From mboxrd@z Thu Jan 1 00:00:00 1970 Return-Path: Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S1759639AbXJQKRv (ORCPT ); Wed, 17 Oct 2007 06:17:51 -0400 Received: (majordomo@vger.kernel.org) by vger.kernel.org id S1754117AbXJQKRm (ORCPT ); Wed, 17 Oct 2007 06:17:42 -0400 Received: from wa-out-1112.google.com ([209.85.146.177]:13610 "EHLO wa-out-1112.google.com" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1751129AbXJQKRl (ORCPT ); Wed, 17 Oct 2007 06:17:41 -0400 DomainKey-Signature: a=rsa-sha1; c=nofws; d=gmail.com; s=beta; h=received:message-id:date:from:sender:to:subject:cc:in-reply-to:mime-version:content-type:content-transfer-encoding:content-disposition:references:x-google-sender-auth; b=ng2I49WboMxkU6tkFzHYs/WXKGC2bm55+bUgLn9Gi7TPNbyAAChs3nlidnpK6MpFpNsRAN2xkdcE5JGuvtKpn8Ae+8lCAI80vav4ZIC0LejumnNvzKG5fredy78ddcViGUbDrMltzZWHifywIwIlihI5pKFZVlfEmnHJ5SMPCn0= Message-ID: <3ae72650710170317m75d0a664oe2bfa98ab70298e9@mail.gmail.com> Date: Wed, 17 Oct 2007 12:17:39 +0200 From: "Kay Sievers" To: "Rafael J. Wysocki" Subject: Re: linux-2.6.23-git3: Many sysfs-related warnings in dmesg Cc: "Greg KH" , "Jens Axboe" , "Linus Torvalds" , "Andrew Morton" , "Tejun Heo" , LKML In-Reply-To: <200710170158.10234.rjw@sisk.pl> MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 7bit Content-Disposition: inline References: <200710132126.33085.rjw@sisk.pl> <200710162313.20373.rjw@sisk.pl> <20071016213626.GA21815@kroah.com> <200710170158.10234.rjw@sisk.pl> X-Google-Sender-Auth: d8d40fb32528731a Sender: linux-kernel-owner@vger.kernel.org X-Mailing-List: linux-kernel@vger.kernel.org On 10/17/07, Rafael J. Wysocki wrote: > On Tuesday, 16 October 2007 23:36, Greg KH wrote: > > On Tue, Oct 16, 2007 at 11:13:19PM +0200, Rafael J. Wysocki wrote: > > > On Tuesday, 16 October 2007 22:50, Jens Axboe wrote: > > > > On Tue, Oct 16 2007, Greg KH wrote: > > > > > On Tue, Oct 16, 2007 at 10:04:51PM +0200, Rafael J. Wysocki wrote: > > > > > > On Tuesday, 16 October 2007 06:50, Greg KH wrote: > > > > > > > On Sat, Oct 13, 2007 at 09:26:32PM +0200, Rafael J. Wysocki wrote: > > > > > > > > Hi, > > > > > > > > > > > > > > > > There are many traces like this in my dmesg from 2.6.23-git3 (they don't > > > > > > > > appear for vanilla 2.6.23): > > > > > > > > > > > > > > > > <4>sysfs: duplicate filename 'ethxx1' can not be created > > > > > > > > WARNING: at /home/rafael/src/linux-2.6/fs/sysfs/dir.c:425 sysfs_add_one() > > > > > > > > > > > > > > > > Call Trace: > > > > > > > > [] sysfs_add_one+0x5c/0xc9 > > > > > > > > [] sysfs_create_link+0xd1/0x12c > > > > > > > > [] device_rename+0x17a/0x1db > > > > > > > > [] dev_change_name+0x114/0x20c > > > > > > > > [] dev_ifsioc+0x204/0x2d0 > > > > > > > > [] dev_ioctl+0x520/0x633 > > > > > > > > [] sk_alloc+0x37/0x10c > > > > > > > > [] up_read+0x9/0xb > > > > > > > > [] sock_ioctl+0x1fe/0x20c > > > > > > > > [] do_ioctl+0x2a/0x77 > > > > > > > > [] vfs_ioctl+0x251/0x26e > > > > > > > > [] sys_ioctl+0x5f/0x83 > > > > > > > > [] system_call+0x7e/0x83 > > > > > > > > > > > > > > > > net ethxx1: device_rename: sysfs_create_symlink failed (-17) > > > > > > > > sysfs: duplicate filename 'eth1' can not be created > > > > > > > > > > > > > > > > Everything seems to work, but this just looks fishy. > > > > > > > > > > > > > > This is a userspace program renaming your network device to a name that > > > > > > > is already in use. What distro and release is this? > > > > > > > > > > > > openSUSE 10.2, with some updates (not the most recent, though). > > > > > > > > > > Can you try 10.3? I think we fixed this issue there. > > > > > > > > Ehm nope: > > > > > > > > Oct 16 19:36:27 carl kernel: sysfs: duplicate filename 'wlan0_rename' > > > > can not be created > > > > Oct 16 19:36:27 carl kernel: udev: renamed network interface wmaster0 to > > > > eth1 > > > > Oct 16 19:36:27 carl kernel: WARNING: at fs/sysfs/dir.c:425 > > > > sysfs_add_one() > > > > Oct 16 19:36:27 carl kernel: [] sysfs_add_one+0x9f/0xe0 > > > > Oct 16 19:36:27 carl kernel: [] sysfs_create_link+0x86/0x110 > > > > Oct 16 19:36:27 carl kernel: [] device_rename+0x17a/0x1e0 > > > > Oct 16 19:36:27 carl kernel: [] filemap_fault+0x1ff/0x400 > > > > Oct 16 19:36:27 carl kernel: [] dev_change_name+0xb0/0x210 > > > > Oct 16 19:36:27 carl kernel: [] dev_ioctl+0x3c9/0x530 > > > > Oct 16 19:36:27 carl kernel: [] > > > > inotify_d_instantiate+0x18/0x80 > > > > Oct 16 19:36:27 carl kernel: [] do_page_fault+0x4a4/0x6f0 > > > > Oct 16 19:36:27 carl kernel: [] sock_ioctl+0x0/0x220 > > > > Oct 16 19:36:27 carl kernel: [] do_ioctl+0x2b/0x90 > > > > Oct 16 19:36:27 carl kernel: [] sys_socket+0x2b/0x50 > > > > Oct 16 19:36:27 carl kernel: [] vfs_ioctl+0x21e/0x2a0 > > > > Oct 16 19:36:27 carl kernel: [] sys_ioctl+0x3d/0x70 > > > > Oct 16 19:36:27 carl kernel: [] sysenter_past_esp+0x6b/0xa1 > > > > Oct 16 19:36:27 carl kernel: ======================= > > > > Oct 16 19:36:27 carl kernel: net wlan0_rename: device_rename: > > > > sysfs_create_symlink failed (-17) > > > > > > Plus it doesn't happen on 2.6.23 on my system, so I count it as a regression. > > > > It's a new warning I've added to the kernel to catch stupid userspace > > and kernel code that is quietly failing when they are doing things > > incorrectly with kobjects and sysfs. > > > > So it's not really a "regression" in that we never notified anyone of > > this kind of thing before, think of it as a new feature :) > > Well, it seems to be a little bit over the top, then. ;-) In the network renaming it's a bit noisy, yes, but for other users which we discovered by this, it already exposed real bugs. Kay