From mboxrd@z Thu Jan  1 00:00:00 1970
Return-Path: <linux-fsdevel-owner@vger.kernel.org>
Received: from mail-co1nam03on0112.outbound.protection.outlook.com ([104.47.40.112]:27472
	"EHLO NAM03-CO1-obe.outbound.protection.outlook.com"
	rhost-flags-OK-OK-OK-FAIL) by vger.kernel.org with ESMTP
	id S1751451AbcGSTHq (ORCPT <rfc822;linux-fsdevel@vger.kernel.org>);
	Tue, 19 Jul 2016 15:07:46 -0400
Message-ID: <578E7A79.7080003@hpe.com>
Date: Tue, 19 Jul 2016 15:07:37 -0400
From: Waiman Long <waiman.long@hpe.com>
MIME-Version: 1.0
To: Al Viro <viro@ZenIV.linux.org.uk>
CC: Jan Kara <jack@suse.com>, Jeff Layton <jlayton@poochiereds.net>,
	"J. Bruce Fields" <bfields@fieldses.org>,
	Tejun Heo <tj@kernel.org>,
	Christoph Lameter <cl@linux-foundation.org>,
	<linux-fsdevel@vger.kernel.org>, <linux-kernel@vger.kernel.org>,
	Ingo Molnar <mingo@redhat.com>,
	Peter Zijlstra <peterz@infradead.org>,
	Andi Kleen <andi@firstfloor.org>,
	Dave Chinner <dchinner@redhat.com>,
	Boqun Feng <boqun.feng@gmail.com>,
	Scott J Norton <scott.norton@hpe.com>,
	Douglas Hatch <doug.hatch@hpe.com>
Subject: Re: [PATCH v3 4/4] vfs: Use dlock list for superblock's inode list
References: <1468604383-40362-1-git-send-email-Waiman.Long@hpe.com> <1468604383-40362-5-git-send-email-Waiman.Long@hpe.com> <20160719052342.GF2356@ZenIV.linux.org.uk>
In-Reply-To: <20160719052342.GF2356@ZenIV.linux.org.uk>
Content-Type: text/plain; charset="ISO-8859-1"; format=flowed
Content-Transfer-Encoding: 7bit
Sender: linux-fsdevel-owner@vger.kernel.org
List-ID: <linux-fsdevel.vger.kernel.org>

On 07/19/2016 01:23 AM, Al Viro wrote:
> On Fri, Jul 15, 2016 at 01:39:43PM -0400, Waiman Long wrote:
>>   void iterate_bdevs(void (*func)(struct block_device *, void *), void *arg)
>>   {
>>   	struct inode *inode, *old_inode = NULL;
>> +	DEFINE_DLOCK_LIST_ITER(iter);
>>
>> -	spin_lock(&blockdev_superblock->s_inode_list_lock);
>> -	list_for_each_entry(inode,&blockdev_superblock->s_inodes, i_sb_list) {
>> -		struct address_space *mapping = inode->i_mapping;
>> +	while (dlock_list_next(&blockdev_superblock->s_inodes,&iter)) {
>> +		struct address_space *mapping;
>>
>> +		inode   = list_entry(iter.curr, struct inode, i_sb_list);
>> +		mapping = inode->i_mapping;
> TBH, I would very much prefer something like
> 	DEFINE_DLOCK_LIST_ITER(iter,&blockdev_superblock->s_inodes);
>
> 	dlist_for_each_entry(inode,&iter, i_sb_list) {
> 		mapping = inode->i_mapping;
>

Sure. I will make the necessary changes to make it happen. Thanks for 
the suggestion.

>> -		spin_unlock(&blockdev_superblock->s_inode_list_lock);
>> +		spin_unlock(iter.lock);
> ... and this might be worth dlist_{un,re}lock(&iter);

Good point. Will change the code accordingly.

Cheers,
Longman