From mboxrd@z Thu Jan  1 00:00:00 1970
Received: (from majordomo@localhost) by pauillac.inria.fr (8.7.6/8.7.3) id QAA10365; Wed, 31 Oct 2001 16:20:02 +0100 (MET)
Received: from concorde.inria.fr (concorde.inria.fr [192.93.2.39]) by pauillac.inria.fr (8.7.6/8.7.3) with ESMTP id QAA10325 for <caml-list@pauillac.inria.fr>; Wed, 31 Oct 2001 16:20:01 +0100 (MET)
Received: from bastion.artisan.com ([209.144.161.130])
	by concorde.inria.fr (8.11.1/8.10.0) with ESMTP id f9VFJx128850
	for <caml-list@inria.fr>; Wed, 31 Oct 2001 16:20:00 +0100 (MET)
Received: from ypmaster.artisan.com (ypmaster [172.16.2.1])
	by bastion.artisan.com (8.9.2/8.9.2) with ESMTP id HAA19969;
	Wed, 31 Oct 2001 07:20:21 -0800 (PST)
Received: from brainiac.artisan.com (brainiac [172.16.10.88])
	by ypmaster.artisan.com (8.9.2/8.9.2) with ESMTP id HAA06967;
	Wed, 31 Oct 2001 07:19:34 -0800 (PST)
Received: (from johnm@localhost)
	by brainiac.artisan.com (8.9.2/8.9.2) id HAA03056;
	Wed, 31 Oct 2001 07:19:34 -0800 (PST)
X-Authentication-Warning: brainiac.artisan.com: johnm set sender to johnm@artisan.com using -f
From: John Gerard Malecki <johnm@artisan.com>
MIME-Version: 1.0
Content-Type: text/plain; charset=us-ascii
Content-Transfer-Encoding: 7bit
Message-ID: <15328.5766.464942.361033@brainiac.artisan.com>
Date: Wed, 31 Oct 2001 07:19:34 -0800
To: Dmitry Bely <dbely@mail.ru>
Cc: caml-list@inria.fr
Subject: Re: [Caml-list] Camlidl and order of include files
In-Reply-To: <elnkcduj.fsf@mail.ru>
References: <15327.13008.977353.705352@brainiac.artisan.com>
	<elnkcduj.fsf@mail.ru>
X-Mailer: VM 6.92 under Emacs 20.4.1
X-Organization: Artisan Components, Inc.
Sender: owner-caml-list@pauillac.inria.fr
Precedence: bulk

Hi Dmitry,

I am satisfied with simply using 

  quote(h,"#include <unistd.h>")

Canlidl does not need to know the value of _SC_NPROCESSORS_ONLN only
that the type is (something like) an enum.

No need for a camlidl bug report but the manual could benefit from an
index of the keywords.

Thanks for you help.

Dmitry Bely wrote (2001-10-31T13:18:12+0300):
 > John Gerard Malecki <johnm@artisan.com> writes:
 > 
 > > I just tried camlidl and am not sure if I am using it wrong or if I
 > > have encountered a bug.  I want to access the (solaris) sysconf system
 > > routine to find out how many processors on a machine.  I wrote this
 > > idl file
 > > 
 > >   quote(C,"#include <unistd.h>")
 > >   
 > >   enum sysconf_names {
 > >     SC_NPROCESSOR_ONLN = _SC_NPROCESSORS_ONLN
 > >   };
 > 
 > 
 > Use quote(h,"#include <unistd.h>")
 > 
 > and then C stubs will compile, but your code still will be incorrect. How
 > the generated Caml part will know the value of _SC_NPROCESSORS_ONLN?
 > Obviously, it is not defined for Camlidl. It should generate an error
 > message here. It does not, and that's the bug in Camlidl (you can submit a
 > bug report). As for correct solution, you should either convert <unistd.h>
 > to IDL and then use
 > 
 > #include unistd.idl
 >    
 > enum sysconf_names {
 >    SC_NPROCESSOR_ONLN = _SC_NPROCESSORS_ONLN
 > };
 > 
 > or resolve _SC_NPROCESSORS_ONLN macro manually:
 > 
 > quote(C,"#include <unistd.h>")
 > 
 > enum sysconf_names {
 >    SC_NPROCESSOR_ONLN = 9
 > };
 > 
 > Hope to hear from you soon,
 > Dmitry
-------------------
Bug reports: http://caml.inria.fr/bin/caml-bugs  FAQ: http://caml.inria.fr/FAQ/
To unsubscribe, mail caml-list-request@inria.fr  Archives: http://caml.inria.fr