From owner-freebsd-ports@FreeBSD.ORG Mon Jan 23 06:19:58 2006 Return-Path: X-Original-To: freebsd-ports@freebsd.org Delivered-To: freebsd-ports@freebsd.org Received: from mx1.FreeBSD.org (mx1.freebsd.org [216.136.204.125]) by hub.freebsd.org (Postfix) with ESMTP id E178316A41F for ; Mon, 23 Jan 2006 06:19:58 +0000 (GMT) (envelope-from eggert@CS.UCLA.EDU) Received: from kiwi.cs.ucla.edu (Kiwi.CS.UCLA.EDU [131.179.128.19]) by mx1.FreeBSD.org (Postfix) with ESMTP id 7661043D46 for ; Mon, 23 Jan 2006 06:19:58 +0000 (GMT) (envelope-from eggert@CS.UCLA.EDU) Received: from penguin.cs.ucla.edu (Penguin.CS.UCLA.EDU [131.179.64.200]) by kiwi.cs.ucla.edu (8.11.7p1+Sun/8.11.7/UCLACS-5.2) with ESMTP id k0N6JYH09310; Sun, 22 Jan 2006 22:19:34 -0800 (PST) Received: from eggert by penguin.cs.ucla.edu with local (Exim 4.50) id 1F0v38-0005ro-7w; Sun, 22 Jan 2006 22:19:34 -0800 To: Kris Kennaway References: <200601131825.SAA21164@sopwith.solgatos.com> <87r772obn8.fsf@penguin.cs.ucla.edu> <20060120211940.GA43031@xor.obsecurity.org> <87u0bymq7x.fsf@penguin.cs.ucla.edu> <20060122044224.GA81690@xor.obsecurity.org> From: Paul Eggert Date: Sun, 22 Jan 2006 22:19:34 -0800 In-Reply-To: <20060122044224.GA81690@xor.obsecurity.org> (Kris Kennaway's message of "Sat, 21 Jan 2006 23:42:24 -0500") Message-ID: <87lkx7zcsp.fsf@penguin.cs.ucla.edu> User-Agent: Gnus/5.1007 (Gnus v5.10.7) Emacs/21.4 (gnu/linux) MIME-Version: 1.0 Content-Type: text/plain; charset=us-ascii Cc: freebsd@sopwith.solgatos.com, bug-bison@gnu.org, freebsd-ports@freebsd.org Subject: Re: portability fix for bison-1.75 X-BeenThere: freebsd-ports@freebsd.org X-Mailman-Version: 2.1.5 Precedence: list List-Id: Porting software to FreeBSD List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , X-List-Received-Date: Mon, 23 Jan 2006 06:19:59 -0000 Kris Kennaway writes: > there are still two failures: > > lang/pike70 (ftp://ftp.oav.net/pike/7.0/) > > security/bro (ftp://ftp.ee.lbl.gov/) Thanks for reporting these. It's the first I've heard of either problem. > Perhaps these are the fault of the software Yes, in both cases. I'll follow up separately for each one. Briefly, Pike is doing a "#define short int", which violates section 7.1.3 of the C Standard, and bro is doing a "#define YYLLOC_DEFAULT(Current, Rhs, N) Current = Rhs[N];", which doesn't properly parenthesize the right-hand side of the macro. I'll check whether the bugs are still in the latest versions of these programs, and if so I'll send a bug report to the upstream developers. Also, I'll install workarounds for this into Bison (by adding an "#undef short" to defeat Pike's "#define", and by overparenthesizing the args passed to YYLLOC_DEFAULT), so Bison 2.2 (when it comes out) should work even with these ancient and buggy distributions. However.... > but until they're both fixed (either by a patch to the software or a > fix to bison) As far as I can tell, neither Pike 7.0 nor bro 0.8 is currently being maintained by the upstream developer, so what procedure would you prefer here, to make Bison 2.1 acceptable to you? Should I send you a patch for these programs personally? Here are my sources for saying the old program versions aren't being maintained. says that Pike 7.0 is no longer being updated. Presumably people should be using 7.6.6, the current official release, or 7.4.44, the previous (and still-maintained) release. says that the port is using , but that URL no longer exists. says that no further development is being done to the 0.8 release.