220 13944 <97718d4b-0b69-40c1-b3e3-e0ab196d6f10@isocpp.org> article Path: news.gmane.org!not-for-mail From: Myriachan Newsgroups: gmane.comp.lang.c++.isocpp.proposals Subject: Re: Reconstruct function idea Date: Wed, 15 Oct 2014 19:03:00 -0700 (PDT) Lines: 150 Approved: news@gmane.org Message-ID: <97718d4b-0b69-40c1-b3e3-e0ab196d6f10@isocpp.org> References: <8a74f08b-7453-4d1a-9f20-bc1d3866617c@isocpp.org> <27e2ae9c-8b1f-4688-9bb3-9caef6bf4f54@isocpp.org> Reply-To: std-proposals@isocpp.org NNTP-Posting-Host: plane.gmane.org Mime-Version: 1.0 Content-Type: multipart/alternative; boundary="----=_Part_123_1538419150.1413424980736" X-Trace: ger.gmane.org 1413424996 7712 80.91.229.3 (16 Oct 2014 02:03:16 GMT) X-Complaints-To: usenet@ger.gmane.org NNTP-Posting-Date: Thu, 16 Oct 2014 02:03:16 +0000 (UTC) To: std-proposals@isocpp.org Original-X-From: std-proposals+bncBDKLT4PURQHRBVWO7SQQKGQEEETLDMQ@isocpp.org Thu Oct 16 04:03:04 2014 Return-path: Envelope-to: gclcip-std-proposals@m.gmane.org Original-Received: from mail-vc0-f198.google.com ([209.85.220.198]) by plane.gmane.org with esmtp (Exim 4.69) (envelope-from ) id 1XeaPA-0008UU-6M for gclcip-std-proposals@m.gmane.org; Thu, 16 Oct 2014 04:03:04 +0200 Original-Received: by mail-vc0-f198.google.com with SMTP id hq12sf4991233vcb.5 for ; Wed, 15 Oct 2014 19:03:03 -0700 (PDT) DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=gmail.com; s=20120113; h=date:from:to:message-id:in-reply-to:references:subject:mime-version :x-original-sender:reply-to:precedence:mailing-list:list-id :list-post:list-help:list-archive:list-subscribe:list-unsubscribe :content-type; bh=djGJnoBxLgVFqmAfDg3cWWAocS69OMyz79r+jw9aNOg=; b=S02B16avkgy5HoEV5smkBlB5l8oahsLiiz3s4QRqdKiD0YNKx0rLdjjjfu+YfEXSD7 jVpdDJWXWGtFVa8XxSLVMusHOI1zhRb/gLYJlDZ+3AHgFkJDwyFLGtOszPpSCpzpjLBx ww3PiFOoy6dkoMu47QycxncroVICqrvJTPQIyKbUReqLKdkzhpyaYhpVRofSobSyEX90 iAYrBIWmlyug8tL2Z8wj14dBikw3HV7Ps1w8s6d41AK8hg50Gw6fEyDYnjbFzWL0hZJu E1mHa3qVIS8HmAIiIVh9ENKkRFo5GEn3Pvnkeo58YbU+fHHhbDB/hI43l2dqb2DDLHVS I7mw== X-Google-DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=1e100.net; s=20130820; h=x-gm-message-state:date:from:to:message-id:in-reply-to:references :subject:mime-version:x-original-sender:reply-to:precedence :mailing-list:list-id:list-post:list-help:list-archive :list-subscribe:list-unsubscribe:content-type; bh=djGJnoBxLgVFqmAfDg3cWWAocS69OMyz79r+jw9aNOg=; b=Bo9sDqYmy5wb3pOaCnAaiJ8hnEwprFklF0+oAA7q5ycTAtsS3hZ7xaUP67e8+m4Fp8 oC7TjIwA7zAcnQQU1R/uj1XRwCmxdXh1uXjSAO/87sBTgGDF5x7T91LIld2lhX0BVjGp dYwI6U+XBCp13qV08sy770nRGatVQGrGmvRnIXvEY+pkZKmpeTp9jrkQVSfurLW0oiD4 lVFUq6Gk1w5jbMiG1zFcwkS65op9WAg3S4HywXrTYxFAbLzyKvIrpiHjnCaX6k6lDWOd nk3z/lraztuBQTljc1x2BGaM0ol3JTANiR8wGtP9R0ahjfOrx86djoPhmEPuE5jOrgkP cXiw== X-Gm-Message-State: ALoCoQnQZHJT3YDRy2W/EEQ5yZaUksJhs+Lk4KZI0iMJfQgvF7ITG1Fz0gFbxPAQLoe9HFFdQ2xe X-Received: by 10.236.198.165 with SMTP id v25mr11392242yhn.51.1413424983251; Wed, 15 Oct 2014 19:03:03 -0700 (PDT) X-BeenThere: std-proposals@isocpp.org Original-Received: by 10.50.107.72 with SMTP id ha8ls450203igb.10.gmail; Wed, 15 Oct 2014 19:03:02 -0700 (PDT) X-Received: by 10.50.134.69 with SMTP id pi5mr271019igb.2.1413424982510; Wed, 15 Oct 2014 19:03:02 -0700 (PDT) In-Reply-To: X-Original-Sender: myriachan@gmail.com Precedence: list Mailing-list: list std-proposals@isocpp.org; contact std-proposals+owners@isocpp.org List-ID: X-Google-Group-Id: 399137483710 List-Post: , List-Help: , List-Archive: List-Subscribe: , List-Unsubscribe: , Xref: news.gmane.org gmane.comp.lang.c++.isocpp.proposals:13944 Archived-At: ------=_Part_123_1538419150.1413424980736 Content-Type: text/plain; charset=UTF-8 On Wednesday, October 15, 2014 3:31:39 PM UTC-7, Miro Knejp wrote: > > On 16 Oct 2014, at 00:22 , Myriachan > > wrote: > > The main reason to use something like this is for copy constructors. > Sometimes, the most convenient way to implement a copy or move constructor > is like this: > > T &T::operator =(const T ©) > { > if (© != this) > { > this->~T(); > new(this) T(copy); > } > return *this; > } > > ...and equivalently with std::forward for move constructors. > > > This would never pass my code review. If that constructor throws you're > doomed. Use copy-and-swap instead, it provides strong exception guarantee > and won't leave you with a destroyed object. Move constructors are a > different game though and the compiler-defaulted one should do it's job > just fine in most cases. > > T& T::operator =(const T& copy) > { > if (© != this) > { > T x{copy}; > swap(*this, x); > } > return *this; > } > > I suppose the difference is that I work on a project that disables C++ exceptions because they're too slow. swap would be slow or unusable if there's either no overload for it nor move constructors, which for a codebase of millions of lines written before C++11 is likely. Enabling C++ exception handling is a massive performance decrease in x86-32 Windows. If it's enabled in the compiler, even if no exceptions are used, every function with an automatic variable whose type has a destructor will be ~20 instructions longer due to setting up the exception frame. Melissa -- --- You received this message because you are subscribed to the Google Groups "ISO C++ Standard - Future Proposals" group. To unsubscribe from this group and stop receiving emails from it, send an email to std-proposals+unsubscribe@isocpp.org. To post to this group, send email to std-proposals@isocpp.org. Visit this group at http://groups.google.com/a/isocpp.org/group/std-proposals/. ------=_Part_123_1538419150.1413424980736 Content-Type: text/html; charset=UTF-8 Content-Transfer-Encoding: quoted-printable
On Wednesday, October 15, 2014 3:31:39 PM UTC-7, Miro Knej= p wrote:
On 16 Oct 2014, at 00:22 , Myriachan <myri...@gmail.com> wrote:
The main reason to use something like this= is for copy constructors.  Sometimes, the most convenient way to impl= ement a copy or move constructor is like this:

T &T::operator =3D(const T &copy){
    if (&copy !=3D this)
    {
        this<= span style=3D"color:#660">->~T();
        new(this) T(copy);
    }
    return *this;
}

...and equivalently with std::forward for move constructors.
=

This would never pass my code revie= w. If that constructor throws you're doomed. Use copy-and-swap instead, it = provides strong exception guarantee and won't leave you with a destroyed ob= ject. Move constructors are a different game though and the compiler-defaul= ted one should do it's job just fine in most cases.

T& T::= operator =3D(con= st T& copy= )
{
    if (&cop= y !=3D this)
    {
&= nbsp;       T x{co= py};
        swap(*this, x);
   = ; }
    = return *this;
}


I suppose the difference is that I work on a project that disabl= es C++ exceptions because they're too slow.  swap would be slow or unu= sable if there's either no overload for it nor move constructors, which for= a codebase of millions of lines written before C++11 is likely.

Ena= bling C++ exception handling is a massive performance decrease in x86-32 Wi= ndows.  If it's enabled in the compiler, even if no exceptions are use= d, every function with an automatic variable whose type has a destructor wi= ll be ~20 instructions longer due to setting up the exception frame.
Melissa

--

---
You received this message because you are subscribed to the Google Groups &= quot;ISO C++ Standard - Future Proposals" group.
To unsubscribe from this group and stop receiving emails from it, send an e= mail to std-proposa= ls+unsubscribe@isocpp.org.
To post to this group, send email to std-proposals@isocpp.org.
Visit this group at http://groups.google.com/a/isocpp.org/group/std-proposals/<= /a>.
------=_Part_123_1538419150.1413424980736-- .