From mboxrd@z Thu Jan 1 00:00:00 1970 Return-Path: Received: (qmail 7076 invoked by alias); 31 Aug 2004 01:06:09 -0000 Mailing-List: contact gcc-help@gcc.gnu.org; run by ezmlm Precedence: bulk List-Archive: List-Post: List-Help: Sender: gcc-owner@gcc.gnu.org Received: (qmail 7056 invoked from network); 31 Aug 2004 01:06:08 -0000 Received: from unknown (HELO vlsi1.ultra.nyu.edu) (128.122.140.213) by sourceware.org with SMTP; 31 Aug 2004 01:06:08 -0000 Received: by vlsi1.ultra.nyu.edu (4.1/1.34) id AA01650; Mon, 30 Aug 04 21:08:59 EDT Date: Tue, 31 Aug 2004 01:25:00 -0000 From: kenner@vlsi1.ultra.nyu.edu (Richard Kenner) Message-Id: <10408310108.AA01650@vlsi1.ultra.nyu.edu> To: zack@codesourcery.com Subject: Re: Ada policy Cc: gcc@gcc.gnu.org X-SW-Source: 2004-08/txt/msg01558.txt.bz2 Note that that was in the context of a discussion of whether configure should refuse to enable Ada if the "wrong" bootstrap compiler was found. You missed the point that "wrong" meant one of the compilers where the you'd end up building the C and Ada parts with different versions of GCC. That caused EH problems within the compiler. But that doesn't mean you can't use pragma Unreferenced; it just has to be properly conditionalized so that GCC 3.(<3) somehow sees pragma Warnings (Off) instead. You can't "conditionalize" in Ada, "properly" or not. It was a requirement of the language that it *not* permit conditionalization. The reason was to avoid situations where you'd get compilation errors with one set of options and not with others. The requirements for the language that became Ada forbid that. There's simply no way to avoid the sort of problem you are trying to work around.