[Bug c++/56176] New: internal compiler error with call to default constructor of inner struct having a std::function member attribute initialization

[Bug c++/56176] New: internal compiler error with call to default constructor of inner struct having a std::function member attribute initialization

[benoit.bayol at gmail dot com]

http://gcc.gnu.org/bugzilla/show_bug.cgi?id=56176

             Bug #: 56176
           Summary: internal compiler error with call to default
                    constructor of inner struct having a std::function
                    member attribute initialization
    Classification: Unclassified
           Product: gcc
           Version: 4.7.2
            Status: UNCONFIRMED
          Severity: blocker
          Priority: P3
         Component: c++
        AssignedTo: unassigned@gcc.gnu.org
        ReportedBy: benoit.bayol@gmail.com


Hello,

Here is the test code :

#include <functional>

typedef std::function<auto (int, int) -> int> fn_type;

auto f1 = [](int,int){return 0;};

struct B
{
    typedef struct
        {
            fn_type _f = f1;
    } A;
    B(A a = A()){}
};

[Bug c++/56176] internal compiler error with call to default constructor of inner struct having a std::function member attribute initialization

[benoit.bayol at gmail dot com]

http://gcc.gnu.org/bugzilla/show_bug.cgi?id=56176

--- Comment #1 from Benoit Bayol <benoit.bayol at gmail dot com> 2013-02-01 16:05:31 UTC ---
I have just noticed that the example code can be simpler :

struct B
{
    typedef struct
        {
            int a = 0;
    } A;
    B(A a = A()){}
};

[Bug c++/56176] internal compiler error with call to default constructor of inner struct having a std::function member attribute initialization

[redi at gcc dot gnu.org]

http://gcc.gnu.org/bugzilla/show_bug.cgi?id=56176

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

           What    |Removed                     |Added
----------------------------------------------------------------------------
           Severity|blocker                     |normal

[Bug c++/56176] ICE with call to synthesized default constructor of inner struct with NSDMI

[redi at gcc dot gnu.org]

http://gcc.gnu.org/bugzilla/show_bug.cgi?id=56176

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

           What    |Removed                     |Added
----------------------------------------------------------------------------
           Keywords|                            |ice-on-invalid-code
             Status|UNCONFIRMED                 |NEW
   Last reconfirmed|                            |2013-02-01
      Known to work|                            |4.8.0
            Summary|internal compiler error     |ICE with call to
                   |with call to default        |synthesized default
                   |constructor of inner struct |constructor of inner struct
                   |having a std::function      |with NSDMI
                   |member attribute            |
                   |initialization              |
     Ever Confirmed|0                           |1
      Known to fail|                            |4.7.2

--- Comment #2 from Jonathan Wakely <redi at gcc dot gnu.org> 2013-02-01 16:11:33 UTC ---
Trunk rejects it as invalid:

t.cc: In constructor 'constexpr B::A::A()':
t.cc:4:9: error: constructor required before non-static data member for
'B::A::a' has been parsed
         {
         ^
t.cc: At global scope:
t.cc:7:15: note: synthesized method 'constexpr B::A::A()' first required here
     B(A a = A()){}
               ^

[Bug c++/56176] ICE with call to synthesized default constructor of inner struct with NSDMI

[paolo.carlini at oracle dot com]

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=56176

--- Comment #3 from Paolo Carlini <paolo.carlini at oracle dot com> ---
I'm going to add the reduced testcase and close this as fixed.

[Bug c++/56176] ICE with call to synthesized default constructor of inner struct with NSDMI

[paolo.carlini at oracle dot com]

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=56176

Paolo Carlini <paolo.carlini at oracle dot com> changed:

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

--- Comment #4 from Paolo Carlini <paolo.carlini at oracle dot com> ---
In fact we already got nsdmi-defer6.C.