From mboxrd@z Thu Jan 1 00:00:00 1970 Return-Path: Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S266506AbUBLW6y (ORCPT ); Thu, 12 Feb 2004 17:58:54 -0500 Received: (majordomo@vger.kernel.org) by vger.kernel.org id S266656AbUBLW6y (ORCPT ); Thu, 12 Feb 2004 17:58:54 -0500 Received: from mail-07.iinet.net.au ([203.59.3.39]:57531 "HELO mail.iinet.net.au") by vger.kernel.org with SMTP id S266506AbUBLW6x (ORCPT ); Thu, 12 Feb 2004 17:58:53 -0500 Message-ID: <402C050B.2040803@cyberone.com.au> Date: Fri, 13 Feb 2004 09:58:19 +1100 From: Nick Piggin User-Agent: Mozilla/5.0 (X11; U; Linux i686; en-US; rv:1.6) Gecko/20040122 Debian/1.6-1 X-Accept-Language: en MIME-Version: 1.0 To: Rudo Thomas CC: linux-kernel@vger.kernel.org Subject: Re: bug, or is it? - SCHED_RR and FPU related References: <20040212205708.GA1679@ss1000.ms.mff.cuni.cz> In-Reply-To: <20040212205708.GA1679@ss1000.ms.mff.cuni.cz> Content-Type: text/plain; charset=us-ascii; format=flowed Content-Transfer-Encoding: 7bit Sender: linux-kernel-owner@vger.kernel.org X-Mailing-List: linux-kernel@vger.kernel.org Rudo Thomas wrote: >Hello. > >I have stumbled upon a grave bug, I think. I can reproduce a hard lock-up using >xmms and a buggy plugin - only SysRq-B helps. I managed to narrow it down to >"0.5/NAN"-like operation. It only causes a complete hang when xmms is run with >SCHED_RR priority. Otherwise, only xmms hangs. Unfortunately, I was not able to >create a working proof-of-concept program. > >The question is - is this a kernel bug, OR something that cannot be prevented >from happening when using SCHED_RR policy? > >Tried with 2.6.3-rc[12] and glibc-2.3.2 (-r9 gentoo) with nptl support (if that >matters). > >Have a nice day. > > Hi Rudo, xmms probably goes into an infinite loop, preventing anything else from being scheduled, right? If that's the case then this is correct behaviour.