From mboxrd@z Thu Jan 1 00:00:00 1970 Return-Path: Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S263861AbUDQLs6 (ORCPT ); Sat, 17 Apr 2004 07:48:58 -0400 Received: (majordomo@vger.kernel.org) by vger.kernel.org id S263926AbUDQLs6 (ORCPT ); Sat, 17 Apr 2004 07:48:58 -0400 Received: from caramon.arm.linux.org.uk ([212.18.232.186]:44304 "EHLO caramon.arm.linux.org.uk") by vger.kernel.org with ESMTP id S263861AbUDQLs4 (ORCPT ); Sat, 17 Apr 2004 07:48:56 -0400 Date: Sat, 17 Apr 2004 12:48:50 +0100 From: Russell King To: Herbert Xu Cc: Rolf Kutz , 244207@bugs.debian.org, Andrew Morton , Linux Kernel Mailing List Subject: Re: Bug#244207: kernel-source-2.6.5: mwave gives warning on suspend Message-ID: <20040417124850.C14786@flint.arm.linux.org.uk> Mail-Followup-To: Herbert Xu , Rolf Kutz , 244207@bugs.debian.org, Andrew Morton , Linux Kernel Mailing List References: <20040417104311.9C13A1D802@jamaika.kutz.dyndns.org> <20040417113918.GA4846@gondor.apana.org.au> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Disposition: inline User-Agent: Mutt/1.2.5.1i In-Reply-To: <20040417113918.GA4846@gondor.apana.org.au>; from herbert@gondor.apana.org.au on Sat, Apr 17, 2004 at 09:39:18PM +1000 Sender: linux-kernel-owner@vger.kernel.org X-Mailing-List: linux-kernel@vger.kernel.org On Sat, Apr 17, 2004 at 09:39:18PM +1000, Herbert Xu wrote: > On Sat, Apr 17, 2004 at 12:43:11PM +0200, Rolf Kutz wrote: > > Package: kernel-source-2.6.5 > > Version: 2.6.5-1 > > Severity: normal > > > > The mwave module gives the following warning on suspend: > > > > Apr 16 09:55:13 localhost kernel: Device 'mwave' does not have a release() funct > > ion, it is broken and must be fixed. > > Apr 16 09:55:13 localhost kernel: Badness in device_release at drivers/base/core > > .c:85 > > Thanks for the report. > > This patch should shut the warning up. And that's all it does. It doesn't stop the oops which potentically can happen when the struct device is freed (by the module being unloaded) while there is still a reference to the struct device. This is the whole point of the warning - to warn that the necessary release functionality is not present, and that the way the device is being used is buggy. Providing an empty release function is not a solution - it merely papers over the warning leaving the real bug behind. -- Russell King Linux kernel 2.6 ARM Linux - http://www.arm.linux.org.uk/ maintainer of: 2.6 PCMCIA - http://pcmcia.arm.linux.org.uk/ 2.6 Serial core