Message-ID: <4860CEF2.4010003@fr.ibm.com>
Date: Tue, 24 Jun 2008 12:39:46 +0200
From: Daniel Lezcano <dlezcano@fr.ibm.com>
User-Agent: Thunderbird 2.0.0.14 (X11/20080501)
MIME-Version: 1.0
To: Tejun Heo <htejun@gmail.com>
CC: Benjamin Thery <benjamin.thery@bull.net>,
       Greg Kroah-Hartman <gregkh@suse.de>,
       Andrew Morton <akpm@linux-foundation.org>,
       Eric Biederman <ebiederm@xmission.com>, Serge Hallyn <serue@us.ibm.com>,
       linux-kernel@vger.kernel.org, Al Viro <viro@ftp.linux.org.uk>,
       Linux Containers <containers@lists.osdl.org>
Subject: Re: [PATCH 05/11] sysfs: sysfs_chmod_file handle multiple superblocks
References: <20080618170729.808539948@theryb.frec.bull.fr> <20080618170730.853353875@theryb.frec.bull.fr> <485DD93A.6020600@gmail.com> <486018CB.80005@fr.ibm.com> <48607BE7.2060203@gmail.com>
In-Reply-To: <48607BE7.2060203@gmail.com>
Content-Type: text/plain; charset=UTF-8; format=flowed
Content-Transfer-Encoding: 7bit
Sender: linux-kernel-owner@vger.kernel.org
Content-Length: 1607
Lines: 40

Tejun Heo wrote:
> Hello,
> 
> Daniel Lezcano wrote:
>>> I think it would be great if sysfs_chmod_file can do all-or-nothing
>>> instead of failing half way through but given the interface of
>>> notify_change(), it could be difficult to implement.  Any ideas?
>> Is it acceptable to queue the notifications in a list until we are in
>> the loop and loop again to notify when exiting the first loop without
>> error ?
> 
> Can you please take a look at the following patch?
> 
> http://article.gmane.org/gmane.linux.file-systems/24484
> 
> Which replaces notify_change() call to two calls to sysfs_setattr() and
> fsnotify_change().  The latter never fails and the former should always
> succeed if inode_change_ok() succeeds (inode_setattr() never fails
> unless the size is changing), so I think the correct thing to do is...
> 
> * Separate out sysfs_do_setattr() which doesn't do inode_change_ok() and
> just sets the attributes.  Making it a void function which triggers
> WARN_ON() when inode_setattr() fails would be a good idea.
> 
> * Implement sysfs_chmod_file() in similar way rename/move are
> implemented - allocate all resources and check conditions and then iff
> everything looks okay commit the operation by calling sysfs_do_setattr().
> 
> How does that sound?

Much better than my first proposition :)

I will do a separate patchset.

Thanks.
--
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/