From owner-freebsd-security Wed Jan 27 07:41:31 1999 Return-Path: Received: (from majordom@localhost) by hub.freebsd.org (8.8.8/8.8.8) id HAA29256 for freebsd-security-outgoing; Wed, 27 Jan 1999 07:41:31 -0800 (PST) (envelope-from owner-freebsd-security@FreeBSD.ORG) Received: from trooper.velocet.ca (host-034.canadiantire.ca [209.146.201.34]) by hub.freebsd.org (8.8.8/8.8.8) with ESMTP id HAA29245 for ; Wed, 27 Jan 1999 07:41:28 -0800 (PST) (envelope-from dgilbert@trooper.velocet.ca) Received: (from dgilbert@localhost) by trooper.velocet.ca (8.8.7/8.8.7) id KAA20238; Wed, 27 Jan 1999 10:40:47 -0500 (EST) From: David Gilbert MIME-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit Message-ID: <13999.13182.831576.209183@trooper.velocet.ca> Date: Wed, 27 Jan 1999 10:40:46 -0500 (EST) To: Bruce Evans Cc: security@FreeBSD.ORG Subject: Re: signal handling in urandom can cause lockup In-Reply-To: <199901271235.XAA19607@godzilla.zeta.org.au> References: <199901271235.XAA19607@godzilla.zeta.org.au> X-Mailer: VM 6.62 under Emacs 19.34.2 Sender: owner-freebsd-security@FreeBSD.ORG Precedence: bulk X-Loop: FreeBSD.org >>>>> "Bruce" == Bruce Evans writes: Bruce> My fix blocks most i/o hog processes in uiomove() and Bruce> associated functions if they would be rescheduled if they were Bruce> running in user mode. Unfortunately, signals can't be handled Bruce> at this level since it would be surprising if disk i/o could be Bruce> aborted by a signal. My fix only checks for signals for Bruce> /dev/urandom. I don't know of any other devices that need it. This is very interesting. For some time (I use FreeBSD on my desktop as well as for 50 odd servers) my desktop machine has been exhibiting a freeze that seems to coinside with intense swapping or writing to the disk. This 'freeze' survived the upgrade from 2.2.7 to 3.0. I can trigger it by saving a very large file in emacs (my mail buffer is usually sufficient) and I can experience often when a lot of swapping is going on. The machine has 128M of memory and 1G of swap across 2 IDE drives. The system is used by myself and 4 xterms (often all running netscape), so the swap can at times be quite busy. Now... my take on the situation had been that IDE is just evil. I have burned out 3 drives (of different brands) in as many months (they ranged from 1 month to 1 year old). I have on order a 9G SCSI drive to become the main system and swap drive under the assumption that scsi drives are going to be more dependable and will not freeze the system. This is not entirely without precident. I switched my CDROM to be SCSI a couple of months ago. Before the change, I could get 10+ second freezes while the cdrom attempted to read a bad CD. Now everything flows 100% smoothly, bad CD or not. However, it does strike me as bogus that this kind of blocking happens. Maybe this type of patch should be applied to devices in general. Dave. -- ============================================================================ |David Gilbert, Velocet Communications. | Two things can only be | |Mail: dgilbert@velocet.net | equal if and only if they | |http://www.velocet.net/~dgilbert | are precisely opposite. | =========================================================GLO================ To Unsubscribe: send mail to majordomo@FreeBSD.org with "unsubscribe freebsd-security" in the body of the message