From owner-freebsd-security Sat Nov 23 22:50:12 2002 Delivered-To: freebsd-security@freebsd.org Received: from mx1.FreeBSD.org (mx1.freebsd.org [216.136.204.125]) by hub.freebsd.org (Postfix) with ESMTP id 0DF4F37B404 for ; Sat, 23 Nov 2002 22:50:10 -0800 (PST) Received: from HAL9000.homeunix.com (12-232-220-15.client.attbi.com [12.232.220.15]) by mx1.FreeBSD.org (Postfix) with ESMTP id 5165F43E9C for ; Sat, 23 Nov 2002 22:50:09 -0800 (PST) (envelope-from dschultz@uclink.Berkeley.EDU) Received: from HAL9000.homeunix.com (localhost [127.0.0.1]) by HAL9000.homeunix.com (8.12.6/8.12.5) with ESMTP id gAO6o3Uf002769; Sat, 23 Nov 2002 22:50:03 -0800 (PST) (envelope-from dschultz@uclink.Berkeley.EDU) Received: (from das@localhost) by HAL9000.homeunix.com (8.12.6/8.12.5/Submit) id gAO6o1kE002760; Sat, 23 Nov 2002 22:50:01 -0800 (PST) (envelope-from dschultz@uclink.Berkeley.EDU) Date: Sat, 23 Nov 2002 22:50:01 -0800 From: David Schultz To: Sheldon Hearn Cc: Mike Silbersack , "David G. Andersen" , freebsd-security@FreeBSD.ORG Subject: Re: File table exhaustion patch Message-ID: <20021124065001.GA2683@HAL9000.homeunix.com> Mail-Followup-To: Sheldon Hearn , Mike Silbersack , "David G. Andersen" , freebsd-security@FreeBSD.ORG References: <20021121105204.B75421@cs.utah.edu> <20021121152539.U44884-100000@patrocles.silby.com> <20021122080515.GQ36738@starjuice.net> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Disposition: inline In-Reply-To: <20021122080515.GQ36738@starjuice.net> Sender: owner-freebsd-security@FreeBSD.ORG Precedence: bulk List-ID: List-Archive: (Web Archive) List-Help: (List Instructions) List-Subscribe: List-Unsubscribe: X-Loop: FreeBSD.org Thus spake Sheldon Hearn : > On (2002/11/21 15:29), Mike Silbersack wrote: > > > HOWEVER, we're in a code freeze leading up to 5.0-release, and local DoSes > > aren't a critical bug. > > Is that the official FreeBSD SO team viewpoint on local DoS > vulnerabilities? DoS attacks are incredibly hard to address in general, and I have yet to see a multiuser system that isn't vulnerable to at least several of them. Given that FreeBSD has always been ``vulnerable'' to file table exhaustion, waiting a few weeks isn't going to be the end of the world[1]. My favorite example of a local DoS attack is: while (1) mkdir t && cd t I ``discovered'' this one about a year ago, then found that Dennis Ritchie had pointed it out in the early 1970's. It reliably crashes most systems, often causing massive filesystem corruption. Until someone fixes the scores of known DoS attacks that already exist, I'm not willing to consider any particular attack to be high-priority. [1] These days, the size limit on the file table is administrative anyway, since the table is a hash table. Of course, it doesn't auto-resize if you grow it by an order of magnitude at runtime. To Unsubscribe: send mail to majordomo@FreeBSD.org with "unsubscribe freebsd-security" in the body of the message