[Bug c++/95564] New: GCC rejects lambda expression with "noexcept(1+1)"

[Bug c++/95564] New: GCC rejects lambda expression with "noexcept(1+1)"

[haoxintu at gmail dot com]

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=95564

            Bug ID: 95564
           Summary: GCC rejects lambda expression with "noexcept(1+1)"
           Product: gcc
           Version: 11.0
            Status: UNCONFIRMED
          Severity: normal
          Priority: P3
         Component: c++
          Assignee: unassigned at gcc dot gnu.org
          Reporter: haoxintu at gmail dot com
  Target Milestone: ---

This code test.cc

int main(){
    auto a = [ & ] (...) noexcept(1+1) {};
    return 0;
}

$g++-trunk test.cc
test.cc:2:38: error: narrowing conversion of '2' from 'int' to 'bool'
[-Wnarrowing]
    2 |     auto a = [ & ] (...) noexcept(1+1) {};
      |                                  

GCC fails to compile it, while clang, icc, or msvc accept it.

[Bug c++/95564] GCC rejects lambda expression with "noexcept(1+1)"

[mpolacek at gcc dot gnu.org]

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=95564

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

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

--- Comment #1 from Marek Polacek <mpolacek at gcc dot gnu.org> ---
There was some discussion on the Core C++ mailing list about whether a
conversion to bool should be narrowing, and though opinions differed, I think
gcc isn't incorrect in rejecting this code.

[Bug c++/95564] GCC rejects lambda expression with "noexcept(1+1)"

[pinskia at gcc dot gnu.org]

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=95564

Andrew Pinski <pinskia at gcc dot gnu.org> changed:

           What    |Removed                     |Added
----------------------------------------------------------------------------
           See Also|                            |https://gcc.gnu.org/bugzill
                   |                            |a/show_bug.cgi?id=87724

--- Comment #2 from Andrew Pinski <pinskia at gcc dot gnu.org> ---
GCC is inconsitent here. with static_assert and constexpr GCC accepts it (PR
87724) while rejects it for noexcept.

[Bug c++/95564] GCC rejects lambda expression with "noexcept(1+1)"

[de34 at live dot cn]

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=95564

Jiang An <de34 at live dot cn> changed:

           What    |Removed                     |Added
----------------------------------------------------------------------------
                 CC|                            |de34 at live dot cn

--- Comment #3 from Jiang An <de34 at live dot cn> ---
(In reply to Andrew Pinski from comment #2)
> GCC is inconsitent here. with static_assert and constexpr GCC accepts it (PR
> 87724) while rejects it for noexcept.

Currently narrowing conversions are still forbidden in noexcept and explicit
specifiers.

The inconsistency is acknowledged in C++23 (WG21-P1401R5), but it seems that
P1401 is not a DR, as the related CWG 2320 is closed as extension. I don't know
whether we should reject narrowing conversions in static_assert/if constexpr in
earlier modes.

[Bug c++/95564] GCC rejects lambda expression with "noexcept(1+1)"

[jason at gcc dot gnu.org]

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=95564

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

           What    |Removed                     |Added
----------------------------------------------------------------------------
                 CC|                            |jason at gcc dot gnu.org
         Resolution|---                         |INVALID
             Status|UNCONFIRMED                 |RESOLVED

--- Comment #4 from Jason Merrill <jason at gcc dot gnu.org> ---
(In reply to Jiang An from comment #3)
> (In reply to Andrew Pinski from comment #2)
> > GCC is inconsitent here. with static_assert and constexpr GCC accepts it (PR
> > 87724) while rejects it for noexcept.
> 
> Currently narrowing conversions are still forbidden in noexcept and explicit
> specifiers.
> 
> The inconsistency is acknowledged in C++23 (WG21-P1401R5), but it seems that
> P1401 is not a DR, as the related CWG 2320 is closed as extension. I don't
> know whether we should reject narrowing conversions in static_assert/if
> constexpr in earlier modes.

Narrowing conversions were fine in earlier standards, it was an accidental
change that made them ill-formed, which was fixed.

Clang now also rejects the testcase.