Return-Path: Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S1759627AbYBGHUU (ORCPT ); Thu, 7 Feb 2008 02:20:20 -0500 Received: (majordomo@vger.kernel.org) by vger.kernel.org id S1758807AbYBGHT6 (ORCPT ); Thu, 7 Feb 2008 02:19:58 -0500 Received: from mx1.suse.de ([195.135.220.2]:54694 "EHLO mx1.suse.de" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1758389AbYBGHT4 (ORCPT ); Thu, 7 Feb 2008 02:19:56 -0500 Date: Wed, 6 Feb 2008 23:18:07 -0800 From: Greg KH To: David Miller Cc: linux-kernel@vger.kernel.org, kay.sievers@vrfy.org Subject: Re: partition sysfs OOPS in current GIT Message-ID: <20080207071807.GA13866@suse.de> References: <20080207064210.GA22896@suse.de> <20080206.223944.200455039.davem@davemloft.net> <20080207070044.GA4851@suse.de> <20080206.231144.174648341.davem@davemloft.net> MIME-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Disposition: inline In-Reply-To: <20080206.231144.174648341.davem@davemloft.net> User-Agent: Mutt/1.5.16 (2007-06-09) Sender: linux-kernel-owner@vger.kernel.org List-ID: X-Mailing-List: linux-kernel@vger.kernel.org Content-Length: 2613 Lines: 76 On Wed, Feb 06, 2008 at 11:11:44PM -0800, David Miller wrote: > From: Greg KH > Date: Wed, 6 Feb 2008 23:00:44 -0800 > > > On Wed, Feb 06, 2008 at 10:39:44PM -0800, David Miller wrote: > > > From: Greg KH > > > Date: Wed, 6 Feb 2008 22:42:10 -0800 > > > > > > > On Wed, Feb 06, 2008 at 10:05:44PM -0800, David Miller wrote: > > > > > I found the problem, it's the "whole_disk" partition attribute. > > > > > > > > I don't understand that code at all, on 2.6.24, what does reading that > > > > file give you? At first glance, I don't see how that file would spit > > > > out anything and not give you the same kind of oops. > > > > > > > > you are in a maze of kobject pointers, all alike... > > > > > > It's supposed to just exist, and be an empty zero length file. > > > That's why it's given no ->show method pointer. > > > > > > It's existence just means that the partition is a "whole disk" > > > partition type. > > > > Can you try this patch to see if it solves the oops, and that the file > > is still there and works properly? > > It doesn't crash, but the file returns -EIO instead of zero when read. How about this attempt? thanks for your patience, greg "Kay owes me a beer" k-h --- fs/partitions/check.c | 17 ++++++++++------- 1 file changed, 10 insertions(+), 7 deletions(-) --- a/fs/partitions/check.c +++ b/fs/partitions/check.c @@ -319,6 +319,14 @@ void delete_partition(struct gendisk *di put_device(&p->dev); } +static ssize_t whole_disk_show(struct device *dev, + struct device_attribute *attr, char *buf) +{ + return 0; +} +static DEVICE_ATTR(whole_disk, S_IRUSR | S_IRGRP | S_IROTH, + whole_disk_show, NULL); + void add_partition(struct gendisk *disk, int part, sector_t start, sector_t len, int flags) { struct hd_struct *p; @@ -352,13 +360,8 @@ void add_partition(struct gendisk *disk, device_add(&p->dev); partition_sysfs_add_subdir(p); p->dev.uevent_suppress = 0; - if (flags & ADDPART_FLAG_WHOLEDISK) { - static struct attribute addpartattr = { - .name = "whole_disk", - .mode = S_IRUSR | S_IRGRP | S_IROTH, - }; - err = sysfs_create_file(&p->dev.kobj, &addpartattr); - } + if (flags & ADDPART_FLAG_WHOLEDISK) + err = device_create_file(&p->dev, &dev_attr_whole_disk); /* suppress uevent if the disk supresses it */ if (!disk->dev.uevent_suppress) -- To unsubscribe from this list: send the line "unsubscribe linux-kernel" in the body of a message to majordomo@vger.kernel.org More majordomo info at http://vger.kernel.org/majordomo-info.html Please read the FAQ at http://www.tux.org/lkml/