public inbox for gcc-help@gcc.gnu.org
 help / color / mirror / Atom feed
From: "LAMOME Julien CS-SI" <julien.lamome-cs-si@irsn.fr>
To: "Jonathan Wakely" <jwakely.gcc@gmail.com>
Cc: <gcc-help@gcc.gnu.org>
Subject: RE: [C++] anonymous struct linkage
Date: Fri, 23 Nov 2012 07:58:00 -0000	[thread overview]
Message-ID: <72041471EE7D1540B7ECD771EA430DB70622524E@AMESS102.proton.intra.irsn.fr> (raw)
In-Reply-To: <CAH6eHdQbUGBo-KE=cLOP=Ey=9znerHxUyaGKFPCojHyYyMiQng@mail.gmail.com>

a minimum code to show the problem :
toto.F :
      subroutine toto()
      implicit none
      real*8 variable
      common /demof/variable

         print*,'Fortran variable=',variable
      end

toto.cc: 
#include <iostream>
struct {double variable;}demof;
extern "C" void toto();
int main()
{
	demof.variable=3.14159;
	std::cout<<"C++  variable = "<<demof.variable<<std::endl;
	toto();
	return 0;
}

* Gcc 4.1 give me :
C++  variable = 3.14159
 Fortran variable= 3.14159000000000
* Gcc 4.6 give me:
C++  variable = 3.14159
 Fortran variable= 0.00000000000000

Yours differents solutions work, but I would prefer to use an option to have the old behavior. Because, I have 163 files with anonymous strut to modify. And, each file require special attention because struct are linked, or not, with a COMMON, and COMMON are not necessary include in source, etc.
So, if it isn't possible to use an option, I will have a lot of work now. Else I could modify  progressively.

-----Message d'origine-----
De : Jonathan Wakely [mailto:jwakely.gcc@gmail.com] 
Envoyé : jeudi 22 novembre 2012 22:07
À : LAMOME Julien CS-SI
Cc : gcc-help@gcc.gnu.org
Objet : Re: [C++] anonymous struct linkage

On 22 November 2012 21:03, Jonathan Wakely wrote:
> On 22 November 2012 15:44, LAMOME Julien CS-SI
> <julien.lamome-cs-si@irsn.fr> wrote:
>> Hi,
>> we have a code mixing C++ and Fortran 77. In fortran, we have some COMMON like :
>> COMMON / DEMOF/VARIABLE
>> Which are include in fortran source file like this :
>> #include <demof.include>
>>
>> We link with C++ code like this :
>> struct { double variable}demof ;
>>
>> The problem is :
>> Between gcc 4.3 and gcc 4.6, the linkage of demof (in C++) change from global to local. This breaks the link between C++ and fortran.
>>
>> How can I restore the global linkage in gcc 4.6 ?
>
> Give the type a name:
>
> struct D { double variable; } demof;
>
> Problem solved.

Alternatively, give it C language linkage:

extern "C" {
  struct { double d; } demof;
}

  reply	other threads:[~2012-11-23  7:58 UTC|newest]

Thread overview: 8+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2012-11-22 15:44 LAMOME Julien CS-SI
2012-11-22 15:49 ` Andrew Haley
2012-11-22 18:53 ` Ángel González
2012-11-22 21:03 ` Jonathan Wakely
2012-11-22 21:06   ` Jonathan Wakely
2012-11-23  7:58     ` LAMOME Julien CS-SI [this message]
2012-11-23  9:37       ` Jonathan Wakely
2012-11-26 15:49       ` Ángel González

Reply instructions:

You may reply publicly to this message via plain-text email
using any one of the following methods:

* Save the following mbox file, import it into your mail client,
  and reply-to-all from there: mbox

  Avoid top-posting and favor interleaved quoting:
  https://en.wikipedia.org/wiki/Posting_style#Interleaved_style

* Reply using the --to, --cc, and --in-reply-to
  switches of git-send-email(1):

  git send-email \
    --in-reply-to=72041471EE7D1540B7ECD771EA430DB70622524E@AMESS102.proton.intra.irsn.fr \
    --to=julien.lamome-cs-si@irsn.fr \
    --cc=gcc-help@gcc.gnu.org \
    --cc=jwakely.gcc@gmail.com \
    /path/to/YOUR_REPLY

  https://kernel.org/pub/software/scm/git/docs/git-send-email.html

* If your mail client supports setting the In-Reply-To header
  via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line before the message body.
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox;
as well as URLs for read-only IMAP folder(s) and NNTP newsgroup(s).