From mboxrd@z Thu Jan  1 00:00:00 1970
Return-Path: <gcc-prs-return-33729-listarch-gcc-prs=gcc.gnu.org@gcc.gnu.org>
Received: (qmail 19131 invoked by alias); 28 Apr 2003 22:46:41 -0000
Mailing-List: contact gcc-prs-help@gcc.gnu.org; run by ezmlm
Precedence: bulk
List-Archive: <http://gcc.gnu.org/ml/gcc-prs/>
List-Post: <mailto:gcc-prs@gcc.gnu.org>
List-Help: <mailto:gcc-prs-help@gcc.gnu.org>
Sender: gcc-prs-owner@gcc.gnu.org
Received: (qmail 19111 invoked by uid 48); 28 Apr 2003 22:46:40 -0000
Date: Mon, 28 Apr 2003 22:46:00 -0000
Message-ID: <20030428224640.19110.qmail@sources.redhat.com>
To: gcc-bugs@gcc.gnu.org, gcc-prs@gcc.gnu.org, jkanze@caicheuvreuse.com,
  nobody@gcc.gnu.org, pcarlini@unitus.it
From: bangerth@dealii.org
Reply-To: bangerth@dealii.org, gcc-bugs@gcc.gnu.org, gcc-prs@gcc.gnu.org,
  jkanze@caicheuvreuse.com, nobody@gcc.gnu.org, pcarlini@unitus.it,
  gcc-gnats@gcc.gnu.org
Subject: Re: c++/10529: C++ front-end troubles with uninitialized_fill_n
X-SW-Source: 2003-04/txt/msg01299.txt.bz2
List-Id: <gcc-prs.sourceware.org>

Synopsis: C++ front-end troubles with uninitialized_fill_n

State-Changed-From-To: open->feedback
State-Changed-By: bangerth
State-Changed-When: Mon Apr 28 22:46:40 2003
State-Changed-Why:
    Hm, I'm not sure I understand what is going on. Your code
    reduces to this:
    -----------------------
    #include <memory>
    
    class X {
      public:
        X(int i);
      private:
        X(const X&);
    };
    
    int main() {
      std::uninitialized_fill_n( (X*)0, 3, X(2) ) ;
    }
    -------------------------
    which indeed fails to compile since X has no public copy
    constructor. However, the draft standard has this wording:
    
    20.4.4.3 uninitialized_fill_n              [lib.uninitialized.fill.n]
    
    template <class ForwardIterator, class Size, class T>
    void uninitialized_fill_n(ForwardIterator first, Size n, const T& x);
    
      Effects:
    
    while (n--)
              new (static_cast<void*>(&*result++))
                      typename iterator_traits<ForwardIterator>::value_type(*first++);
    
    So it uses placement new to generate a -- well, that
    seems like a defect, but if we take the respective
    statement from right above that place (unitialized_fill),
    it reads
    
     new (static_cast<void*>(&*first++))
                      typename iterator_traits<ForwardIterator>::value_type(x);
    
    So that would make a copy of x, no? I would say that then
    we need a public copy constructor. What am I missing?
    
    W.

http://gcc.gnu.org/cgi-bin/gnatsweb.pl?cmd=view%20audit-trail&database=gcc&pr=10529