[Bug c++/60258] Member initialization for atomic fail.

[paolo.carlini at oracle dot com]

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=60258

Paolo Carlini paolo.carlini at oracle dot com changed:

   What|Removed |Added

 Status|NEW |RESOLVED
 Resolution|--- |DUPLICATE

--- Comment #6 from Paolo Carlini paolo.carlini at oracle dot com ---
Dup, already fixed.

*** This bug has been marked as a duplicate of bug 58930 ***

[Bug c++/60258] Member initialization for atomic fail.

[redi at gcc dot gnu.org]

http://gcc.gnu.org/bugzilla/show_bug.cgi?id=60258

Jonathan Wakely redi at gcc dot gnu.org changed:

   What|Removed |Added

 Status|UNCONFIRMED |NEW
   Last reconfirmed||2014-02-19
 Ever confirmed|0   |1

[Bug c++/60258] Member initialization for atomic fail.

[jakub at gcc dot gnu.org]

http://gcc.gnu.org/bugzilla/show_bug.cgi?id=60258

Jakub Jelinek jakub at gcc dot gnu.org changed:

   What|Removed |Added

 CC||jakub at gcc dot gnu.org

--- Comment #1 from Jakub Jelinek jakub at gcc dot gnu.org ---
Why the
auto my_func() - void
rather than
void my_func()
?  The fact that C++11 allows the former form doesn't mean the latter form is
deprecated, and in the case when the return type doesn't depend on argument
types it doesn't make any sense to use that form.

[Bug c++/60258] Member initialization for atomic fail.

[ja.gcc.bugzilla at aptsketch dot com]

http://gcc.gnu.org/bugzilla/show_bug.cgi?id=60258

--- Comment #2 from ja.gcc.bugzilla at aptsketch dot com ---
auto foo() - void
vs
void foo()

is more of just a stylistic issue. There is nothing wrong with both. To me both
are good.

The bug is about std::atomic member initialization. Where if there is member
initialization std::atomic_size_t x{10} the compiler chokes on it.

[Bug c++/60258] Member initialization for atomic fail.

[jakub at gcc dot gnu.org]

http://gcc.gnu.org/bugzilla/show_bug.cgi?id=60258

--- Comment #3 from Jakub Jelinek jakub at gcc dot gnu.org ---
Sure, my comment was about this stylistic issue, not about the real bug (if
any, haven't tried it).  Note your testcase isn't self-contained, you probably
need #include atomic.

[Bug c++/60258] Member initialization for atomic fail.

[daniel.kruegler at googlemail dot com]

http://gcc.gnu.org/bugzilla/show_bug.cgi?id=60258

Daniel Krügler daniel.kruegler at googlemail dot com changed:

   What|Removed |Added

 CC||daniel.kruegler@googlemail.
   ||com

--- Comment #4 from Daniel Krügler daniel.kruegler at googlemail dot com ---
Reduced example:

//--
#include atomic

template class
struct base
{
  std::atomic_bool magic{false};
};

int main() 
{
  basevoid b;
}
//--

Diagnostics (gcc HEAD 4.9.0 20140215 (experimental)):

quote
prog.cc: In constructor 'constexpr basevoid::base()': 
prog.cc:4:8: error: use of deleted function
'std::atomic_bool::atomic_bool(const std::atomic_bool)' 
  struct base 
 ^ 
In file included from prog.cc:1:0: 
/usr/local/gcc-head/include/c++/4.9.0/atomic:62:5: note: declared here
  atomic_bool(const atomic_bool) = delete; 
  ^ 
prog.cc: In function 'int main()': 
prog.cc:11:14: note: synthesized method 'constexpr basevoid::base()' first
required here 
  basevoid b; 
  ^
/quote

The error doesn't occur, if the class template is replaced by a non-template
class.

[Bug c++/60258] Member initialization for atomic fail.

[daniel.kruegler at googlemail dot com]

http://gcc.gnu.org/bugzilla/show_bug.cgi?id=60258

--- Comment #5 from Daniel Krügler daniel.kruegler at googlemail dot com ---
(In reply to Daniel Krügler from comment #4)
 Reduced example:

More reduced by eliminating library dependencies:

//--
struct atomic_bool
{
  atomic_bool(const atomic_bool) = delete;
  constexpr atomic_bool(bool) noexcept { }
};

#define USE_TEMPLATE

#ifdef USE_TEMPLATE
template class
#endif
struct base
{
  atomic_bool magic{false};
};

int main()
{
#ifdef USE_TEMPLATE
  basevoid b;
#else
  base b;
#endif
}
//--