From owner-freebsd-current Thu May 14 04:19:06 1998 Return-Path: Received: (from majordom@localhost) by hub.freebsd.org (8.8.8/8.8.8) id EAA06016 for freebsd-current-outgoing; Thu, 14 May 1998 04:19:06 -0700 (PDT) (envelope-from owner-freebsd-current@FreeBSD.ORG) Received: from ns1.yes.no (ns1.yes.no [195.119.24.10]) by hub.freebsd.org (8.8.8/8.8.8) with ESMTP id EAA06004 for ; Thu, 14 May 1998 04:19:03 -0700 (PDT) (envelope-from eivind@bitbox.follo.net) Received: from bitbox.follo.net (bitbox.follo.net [194.198.43.36]) by ns1.yes.no (8.8.7/8.8.7) with ESMTP id LAA03883; Thu, 14 May 1998 11:18:57 GMT Received: (from eivind@localhost) by bitbox.follo.net (8.8.8/8.8.6) id NAA18972; Thu, 14 May 1998 13:18:56 +0200 (MET DST) Message-ID: <19980514131856.48680@follo.net> Date: Thu, 14 May 1998 13:18:56 +0200 From: Eivind Eklund To: Terry Lambert , Alexander Litvin Cc: current@FreeBSD.ORG Subject: Re: daemons exiting on signal 11 References: <19980513093110.21002@carrier.kiev.ua> <199805140302.UAA04302@usr08.primenet.com> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii X-Mailer: Mutt 0.89.1i In-Reply-To: <199805140302.UAA04302@usr08.primenet.com>; from Terry Lambert on Thu, May 14, 1998 at 03:02:08AM +0000 Sender: owner-freebsd-current@FreeBSD.ORG Precedence: bulk X-Loop: FreeBSD.ORG On Thu, May 14, 1998 at 03:02:08AM +0000, Terry Lambert wrote: > > I thought the problem has gone. But yesterday trying to test > > new softupdates patches, I loaded my machine with 'make -j64 > > buildworld', and at some point cron started to exit with SIGSEGV. > > After makeworld finished problem with cron didn't disappear, > > so I needed to restart cron :-\ > > For readers who might have been alarmed by this: > > Note that this is not indicative of a problem in the soft updates > code. This means that he was out of swap + RAM, and when cron > went to get a page, it was denied, and could not continue to run. The part about the it not having to do with soft updates is correct. The rest is slightly misleading - there is a problem in the VM code (has been there for a few weeks) which is triggered when you run out of memory and swap. After you've been out of memory, the new forks of daemons die with SIGSEGV (seemingly right after the fork - at least before they get anything done). This is not just while you're out of memory - it behaviour continues until you restart the daemons. Eivind. To Unsubscribe: send mail to majordomo@FreeBSD.org with "unsubscribe freebsd-current" in the body of the message