------------------------------------------------------------------
   Target Milestone|11.4                        |11.5

--- Comment #5 from Jakub Jelinek <jakub at gcc dot gnu.org> ---
GCC 11.4 is being released, retargeting bugs to GCC 11.5.