From owner-freebsd-threads@FreeBSD.ORG Wed Nov 3 04:36:53 2004 Return-Path: Delivered-To: freebsd-threads@freebsd.org Received: from mx1.FreeBSD.org (mx1.freebsd.org [216.136.204.125]) by hub.freebsd.org (Postfix) with ESMTP id 65B7616A4CE for ; Wed, 3 Nov 2004 04:36:53 +0000 (GMT) Received: from hellmouth3.gatech.edu (hellmouth3.gatech.edu [130.207.165.163]) by mx1.FreeBSD.org (Postfix) with ESMTP id ECB9243D39 for ; Wed, 3 Nov 2004 04:36:52 +0000 (GMT) (envelope-from gte990t@mail.gatech.edu) Received: from hellmouth3.gatech.edu (localhost [127.0.0.1]) by hellmouth3.gatech.edu (Postfix) with SMTP id 86E4E220ECE for ; Tue, 2 Nov 2004 23:36:52 -0500 (EST) (envelope-from gte990t@mail.gatech.edu) Received: from mailprx4.gatech.edu (mailprx4.prism.gatech.edu [130.207.171.18]) (using TLSv1 with cipher EDH-RSA-DES-CBC3-SHA (168/168 bits)) (verified OK)) by hellmouth3.gatech.edu (Postfix) with ESMTP id 71B19220E8D for ; Tue, 2 Nov 2004 23:36:52 -0500 (EST) (envelope-from gte990t@mail.gatech.edu) Received: from [192.168.0.3] (r58h96.res.gatech.edu [128.61.58.96]) (using TLSv1 with cipher RC4-MD5 (128/128 bits)) (No client certificate requested) (sasl: method=PLAIN, username=gte990t, sender=n/a) by mailprx4.gatech.edu (Postfix) with ESMTP id BB66A3A5EE for ; Tue, 2 Nov 2004 23:36:50 -0500 (EST) (envelope-from gte990t@mail.gatech.edu) From: Jason Harmening Date: Tue, 2 Nov 2004 23:56:06 -0500 User-Agent: KMail/1.7 To: freebsd-threads@freebsd.org MIME-Version: 1.0 Content-Type: text/plain; charset="iso-8859-1" Content-Transfer-Encoding: 7bit Content-Disposition: inline Message-Id: <200411022356.06257.gte990t@mail.gatech.edu> Subject: [FreeBSD 5.3-RC2] Processes STILL hanging in unkillable state X-BeenThere: freebsd-threads@freebsd.org X-Mailman-Version: 2.1.1 Precedence: list List-Id: Threading on FreeBSD List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , X-List-Received-Date: Wed, 03 Nov 2004 04:36:53 -0000 The following message has already been posted to the freebsd-current mailing list. Please note that the unkillable process problem originally reported in 5.3-RC1 is NOT isolated to gdb/kde, and seems to occur with OpenOffice under heavy loads, especially with java running at the same time. Furthermore, 5.3-RC2 does not fix the problem. Several other people have chimed in saying they have the same problem, and I think it would be a BAD idea to release 5.3 without fixing this. I don't have the kernel expertise to directly delve into the problem, but I do have a reasonable amount of systems-level development/troubleshooting experience, so I would be more than willing to help with debugging/diagnosis if someone could point me in the right direction. Thanks, Jason Harmening ---------- Forwarded Message ---------- Subject: Re: [FreeBSD 5.3-RC2] Processes STILL hanging in unkillable state Date: Tuesday 02 November 2004 11:45 From: Jason Harmening To: Alex Dupre Cc: freebsd-current@freebsd.org On Tuesday 02 November 2004 09:00, you wrote: > Jason Harmening wrote: > > I just upgraded to 5.3-RC2, and I'm still running into the problem where > > processes will hang in an unkillable state. In particular, this happens > > for me with OpenOffice under a heavy load. 'ps' reports the state as > > 'TL' and 'top' reports the state as STOP. Neither kill -CONT nor kill > > -KILL will work. > > Same here. With OpenOffice and Java under high load. Yes--I forgot to mention it, but my OpenOffice problems have all seemed to happen at the same time I have some instance of the JVM (native 1.4.2) running. OpenOffice will start up, reach almost the end of its splash screen, and then simply hang in an unkillable state. This has a very sinister side effect in that, when the system does its daily security run (I'm also assuming it tries to clean up useless processes), and there's such an "unkillable" process, the entire system will hang. It's not an abrupt freeze, but it's a gradual "bogging down" that ultimately requires a hard reset. The same thing happens during a shutdown/reboot with an unkillable process, but luckily the system manages to sync the disk before hanging. -------------------------------------------------------