220 14741 <5124499.Y4MVqN9qgV@tjmaciei-mobl4> article Path: news.gmane.org!not-for-mail From: Thiago Macieira Newsgroups: gmane.comp.lang.c++.isocpp.proposals Subject: Re: Allow const-qualified "argv" Date: Tue, 25 Nov 2014 19:31:29 -0800 Lines: 82 Approved: news@gmane.org Message-ID: <5124499.Y4MVqN9qgV@tjmaciei-mobl4> References: <5749465.k8kpGYc6nD@tjmaciei-mobl4> Reply-To: std-proposals@isocpp.org NNTP-Posting-Host: plane.gmane.org Mime-Version: 1.0 Content-Type: text/plain; charset=UTF-8 X-Trace: ger.gmane.org 1416972699 30714 80.91.229.3 (26 Nov 2014 03:31:39 GMT) X-Complaints-To: usenet@ger.gmane.org NNTP-Posting-Date: Wed, 26 Nov 2014 03:31:39 +0000 (UTC) To: std-proposals@isocpp.org Original-X-From: std-proposals+bncBCB4TK757YBRBFMT2WRQKGQEQP7XW4Q@isocpp.org Wed Nov 26 04:31:34 2014 Return-path: Envelope-to: gclcip-std-proposals@m.gmane.org Original-Received: from mail-ee0-f69.google.com ([74.125.83.69]) by plane.gmane.org with esmtp (Exim 4.69) (envelope-from ) id 1XtTKH-0001w6-UQ for gclcip-std-proposals@m.gmane.org; Wed, 26 Nov 2014 04:31:34 +0100 Original-Received: by mail-ee0-f69.google.com with SMTP id d49sf1403125eek.4 for ; Tue, 25 Nov 2014 19:31:33 -0800 (PST) X-Google-DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=1e100.net; s=20130820; h=x-gm-message-state:from:to:subject:date:message-id:user-agent :in-reply-to:references:mime-version:content-type:x-original-sender :x-original-authentication-results:reply-to:precedence:mailing-list :list-id:list-post:list-help:list-archive:list-subscribe :list-unsubscribe; bh=UU8A+E/QT6Td66AxlyI+gTY8EZrTRYwRwpzwiUieZ78=; b=M1+OPTugJ99FZj9hqACOd2VC5ZiTXuNdyzmOL3zJpDXjO7wem7f1on+OGRneHsFpWv FpolJWljGgQhXf0KmNXMZtKWEIciUw77wYijv/eeYY1BDTqF+yVOhgIWlzQVuVSdwMy7 AgCnTrtybHjx07RlycNBJHDCLgF7RQqbqzDZGO/hS9j2QvVEyAHUJyBXzwfLGgSEUNI6 VOgKb3bxltGSNl1a4BCGpmYliZv996DsoBqs8B6xdTWLKWAwABTys/ZL0uVdfo2mX2Lc UsWMBItE6JoqhYI5dJnMctsqpfjJDsjAu/bfsjAXXSDYM1djQy3hxuGcClGjSiPEN2R1 5hBg== X-Gm-Message-State: ALoCoQmA35mFB8QXX1ftsdEoUjiQvUdnfW8L7nPY/z4EgcrC3cN5xCw02BSS39kcyviLcpBrovWN X-Received: by 10.112.154.194 with SMTP id vq2mr2993836lbb.10.1416972693637; Tue, 25 Nov 2014 19:31:33 -0800 (PST) X-BeenThere: std-proposals@isocpp.org Original-Received: by 10.180.83.133 with SMTP id q5ls710668wiy.31.canary; Tue, 25 Nov 2014 19:31:32 -0800 (PST) X-Received: by 10.180.81.102 with SMTP id z6mr37018471wix.69.1416972692830; Tue, 25 Nov 2014 19:31:32 -0800 (PST) Original-Received: from gondolin.macieira.info (gondolin.macieira.info. [78.47.120.188]) by mx.google.com with ESMTP id gd8si20110265wib.41.2014.11.25.19.31.32 for ; Tue, 25 Nov 2014 19:31:32 -0800 (PST) Received-SPF: pass (google.com: domain of thiago@macieira.org designates 78.47.120.188 as permitted sender) client-ip=78.47.120.188; Original-Received: from tjmaciei-mobl4.localnet (unknown [192.55.54.40]) by gondolin.macieira.info (Postfix) with ESMTPSA id 0E3B011B4B0 for ; Tue, 25 Nov 2014 19:31:32 -0800 (PST) User-Agent: KMail/4.14.2 (Linux/3.11.10-21-desktop; KDE/4.14.2; x86_64; ; ) In-Reply-To: X-Original-Sender: thiago@macieira.org X-Original-Authentication-Results: mx.google.com; spf=pass (google.com: domain of thiago@macieira.org designates 78.47.120.188 as permitted sender) smtp.mail=thiago@macieira.org 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:14741 Archived-At: On Tuesday 25 November 2014 12:36:35 Myriachan wrote: > Visual Studio already has a wmain(), which is why I pointed this out. Note > that in Windows, main() is a compiler concept, not a system one; this is > comparable to "start" of crt0.o on the major POSIX systems. > > extern "C" int __cdecl wmain(); > extern "C" int __cdecl wmain(int argc, wchar_t **argv); > extern "C" int __cdecl wmain(int argc, wchar_t **argv, wchar_t **envp); Actually, it's a regular function. The real entry point is WinMain and the runtime needs to split the arguments into argv before proceeding. > > > Actually, the only way this would work is if the compiler emitted an > > external > > symbol called "main" whenever it saw one of those four functions. > > > Unfortunately, it doesn't solve this problem: > How Microsoft's compiler works is that the compiler emits a linker > directive into the .obj file if it sees a function named ::wmain defined. > Specifically, it emits /ENTRY:wmainCRTStartup to change the entry point of > the module to wmainCRTStartup(). wmainCRTStartup() is the "start" > equivalent in Windows. mainCRTStartup(), the default, calls main(), and > wmainCRTStartup() calls wmain(). mainCRTStartup() handles any conversion > if required. If c16main() or c32main() existed, they likely would be based > upon a similar model. wmainCRTStartup would likely use a hack to call > c16main so as to avoid a 4th copy of the startup code. Microsoft's solution would not work for everyone. It implies having control over the linker, which is not always the case for C++ compiler writers. Hence what I said that the solution is to always have a "main", regardless of what the user typed. > > And all along it would allocate more memory for the conversions (well, > > the > > > > > UCS-4 to UTF-16 conversion can be done in-place). > > > > This still requires allocating memory, which may fail. And it requires > > knowing > > what the system locale is in order to convert from 8-bit, before the first > > line > > of user code... > > Systems generally have code before either main() or the global constructors > anyway. It's already the case that in Windows, using main() instead of > wmain() loses information from the command line due to the system code > page/locale. Windows has to allocate memory for the argv buffer > regardless, because Windows has command lines as a singe string, and the C > runtime library needs to split it up into argv format. > > POSIX systems have less to worry about. They have the locale to worry about in your proposal. It might imply having to load a plugin for iconv() before main() gets run in order to do the conversion. This is not impossible, just way too complex for the benefit. > Perhaps wmain(), c16main() and c32main() could be optional but recommended > in the Standard? In other words, specifying (figuratively), "if these > entry points do exist, they function as the following...". Then the standard also needs to define a macro that the compiler will set when it does support those alternate entry points, so code can properly decide whether to have a c16main() or a plain main(). -- Thiago Macieira - thiago (AT) macieira.info - thiago (AT) kde.org Software Architect - Intel Open Source Technology Center PGP/GPG: 0x6EF45358; fingerprint: E067 918B B660 DBD1 105C 966C 33F5 F005 6EF4 5358 -- --- 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/. .