[Bug other/99445] New: [11 Regression] ICE in

[Bug other/99445] New: [11 Regression] ICE in

[doko at debian dot org]

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=99445

            Bug ID: 99445
           Summary: [11 Regression] ICE in
           Product: gcc
           Version: 11.0
            Status: UNCONFIRMED
          Severity: normal
          Priority: P3
         Component: other
          Assignee: unassigned at gcc dot gnu.org
          Reporter: doko at debian dot org
  Target Milestone: ---

Created attachment 50322
  --> https://gcc.gnu.org/bugzilla/attachment.cgi?id=50322&action=edit
preprocessed source, unreduced

seen with trunk 20210227, building the broker package.

$cat foo.ii
template <typename> class decay;
template <class> struct implicit_conversions;
template <class T>
using implicit_conversions_t = typename implicit_conversions<T>::type;
template <class> struct response_type;
template <class Ts, class... Xs>
using response_type_t = typename response_type<Ts, Xs...>::type;
template <class Handle, class... Ts>
    response_type<typename Handle::signatures,
                  implicit_conversions_t<typename decay<Ts>::type>...>
        response_type_t < typename Handle::signatures,
    typename implicit_conversions<typename decay<Ts>::type>::type...

$ g++ -c  -fchecking=2 --param=hash-table-verification-limit=1000 foo.ii
hash table checking failed: equal operator returns true for a pair of values
with a different hash value
foo.ii: In substitution of ‘template<class Ts, class ... Xs> using
response_type_t = typename response_type<Ts, Xs ...>::type [with Ts = typename
Handle::signatures; Xs = {typename implicit_conversions<typename
decay<Ts>::type>::type ...}]’:
foo.ii:12:66:   required from here
foo.ii:7:7: internal compiler error: in hashtab_chk_error, at hash-table.c:137
    7 | using response_type_t = typename response_type<Ts, Xs...>::type;
      |       ^~~~~~~~~~~~~~~
0x120e1f7 hashtab_chk_error()
        ../../src/gcc/hash-table.c:137
0x7863d0 hash_table<spec_hasher, false, xcallocator>::verify(spec_entry*
const&, unsigned int)
        ../../src/gcc/hash-table.h:1033
0x7863d0 hash_table<spec_hasher, false,
xcallocator>::find_with_hash(spec_entry* const&, unsigned int)
        ../../src/gcc/hash-table.h:918
0x14cdeff lookup_template_class_1
        ../../src/gcc/cp/pt.c:9871
0x14fe6fe lookup_template_class(tree_node*, tree_node*, tree_node*, tree_node*,
int, int)
        ../../src/gcc/cp/pt.c:10243
0x14fe6fe tsubst_aggr_type
        ../../src/gcc/cp/pt.c:13576
0x14e2f18 tsubst(tree_node*, tree_node*, int, tree_node*)
        ../../src/gcc/cp/pt.c:16037
0x15b98f7 tsubst_decl
        ../../src/gcc/cp/pt.c:14802
0x160ea6d instantiate_template_1
        ../../src/gcc/cp/pt.c:21118
0x160e884 instantiate_template(tree_node*, tree_node*, int)
        ../../src/gcc/cp/pt.c:21177
0x14e3a38 instantiate_alias_template
        ../../src/gcc/cp/pt.c:21215
0x14e3a38 tsubst(tree_node*, tree_node*, int, tree_node*)
        ../../src/gcc/cp/pt.c:15444
0x14cf10d lookup_template_class_1
        ../../src/gcc/cp/pt.c:9944
0x14cc186 lookup_template_class(tree_node*, tree_node*, tree_node*, tree_node*,
int, int)
        ../../src/gcc/cp/pt.c:10243
0x14cc186 finish_template_type(tree_node*, tree_node*, int)
        ../../src/gcc/cp/semantics.c:3498
0x13e7bd3 cp_parser_template_id
        ../../src/gcc/cp/parser.c:17437
0x13ef262 cp_parser_class_name
        ../../src/gcc/cp/parser.c:24664
0x13e240f cp_parser_qualifying_entity
        ../../src/gcc/cp/parser.c:6994
0x13e240f cp_parser_nested_name_specifier_opt
        ../../src/gcc/cp/parser.c:6676
0x13ded32 cp_parser_nested_name_specifier
        ../../src/gcc/cp/parser.c:6920
Please submit a full bug report,
with preprocessed source if appropriate.

[Bug c++/99445] [11 Regression] ICE in hashtab_chk_error, at hash-table.c:137

[marxin at gcc dot gnu.org]

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=99445

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

           What    |Removed                     |Added
----------------------------------------------------------------------------
             Status|UNCONFIRMED                 |WAITING
     Ever confirmed|0                           |1
   Last reconfirmed|                            |2021-03-08

--- Comment #1 from Martin Liška <marxin at gcc dot gnu.org> ---
Can you please reduce it to a valid test-case?
I see a compilation error for the reduced test-case:

$ g++ pr99445.C -c
pr99445.C:11:3: error: parse error in template argument list
   11 |   response_type_t < typename Handle::signatures,
      |   ^~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
   12 |   typename implicit_conversions<typename decay<Ts>::type>::type...
      |   ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
pr99445.C:12:67: error: invalid declarator at end of input
   12 |   typename implicit_conversions<typename decay<Ts>::type>::type...
      |                                                                   ^

[Bug c++/99445] [11 Regression] ICE in hashtab_chk_error, at hash-table.c:137 since r11-7011-g6e0a231a4aa2407b

[marxin at gcc dot gnu.org]

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=99445

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

           What    |Removed                     |Added
----------------------------------------------------------------------------
            Summary|[11 Regression] ICE in      |[11 Regression] ICE in
                   |hashtab_chk_error, at       |hashtab_chk_error, at
                   |hash-table.c:137            |hash-table.c:137 since
                   |                            |r11-7011-g6e0a231a4aa2407b

--- Comment #2 from Martin Liška <marxin at gcc dot gnu.org> ---
Likely started with r11-7011-g6e0a231a4aa2407b.

[Bug c++/99445] [11 Regression] ICE in hashtab_chk_error, at hash-table.c:137 since r11-7011-g6e0a231a4aa2407b

[rguenth at gcc dot gnu.org]

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=99445

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

           What    |Removed                     |Added
----------------------------------------------------------------------------
           Priority|P3                          |P1
   Target Milestone|---                         |11.0

[Bug c++/99445] [11 Regression] ICE in hashtab_chk_error, at hash-table.c:137 since r11-7011-g6e0a231a4aa2407b

[doko at debian dot org]

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=99445

--- Comment #3 from Matthias Klose <doko at debian dot org> ---
> Can you please reduce it to a valid test-case?

the run took 40 hours. I'll see when I can repeat it.

[Bug c++/99445] [11 Regression] ICE in hashtab_chk_error, at hash-table.c:137 since r11-7011-g6e0a231a4aa2407b

[marxin at gcc dot gnu.org]

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=99445

--- Comment #4 from Martin Liška <marxin at gcc dot gnu.org> ---
(In reply to Matthias Klose from comment #3)
> > Can you please reduce it to a valid test-case?
> 
> the run took 40 hours. I'll see when I can repeat it.

I can imagine. Or you can try to somehow "fix" the reduced test-case so that
it's valid again :)

[Bug c++/99445] [11 Regression] ICE in hashtab_chk_error, at hash-table.c:137 since r11-7011-g6e0a231a4aa2407b

[ppalka at gcc dot gnu.org]

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=99445

--- Comment #5 from Patrick Palka <ppalka at gcc dot gnu.org> ---
Reduced valid testcase:

template <class> struct implicit_conversions;
template <class T>
using implicit_conversions_t = typename implicit_conversions<T>::type;
template <class...> struct response_type;

template <class Handle, class... Ts>
using type1 = response_type<implicit_conversions_t<Ts>...>;

template <class Handle, class... Ts>
using type2 = response_type<typename implicit_conversions<Ts>::type...>;

[Bug c++/99445] [11 Regression] ICE in hashtab_chk_error, at hash-table.c:137 since r11-7011-g6e0a231a4aa2407b

[marxin at gcc dot gnu.org]

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=99445

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

           What    |Removed                     |Added
----------------------------------------------------------------------------
             Status|WAITING                     |NEW

[Bug c++/99445] [11 Regression] ICE in hashtab_chk_error, at hash-table.c:137 since r11-7011-g6e0a231a4aa2407b

[jakub at gcc dot gnu.org]

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=99445

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

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

--- Comment #6 from Jakub Jelinek <jakub at gcc dot gnu.org> ---
(gdb) p debug_tree ($15->args)
 <tree_vec 0x7fffea19ec80 length:1
    elt:0 <type_argument_pack 0x7fffea1b4e70
        type <tree_vec 0x7fffea19eca0 length:1 elt:0 <type_pack_expansion
0x7fffea1b4dc8>>
        VOID
        align:8 warn_if_not_align:0 symtab:0 alias-set -1 canonical-type
0x7fffea1b4e70>>
$23 = void
(gdb) p debug_tree ($16->args)
 <tree_vec 0x7fffea19eb80 length:1
    elt:0 <type_argument_pack 0x7fffea1b4930
        type <tree_vec 0x7fffea19eba0 length:1 elt:0 <type_pack_expansion
0x7fffea1b4888>>
        VOID
        align:8 warn_if_not_align:0 symtab:0 alias-set -1 canonical-type
0x7fffea1b4930>>
$24 = void
(gdb) p iterative_hash_template_arg ($15->args, 0)
$25 = 1060540270
(gdb) p iterative_hash_template_arg ($16->args, 0)
$26 = 2053511766
(gdb) p comp_template_args ($15->args, $16->args, 0, 0, false)
$27 = 1
(gdb) p debug_tree (0x7fffea1b4dc8)
 <type_pack_expansion 0x7fffea1b4dc8
    type <typename_type 0x7fffea1b4738 type VOID
        align:8 warn_if_not_align:0 symtab:0 alias-set -1 structural-equality
context <record_type 0x7fffea1b4690 implicit_conversions>
        fullname <identifier_node 0x7fffea05b300 type
            normal local bindings <(nil)>>
        chain <type_decl 0x7fffea06ee40 type>>
    VOID
    align:8 warn_if_not_align:0 symtab:0 alias-set -1 structural-equality>
$28 = void
(gdb) p debug_tree (0x7fffea1b4888)
 <type_pack_expansion 0x7fffea1b4888
    type <typename_type 0x7fffea1b47e0 implicit_conversions_t VOID
        align:8 warn_if_not_align:0 symtab:0 alias-set -1 structural-equality
context <record_type 0x7fffea1b4690 implicit_conversions>
        fullname <identifier_node 0x7fffea05b300 type
            normal local bindings <(nil)>>
        chain <type_decl 0x7fffea06ee40 type>>
    type_0 type_6 VOID
    align:8 warn_if_not_align:0 symtab:0 alias-set -1 structural-equality>
$29 = void

and flag_concepts is false.  So the template args hash differently but compare
equal.

[Bug c++/99445] [11 Regression] ICE in hashtab_chk_error, at hash-table.c:137 since r11-7011-g6e0a231a4aa2407b

[marxin at gcc dot gnu.org]

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=99445

--- Comment #7 from Martin Liška <marxin at gcc dot gnu.org> ---
@Jason: Can you please take a look?

[Bug c++/99445] [11 Regression] ICE in hashtab_chk_error, at hash-table.c:137 since r11-7011-g6e0a231a4aa2407b

[jason at gcc dot gnu.org]

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=99445

Jason Merrill <jason at gcc dot gnu.org> changed:

           What    |Removed                     |Added
----------------------------------------------------------------------------
           Assignee|unassigned at gcc dot gnu.org      |jason at gcc dot gnu.org
             Status|NEW                         |ASSIGNED

[Bug c++/99445] [11 Regression] ICE in hashtab_chk_error, at hash-table.c:137 since r11-7011-g6e0a231a4aa2407b

[cvs-commit at gcc dot gnu.org]

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=99445

--- Comment #8 from CVS Commits <cvs-commit at gcc dot gnu.org> ---
The master branch has been updated by Jason Merrill <jason@gcc.gnu.org>:

https://gcc.gnu.org/g:a2531859bf5bf6cf1f29c0dca85fd26e80904a5d

commit r11-7931-ga2531859bf5bf6cf1f29c0dca85fd26e80904a5d
Author: Jason Merrill <jason@redhat.com>
Date:   Tue Mar 30 20:31:18 2021 -0400

    c++: Alias template in pack expansion [PR99445]

    In this testcase, iterative_hash_template_arg checks
    alias_template_specialization_p to determine whether to treat a type as a
    dependent alias, and structural_comptypes checks
    dependent_alias_template_spec_p.  Normally that difference isn't a problem
    because canonicalizing template arguments strips non-dependent aliases, but
    that wasn't happening for the pack expansion.  Fixed thus.

    gcc/cp/ChangeLog:

            PR c++/99445
            * tree.c (strip_typedefs): Handle TYPE_PACK_EXPANSION.

    gcc/testsuite/ChangeLog:

            PR c++/99445
            * g++.dg/cpp0x/alias-decl-variadic1.C: New test.

[Bug c++/99445] [11 Regression] ICE in hashtab_chk_error, at hash-table.c:137 since r11-7011-g6e0a231a4aa2407b

[jason at gcc dot gnu.org]

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=99445

Jason Merrill <jason at gcc dot gnu.org> changed:

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

--- Comment #9 from Jason Merrill <jason at gcc dot gnu.org> ---
Fixed.

[Bug c++/99445] [11 Regression] ICE in hashtab_chk_error, at hash-table.c:137 since r11-7011-g6e0a231a4aa2407b

[redi at gcc dot gnu.org]

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=99445

--- Comment #10 from Jonathan Wakely <redi at gcc dot gnu.org> ---
There's a report of a regression caused by this:
https://gcc.gnu.org/pipermail/gcc-help/2022-January/141127.html
I'll ask for it to be reported to bugzilla.

[Bug c++/99445] [11 Regression] ICE in hashtab_chk_error, at hash-table.c:137 since r11-7011-g6e0a231a4aa2407b

[redi at gcc dot gnu.org]

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=99445

Jonathan Wakely <redi at gcc dot gnu.org> changed:

           What    |Removed                     |Added
----------------------------------------------------------------------------
           See Also|                            |https://gcc.gnu.org/bugzill
                   |                            |a/show_bug.cgi?id=104008

--- Comment #11 from Jonathan Wakely <redi at gcc dot gnu.org> ---
(In reply to Jonathan Wakely from comment #10)
> There's a report of a regression caused by this:
> https://gcc.gnu.org/pipermail/gcc-help/2022-January/141127.html
> I'll ask for it to be reported to bugzilla.

That's now PR 104008

[Bug c++/99445] [11 Regression] ICE in hashtab_chk_error, at hash-table.c:137 since r11-7011-g6e0a231a4aa2407b

[cvs-commit at gcc dot gnu.org]

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=99445

--- Comment #12 from CVS Commits <cvs-commit at gcc dot gnu.org> ---
The master branch has been updated by Jason Merrill <jason@gcc.gnu.org>:

https://gcc.gnu.org/g:07be8f8da4c6840a1fd6b2229b147e50cc6f03dc

commit r12-7848-g07be8f8da4c6840a1fd6b2229b147e50cc6f03dc
Author: Jason Merrill <jason@redhat.com>
Date:   Fri Mar 25 11:26:06 2022 -0400

    c++: ICE with alias in pack expansion [PR103769]

    This was breaking because when we stripped the 't' typedef in s<t<Args>...>
    to be s<Args...>, the TYPE_MAIN_VARIANT of "Args..." was still
    "t<Args>...", because type pack expansions are treated as types.  Fixed by
    using the right function to copy a "type".

            PR c++/99445
            PR c++/103769

    gcc/cp/ChangeLog:

            * tree.cc (strip_typedefs): Use build_distinct_type_copy.

    gcc/testsuite/ChangeLog:

            * g++.dg/cpp0x/variadic-alias5.C: New test.

[Bug c++/99445] [11 Regression] ICE in hashtab_chk_error, at hash-table.c:137 since r11-7011-g6e0a231a4aa2407b

[cvs-commit at gcc dot gnu.org]

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=99445

--- Comment #13 from CVS Commits <cvs-commit at gcc dot gnu.org> ---
The releases/gcc-11 branch has been updated by Jason Merrill
<jason@gcc.gnu.org>:

https://gcc.gnu.org/g:3bf2e1bfc9858516e028137b313e4c689b0c8cd4

commit r11-9707-g3bf2e1bfc9858516e028137b313e4c689b0c8cd4
Author: Jason Merrill <jason@redhat.com>
Date:   Fri Mar 25 11:26:06 2022 -0400

    c++: ICE with alias in pack expansion [PR103769]

    This was breaking because when we stripped the 't' typedef in s<t<Args>...>
    to be s<Args...>, the TYPE_MAIN_VARIANT of "Args..." was still
    "t<Args>...", because type pack expansions are treated as types.  Fixed by
    using the right function to copy a "type".

            PR c++/99445
            PR c++/103769

    gcc/cp/ChangeLog:

            * tree.c (strip_typedefs): Use build_distinct_type_copy.

    gcc/testsuite/ChangeLog:

            * g++.dg/cpp0x/variadic-alias5.C: New test.