From mboxrd@z Thu Jan 1 00:00:00 1970 Return-Path: Delivered-To: caml-list@yquem.inria.fr Received: from concorde.inria.fr (concorde.inria.fr [192.93.2.39]) by yquem.inria.fr (Postfix) with ESMTP id AE884BBFB for ; Sun, 19 Jun 2005 12:01:02 +0200 (CEST) Received: from pauillac.inria.fr (pauillac.inria.fr [128.93.11.35]) by concorde.inria.fr (8.13.0/8.13.0) with ESMTP id j5JA12fu000999 for ; Sun, 19 Jun 2005 12:01:02 +0200 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 MAA27879 for ; Sun, 19 Jun 2005 12:01:01 +0200 (MET DST) Received: from yquem.inria.fr (yquem.inria.fr [128.93.8.37]) by concorde.inria.fr (8.13.0/8.13.0) with ESMTP id j5JA0xO9000996; Sun, 19 Jun 2005 12:00:59 +0200 Received: by yquem.inria.fr (Postfix, from userid 18180) id 90961BBFB; Sun, 19 Jun 2005 12:00:59 +0200 (CEST) Date: Sun, 19 Jun 2005 12:00:59 +0200 From: Xavier Leroy To: Aleksey Nogin Cc: Caml List Subject: Re: [Caml-list] OCaml 3.08.3 fails to compile under Fedora Core 4 (GCC 4). Message-ID: <20050619100059.GA20968@yquem.inria.fr> References: <42B4E6C5.2010405@cs.caltech.edu> <20050619074551.GA623@yquem.inria.fr> <42B529CD.6070106@cs.caltech.edu> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Disposition: inline In-Reply-To: <42B529CD.6070106@cs.caltech.edu> User-Agent: Mutt/1.3.28i X-Miltered: at concorde with ID 42B5425E.000 by Joe's j-chkmail (http://j-chkmail.ensmp.fr)! X-Miltered: at concorde with ID 42B5425B.001 by Joe's j-chkmail (http://j-chkmail.ensmp.fr)! X-Spam: no; 0.00; caml-list:01 ocaml:01 gcc:01 ocaml:01 gcc:01 bug:01 runtime:01 otherlibs:01 compile:01 workaround:01 inline:01 asm:02 gnu:03 seems:03 suggests:04 X-Spam-Checker-Version: SpamAssassin 3.0.2 (2004-11-16) on yquem.inria.fr X-Spam-Status: No, score=0.0 required=5.0 tests=none autolearn=disabled version=3.0.2 X-Spam-Level: > I've realized that in Fedora Extras 4 the ocaml package contains the > following patch: > - : "+&r" (a), "+&r" (b), "+&rm" (blen), "+&r" (out) > + : "+&r" (a), "+&r" (b), "+m" (blen), "+&r" (out) > Is this reasonable? Yes, very much so. You'll get better performance than with the workaround I suggested (disabling inline asm). I need to run a few tests with gcc 4, but eventually will fix the problem along these lines. > Also, > http://gcc.gnu.org/bugzilla/show_bug.cgi?id=21291 suggests using -O0 > (which seems to work as well). That wil be costly in performance, especially if -O0 is applied to the whole OCaml runtime system (not just otherlibs/num). - Xavier Leroy