From mboxrd@z Thu Jan 1 00:00:00 1970 Return-Path: Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S269409AbUI3Ss6 (ORCPT ); Thu, 30 Sep 2004 14:48:58 -0400 Received: (majordomo@vger.kernel.org) by vger.kernel.org id S269406AbUI3Ss4 (ORCPT ); Thu, 30 Sep 2004 14:48:56 -0400 Received: from mail.tmr.com ([216.238.38.203]:12295 "EHLO gatekeeper.tmr.com") by vger.kernel.org with ESMTP id S269417AbUI3SsU (ORCPT ); Thu, 30 Sep 2004 14:48:20 -0400 To: linux-kernel@vger.kernel.org Path: not-for-mail From: Bill Davidsen Newsgroups: mail.linux-kernel Subject: Re: mmap() on cdrom files fails in 2.6.9-rc2-mmX Date: Thu, 30 Sep 2004 14:49:32 -0400 Organization: TMR Associates, Inc Message-ID: References: <20040928214246.41b80d30.khali@linux-fr.org> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii; format=flowed Content-Transfer-Encoding: 7bit X-Trace: gatekeeper.tmr.com 1096569646 6527 192.168.12.100 (30 Sep 2004 18:40:46 GMT) X-Complaints-To: abuse@tmr.com User-Agent: Mozilla/5.0 (X11; U; Linux i686; en-US; rv:1.7.3) Gecko/20040913 X-Accept-Language: en-us, en In-Reply-To: <20040928214246.41b80d30.khali@linux-fr.org> Sender: linux-kernel-owner@vger.kernel.org X-Mailing-List: linux-kernel@vger.kernel.org Jean Delvare wrote: > Hi all, > > (Please CC me on reply, I am not subscribed to the LKML.) > > I think I found a bug in 2.6.9-rc2-mm4. It doesn't seem to be able to > mmap() files located on cdroms. Same problem with -mm3 and -mm1. > 2.6.9-rc2 works fine. I reproduced it on two completely different > systems, so I guess it isn't device dependant. > > I tried to get mmap() to fail on other files with similar properties > (different owner, read only) but it won't. The problem seems to affect > only files located on cdroms. > > I also tried on a loop device-mounted cdrom image, in case the problem > was related to the iso9660 filesystem. The problem wouldn't show. It > seems to fail only on real cdroms. > > I searched the broken-out directory but wouldn't find any candidate to > backout. Is your mmap length a multiple of 2k? If not please resize and try again. There is some less traveled code when physical sector size is not 512 bytes. Yes, this is a wild guess, but it's easy to test if you have the kernel built. -- -bill davidsen (davidsen@tmr.com) "The secret to procrastination is to put things off until the last possible moment - but no longer" -me