Re: Is it a bug or doing it wrong (sample code using std::variant)?

[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

 



On 2020-10-28 13:56, Nicolai G. via Gcc-help wrote:
Hi,

hi,

look at this: https://en.cppreference.com/w/cpp/utility/variant/monostate


I've recently switched to GCC 9.3 (the same error on 10.2) and have a
compile issue.
The reproduction is the following:

--------------------------------------
#include <variant>
#include <string>
#include <vector>

struct response_t
{
    struct response_a_t
    {
        std::string raw_data{};
    };

    struct response_b_t {};

    std::variant< response_a_t, response_b_t > resp{};
};

void handle( response_t & r );

response_t test()
{
    response_t r;
    handle( r );
    return r;
}
--------------------------------------
<source>:18:53: error: use of deleted function
'std::variant<_Types>::variant() [with _Types = {response_t::response_a_t,
response_t::response_b_t}]'

   13 |     std::variant< response_a_t, response_b_t > resp{};

      |                                                     ^

In file included from <source>:3:

/opt/compiler-explorer/gcc-10.2.0/include/c++/10.2.0/variant:1347:7: note: 'std::variant<_Types>::variant() [with _Types = {response_t::response_a_t,
response_t::response_b_t}]' is implicitly deleted because the default
definition would be ill-formed:

 1347 |       variant() = default;

      |       ^~~~~~~

/opt/compiler-explorer/gcc-10.2.0/include/c++/10.2.0/variant:1347:7: error: use of deleted function 'constexpr std::_Enable_default_constructor<false,
_Tag>::_Enable_default_constructor() [with _Tag =
std::variant<response_t::response_a_t, response_t::response_b_t>]'

In file included from
/opt/compiler-explorer/gcc-10.2.0/include/c++/10.2.0/variant:38,

                 from <source>:3:

/opt/compiler-explorer/gcc-10.2.0/include/c++/10.2.0/bits/enable_special_members.h:110:15:
note: declared here

  110 |     constexpr _Enable_default_constructor() noexcept = delete;

      |               ^~~~~~~~~~~~~~~~~~~~~~~~~~~

Compiler returned: 1
--------------------------------------

It worked on 8.3. And I don't see anything wrong with this snippet.
if *response_a_t* and *response_b_t* are not nested, then compilation goes
fine.
Compilation is also ok if I put
*response_a_t the following way:*
    struct response_a_t
    {
        std::string raw_data; // no {}
    };

Reproduction on compile explorer: https://godbolt.org/z/ETzsqo

Can you, please, clarify on the subject?



[Index of Archives]     [Linux C Programming]     [Linux Kernel]     [eCos]     [Fedora Development]     [Fedora Announce]     [Autoconf]     [The DWARVES Debugging Tools]     [Yosemite Campsites]     [Yosemite News]     [Linux GCC]

  Powered by Linux