From mboxrd@z Thu Jan 1 00:00:00 1970 Return-Path: Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S264940AbUELCXz (ORCPT ); Tue, 11 May 2004 22:23:55 -0400 Received: (majordomo@vger.kernel.org) by vger.kernel.org id S264941AbUELCXz (ORCPT ); Tue, 11 May 2004 22:23:55 -0400 Received: from mail.tpgi.com.au ([203.12.160.53]:22156 "EHLO mail5.tpgi.com.au") by vger.kernel.org with ESMTP id S264940AbUELCXy (ORCPT ); Tue, 11 May 2004 22:23:54 -0400 From: Nigel Cunningham Reply-To: ncunningham@linuxmail.org To: Linux Mailing List Subject: Preempted by idle thread? Date: Wed, 12 May 2004 12:23:26 +1000 User-Agent: KMail/1.5.3 MIME-Version: 1.0 Content-Type: text/plain; charset="us-ascii" Content-Transfer-Encoding: 7bit Content-Disposition: inline Message-Id: <200405121223.26115.ncunningham@linuxmail.org> X-TPG-Antivirus: Passed Sender: linux-kernel-owner@vger.kernel.org X-Mailing-List: linux-kernel@vger.kernel.org Hi all. I've been seeing some funny bad_scheduling_while_atomic messages that might be my own fault (they only occur after suspending). The funny thing, though, is that they are all the same, and show a calling chain: schedule_tail -> preempt_schedule -> cpu_idle Am I misunderstanding this, or is a thread being preempted by cpu_idle? If that is the case, real work shouldn't be preempted to idle, should it? Nigel