From mboxrd@z Thu Jan 1 00:00:00 1970 From: Paolo Bonzini Subject: Re: [PATCH 5/5 V2] kvm tools: Initialize and use VESA and VNC Date: Wed, 25 May 2011 17:37:50 +0200 Message-ID: <4DDD224E.8070608@redhat.com> References: <20110524085024.GA31453@elte.hu> <4DDB77E5.4080306@redhat.com> <20110524194019.GA27634@elte.hu> <4DDCBC08.3030507@redhat.com> <20110525083200.GF21552@elte.hu> <4DDCC8A0.8050708@redhat.com> <20110525093611.GD28500@elte.hu> <4DDCD364.9090903@redhat.com> <20110525101744.GA30983@elte.hu> <4DDCDD84.8010800@redhat.com> <20110525125335.GD29300@elte.hu> Mime-Version: 1.0 Content-Type: text/plain; charset=ISO-8859-1; format=flowed Content-Transfer-Encoding: 7bit Cc: Sasha Levin , penberg@kernel.org, john@jfloren.net, kvm@vger.kernel.org, asias.hejun@gmail.com, gorcunov@gmail.com, prasadjoshi124@gmail.com To: Ingo Molnar Return-path: Received: from mx1.redhat.com ([209.132.183.28]:9724 "EHLO mx1.redhat.com" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1758013Ab1EYPiI (ORCPT ); Wed, 25 May 2011 11:38:08 -0400 In-Reply-To: <20110525125335.GD29300@elte.hu> Sender: kvm-owner@vger.kernel.org List-ID: On 05/25/2011 02:53 PM, Ingo Molnar wrote: > Obviously the compiler did not provide this feature into a vacuum, it > expected the C library to execute constructors, right? > > So the above dodges my question of why there is no method (in glibc) > to turn auto-execution off and let the app do it. Well, I wasn't there. :) I suppose they simply chose those semantics for C++ constructors which were easiest to implement (i.e. execute before main). ((constructor)) then followed suit. Any further extension in the direction you suggest would have to be designed very carefully in order to coexist with C++ constructors, or it would have to be something completely separate... in the latter case, ((constructor)) already provides the ingredients to do this, even if at a small cost in efficiency. So there is no big incentive to add this further extension to the compiler and/or the C library. Paolo