[Bug libstdc++/64658] New: std::atomic_init() undefined

[Bug libstdc++/64658] New: std::atomic_init() undefined

[redi at gcc dot gnu.org]

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=64658

            Bug ID: 64658
           Summary: std::atomic_init() undefined
           Product: gcc
           Version: 4.9.2
            Status: UNCONFIRMED
          Keywords: rejects-valid
          Severity: normal
          Priority: P3
         Component: libstdc++
          Assignee: unassigned at gcc dot gnu.org
          Reporter: redi at gcc dot gnu.org

#include <atomic>

int main()
{
  std::atomic_int a;
  atomic_init(&a, 0);
}

$ g++ -std=c++11 at.cc
In file included from at.cc:1:0:
/home/jwakely/gcc/5/include/c++/5.0.0/atomic:924:5: warning: inline function
‘void std::atomic_init(std::atomic<_ITp>*, _ITp) [with _ITp = int]’ used but
never defined
     atomic_init(atomic<_ITp>* __a, _ITp __i) noexcept;
     ^
/tmp/ccZgPnay.o: In function `main':
/tmp/at.cc:6: undefined reference to `void
std::atomic_init<int>(std::atomic<int>*, int)'
collect2: error: ld returned 1 exit status
>From gcc-bugs-return-473732-listarch-gcc-bugs=gcc.gnu.org@gcc.gnu.org Sun Jan 18 17:24:22 2015
Return-Path: <gcc-bugs-return-473732-listarch-gcc-bugs=gcc.gnu.org@gcc.gnu.org>
Delivered-To: listarch-gcc-bugs@gcc.gnu.org
Received: (qmail 10457 invoked by alias); 18 Jan 2015 17:24:04 -0000
Mailing-List: contact gcc-bugs-help@gcc.gnu.org; run by ezmlm
Precedence: bulk
List-Id: <gcc-bugs.gcc.gnu.org>
List-Archive: <http://gcc.gnu.org/ml/gcc-bugs/>
List-Post: <mailto:gcc-bugs@gcc.gnu.org>
List-Help: <mailto:gcc-bugs-help@gcc.gnu.org>
Sender: gcc-bugs-owner@gcc.gnu.org
Delivered-To: mailing list gcc-bugs@gcc.gnu.org
Received: (qmail 10278 invoked by uid 48); 18 Jan 2015 17:22:57 -0000
From: "redi at gcc dot gnu.org" <gcc-bugzilla@gcc.gnu.org>
To: gcc-bugs@gcc.gnu.org
Subject: [Bug libstdc++/64658] std::atomic_init() undefined
Date: Sun, 18 Jan 2015 17:24:00 -0000
X-Bugzilla-Reason: CC
X-Bugzilla-Type: changed
X-Bugzilla-Watch-Reason: None
X-Bugzilla-Product: gcc
X-Bugzilla-Component: libstdc++
X-Bugzilla-Version: 4.9.2
X-Bugzilla-Keywords: rejects-valid
X-Bugzilla-Severity: normal
X-Bugzilla-Who: redi at gcc dot gnu.org
X-Bugzilla-Status: NEW
X-Bugzilla-Priority: P3
X-Bugzilla-Assigned-To: unassigned at gcc dot gnu.org
X-Bugzilla-Target-Milestone: ---
X-Bugzilla-Flags:
X-Bugzilla-Changed-Fields: bug_status cf_reconfirmed_on everconfirmed
Message-ID: <bug-64658-4-0G1E9KUVlx@http.gcc.gnu.org/bugzilla/>
In-Reply-To: <bug-64658-4@http.gcc.gnu.org/bugzilla/>
References: <bug-64658-4@http.gcc.gnu.org/bugzilla/>
Content-Type: text/plain; charset="UTF-8"
Content-Transfer-Encoding: 7bit
X-Bugzilla-URL: http://gcc.gnu.org/bugzilla/
Auto-Submitted: auto-generated
MIME-Version: 1.0
X-SW-Source: 2015-01/txt/msg01726.txt.bz2
Content-length: 1300

https://gcc.gnu.org/bugzilla/show_bug.cgi?idd658

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

           What    |Removed                     |Added
----------------------------------------------------------------------------
             Status|UNCONFIRMED                 |NEW
   Last reconfirmed|                            |2015-01-18
     Ever confirmed|0                           |1

--- Comment #1 from Jonathan Wakely <redi at gcc dot gnu.org> ---
Although atomic_init doesn't need to be atomic the simplest way to implement it
is just with a relaxed store:

--- a/libstdc++-v3/include/std/atomic
+++ b/libstdc++-v3/include/std/atomic
@@ -921,11 +921,13 @@ _GLIBCXX_BEGIN_NAMESPACE_VERSION

   template<typename _ITp>
     inline void
-    atomic_init(atomic<_ITp>* __a, _ITp __i) noexcept;
+    atomic_init(atomic<_ITp>* __a, _ITp __i) noexcept
+    { __a->store(__i, memory_order_relaxed); }

   template<typename _ITp>
     inline void
-    atomic_init(volatile atomic<_ITp>* __a, _ITp __i) noexcept;
+    atomic_init(volatile atomic<_ITp>* __a, _ITp __i) noexcept
+    { __a->store(__i, memory_order_relaxed); }

   template<typename _ITp>
     inline void


A better implementation would make atomic_init friend of std::atomic so it can
set the private data member directly.

[Bug libstdc++/64658] std::atomic_init() undefined

[redi at gcc dot gnu.org]

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=64658

--- Comment #2 from Jonathan Wakely <redi at gcc dot gnu.org> ---
Testcase for 4.9 (which doesn't have the fix for PR64940 that allows
std::atomic_int t o be used interchangeably with std::atomic<int>):

#include <atomic>

int main()
{
  std::atomic<int> a;
  atomic_init(&a, 0);
}

[Bug libstdc++/64658] std::atomic_init() undefined

[redi at gcc dot gnu.org]

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=64658

--- Comment #3 from Jonathan Wakely <redi at gcc dot gnu.org> ---
Author: redi
Date: Tue Jan 20 11:50:43 2015
New Revision: 219886

URL: https://gcc.gnu.org/viewcvs?rev=219886&root=gcc&view=rev
Log:
    PR libstdc++/64658
    * include/std/atomic (atomic_init): Define.
    * testsuite/29_atomics/atomic/64658.cc: New.

Added:
    trunk/libstdc++-v3/testsuite/29_atomics/atomic/64658.cc
Modified:
    trunk/libstdc++-v3/ChangeLog
    trunk/libstdc++-v3/include/std/atomic

[Bug libstdc++/64658] std::atomic_init() undefined

[redi at gcc dot gnu.org]

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=64658

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

           What    |Removed                     |Added
----------------------------------------------------------------------------
             Status|NEW                         |RESOLVED
         Resolution|---                         |FIXED
   Target Milestone|---                         |5.0

--- Comment #4 from Jonathan Wakely <redi at gcc dot gnu.org> ---
fixed - probably not worth backporting since noone has noticed it's missing
(and these functions only really exist because of the original plan for C
compatibility, in C++ you can just use the constructor)

Re: [Bug libstdc++/64658] std::atomic_init() undefined

[Jon Chesterfield]

> --- Comment #4 from Jonathan Wakely <redi at gcc dot gnu.org> ---
> fixed - probably not worth backporting since noone has noticed it's missing


For what it's worth, at least one of your users (me) has noticed this. 
I'll restructure code as necessary to use the constructor instead, i.e. 
std::atomic<std::size_t> foo(0);

Regards,

Jon