[Bug c++/106309] New: ICE: error reporting routines re-entered

[Bug c++/106309] New: ICE: error reporting routines re-entered

[raffael at casagrande dot ch]

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=106309

            Bug ID: 106309
           Summary: ICE: error reporting routines re-entered
           Product: gcc
           Version: 12.1.0
            Status: UNCONFIRMED
          Severity: normal
          Priority: P3
         Component: c++
          Assignee: unassigned at gcc dot gnu.org
          Reporter: raffael at casagrande dot ch
  Target Milestone: ---

I tried for over 2 hours to get a minimal test case but it is quite involved
and for now I give up. I've created a file with the preprocessed source using
`-freport-bug` and uploaded it to my dropbox:
https://www.dropbox.com/s/oxx299jhx7yi3b2/ccFsUsUT.out?dl=0 . Unfortunately the
file is 11Kb large so I cannot attach it here in bugzilla.

The error message that GCC produces is the following:
------------------------------------------------------
‘
internal compiler error: error reporting routines re-entered.
0x1b9ca82 warning_at(unsigned int, int, char const*, ...)
        ???:0
0x95317a shorten_compare(unsigned int, tree_node**, tree_node**, tree_node**,
tree_code*)
        ???:0
0x915581 cp_build_binary_op(op_location_t const&, tree_code, tree_node*,
tree_node*, int)
        ???:0
0x75f23c build_new_op(op_location_t const&, tree_code, int, tree_node*,
tree_node*, tree_node*, tree_node*, tree_node**, int)
        ???:0
0x90c45f build_x_binary_op(op_location_t const&, tree_code, tree_node*,
tree_code, tree_node*, tree_code, tree_node*, tree_node**, int)
        ???:0
0x8b00a7 tsubst_copy_and_build(tree_node*, tree_node*, int, tree_node*, bool,
bool)
        ???:0
0x78f993 constraints_satisfied_p(tree_node*, tree_node*)
        ???:0
0x8b9f69 tsubst(tree_node*, tree_node*, int, tree_node*)
        ???:0
0x1ba82ac pp_format(pretty_printer*, text_info*)
        ???:0
0x1ba9830 pp_verbatim(pretty_printer*, char const*, ...)
        ???:0
0x1b9bfa1 diagnostic_report_diagnostic(diagnostic_context*, diagnostic_info*)
        ???:0
0x1b9cd76 error(char const*, ...)
        ???:0
0x8ade04 do_auto_deduction(tree_node*, tree_node*, tree_node*, int,
auto_deduction_context, tree_node*, int)
        ???:0
0x924d0a build_functional_cast(unsigned int, tree_node*, tree_node*, int)
        ???:0
0x8b1c54 tsubst_copy_and_build(tree_node*, tree_node*, int, tree_node*, bool,
bool)
        ???:0
0x8c0d29 instantiate_decl(tree_node*, bool, bool)
        ???:0
0x8d575b instantiate_pending_templates(int)
        ???:0
0x7db28f c_parse_final_cleanups()
        ???:0
Please submit a full bug report, with preprocessed source.
Please include the complete backtrace with any bug report.
See <https://gcc.gnu.org/bugs/> for instructions.
----------------------------------------------------------------------

As far as I can tell, the file which I uploaded to Dropbox should contain all
the information which is necessary to reproduce the bug (e.g. command line and
precompiled sources).
If you need more information to reproduce the bug, please let me know.

[Bug c++/106309] ICE: error reporting routines re-entered

[mpolacek at gcc dot gnu.org]

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=106309

Marek Polacek <mpolacek at gcc dot gnu.org> changed:

           What    |Removed                     |Added
----------------------------------------------------------------------------
             Status|UNCONFIRMED                 |NEW
   Last reconfirmed|                            |2022-07-14
     Ever confirmed|0                           |1
           Keywords|                            |ice-on-invalid-code
                 CC|                            |mpolacek at gcc dot gnu.org

--- Comment #1 from Marek Polacek <mpolacek at gcc dot gnu.org> ---
Confirmed, thanks.

I think the problem is ultimately that shorten_compare doesn't have the
'complain' parameter, so when it's called while we're outputting another
diagnostic and shorten_compare attempts to warn, we run into the crash above.

[Bug c++/106309] ICE: error reporting routines re-entered

[marxin at gcc dot gnu.org]

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=106309

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

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

--- Comment #2 from Martin Liška <marxin at gcc dot gnu.org> ---
It fails for me with the following error:

In file included from /opt/feen_rsync/lib/feen/fe/fe.h:22,
                 from
/opt/feen_rsync/lib/feen/fe/test/h1_hierarchic_fe_space_tests.cc:9:
/opt/feen_rsync/lib/feen/fe/h1_hierarchic_fe_space.h: In instantiation of
‘feen::fe::H1HierarchicFeSpace::fe_t<0>
feen::fe::H1HierarchicFeSpace::Restrict(ENTITY&&) const [with ENTITY =
feen::mesh::single_ref_el_hanging::Cell<feen::base::RefEl{feen::base::RefElType::kHex},
feen::geom::Hex8>&; fe_t<0> = feen::fe::H1HierarchicFiniteElement<int>]’:
/opt/feen_rsync/lib/feen/fe/test/h1_hierarchic_fe_space_tests.cc:103:34:  
required from here
/opt/feen_rsync/lib/feen/fe/h1_hierarchic_fe_space.h:66:12: error: class
template argument deduction failed:
/opt/feen_rsync/lib/feen/fe/h1_hierarchic_fe_space.h:66:12: error: no matching
function for call to
‘H1HierarchicFiniteElement(feen::mesh::single_ref_el_hanging::Cell<feen::base::RefEl{feen::base::RefElType::kHex},
feen::geom::Hex8>&)’
/opt/feen_rsync/lib/feen/fe/h1_hierarchic_fe_space.h:23:3: note: candidate:
‘template<class ENTITY, class ENTITY2> H1HierarchicFiniteElement(ENTITY2&&)->
feen::fe::H1HierarchicFiniteElement<ENTITY>’
/opt/feen_rsync/lib/feen/fe/h1_hierarchic_fe_space.h:23:3: note:   template
argument deduction/substitution failed:
/opt/feen_rsync/lib/feen/fe/h1_hierarchic_fe_space.h:66:12: note:   couldn’t
deduce template parameter ‘ENTITY’
/opt/feen_rsync/lib/feen/fe/h1_hierarchic_fe_space.h:19:7: note: candidate:
‘template<class ENTITY>
H1HierarchicFiniteElement(feen::fe::H1HierarchicFiniteElement<ENTITY>)->
feen::fe::H1HierarchicFiniteElement<ENTITY>’
/opt/feen_rsync/lib/feen/fe/h1_hierarchic_fe_space.h:19:7: note:   template
argument deduction/substitution failed:
/opt/feen_rsync/lib/feen/fe/h1_hierarchic_fe_space.h:66:12: note:  
‘feen::mesh::single_ref_el_hanging::Cell<feen::base::RefEl{feen::base::RefElType::kHex},
feen::geom::Hex8>’ is not derived from
‘feen::fe::H1HierarchicFiniteElement<ENTITY>’

[Bug c++/106309] ICE: error reporting routines re-entered

[asolokha at gmx dot com]

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=106309

Arseny Solokha <asolokha at gmx dot com> changed:

           What    |Removed                     |Added
----------------------------------------------------------------------------
                 CC|                            |asolokha at gmx dot com

--- Comment #3 from Arseny Solokha <asolokha at gmx dot com> ---
(In reply to Martin Liška from comment #2)
> It fails for me with the following error:

gcc ICEs as reported when invoked like this:

% g++-13.0.0 -march=cascadelake -std=c++20 -g -c ccFsUsUT.cc

[Bug c++/106309] ICE: error reporting routines re-entered

[mpolacek at gcc dot gnu.org]

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=106309

Marek Polacek <mpolacek at gcc dot gnu.org> changed:

           What    |Removed                     |Added
----------------------------------------------------------------------------
           Keywords|                            |needs-reduction
             Status|WAITING                     |NEW

--- Comment #4 from Marek Polacek <mpolacek at gcc dot gnu.org> ---
Unfortunately, after 12 hours of reducing, I'm still at 154457 lines :(.

[Bug c++/106309] ICE: error reporting routines re-entered

[mpolacek at gcc dot gnu.org]

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=106309

--- Comment #5 from Marek Polacek <mpolacek at gcc dot gnu.org> ---
Created attachment 53302
  --> https://gcc.gnu.org/bugzilla/attachment.cgi?id=53302&action=edit
x.ii.xz

Where I got so far.

[Bug c++/106309] ICE: error reporting routines re-entered

[mpolacek at gcc dot gnu.org]

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=106309

--- Comment #6 from Marek Polacek <mpolacek at gcc dot gnu.org> ---
Martin, did you have any luck reducing this further?

[Bug c++/106309] ICE: error reporting routines re-entered

[marxin at gcc dot gnu.org]

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=106309

--- Comment #7 from Martin Liška <marxin at gcc dot gnu.org> ---
No, sorry, I was not lucky.