From mboxrd@z Thu Jan  1 00:00:00 1970
From: Veaceslav Falico <vfalico@gmail.com>
Subject: Re: [PATCH net-next 3/3] bonding: add proper __rcu annotation for
 curr_active_slave
Date: Tue, 15 Jul 2014 13:15:22 +0200
Message-ID: <20140715111522.GE28235@mikrodark.usersys.redhat.com>
References: <1405417604-7828-1-git-send-email-edumazet@google.com>
 <1405417604-7828-4-git-send-email-edumazet@google.com>
 <20140715103858.GD28235@mikrodark.usersys.redhat.com>
 <1405421569.10255.29.camel@edumazet-glaptop2.roam.corp.google.com>
Mime-Version: 1.0
Content-Type: text/plain; charset=us-ascii; format=flowed
Cc: Eric Dumazet <edumazet@google.com>,
	"David S. Miller" <davem@davemloft.net>, netdev@vger.kernel.org,
	Jay Vosburgh <j.vosburgh@gmail.com>,
	Andy Gospodarek <andy@greyhouse.net>,
	Nikolay Aleksandrov <nikolay@redhat.com>,
	Mahesh Bandewar <maheshb@google.com>
To: Eric Dumazet <eric.dumazet@gmail.com>
Return-path: <netdev-owner@vger.kernel.org>
Received: from mail-wg0-f44.google.com ([74.125.82.44]:40470 "EHLO
	mail-wg0-f44.google.com" rhost-flags-OK-OK-OK-OK) by vger.kernel.org
	with ESMTP id S1758039AbaGOLSV (ORCPT
	<rfc822;netdev@vger.kernel.org>); Tue, 15 Jul 2014 07:18:21 -0400
Received: by mail-wg0-f44.google.com with SMTP id m15so5391036wgh.27
        for <netdev@vger.kernel.org>; Tue, 15 Jul 2014 04:18:20 -0700 (PDT)
Content-Disposition: inline
In-Reply-To: <1405421569.10255.29.camel@edumazet-glaptop2.roam.corp.google.com>
Sender: netdev-owner@vger.kernel.org
List-ID: <netdev.vger.kernel.org>

On Tue, Jul 15, 2014 at 12:52:49PM +0200, Eric Dumazet wrote:
>On Tue, 2014-07-15 at 12:38 +0200, Veaceslav Falico wrote:
>> On Tue, Jul 15, 2014 at 02:46:44AM -0700, Eric Dumazet wrote:
>> >RCU was added to bonding in linux-3.12 but lacked proper sparse annotations.
>> >
>> >Using __rcu annotation actually helps to spot all accesses to bond->curr_active_slave
>> >are correctly protected, with LOCKDEP support.
>> >
>> >Signed-off-by: Eric Dumazet <edumazet@google.com>
>>
>> Thanks a lot for cleaning this up, it's really a huge mess here...
>>
>> One question though:
>>
>> ...snip...
>> >@@ -2046,7 +2051,7 @@ static void bond_miimon_commit(struct bonding *bond)
>> > 				bond_alb_handle_link_change(bond, slave,
>> > 							    BOND_LINK_DOWN);
>> >
>> >-			if (slave == bond->curr_active_slave)
>> >+			if (slave == rcu_access_pointer(bond->curr_active_slave))
>>
>> I guess you've meant rtnl_dereference()? As bond_miimon_commit() is ujnder
>> rtnl, not under rcu.
>
>This does not really matter here, as we only perform an equality test.
>
>(We dont read the pointer, then dereference it)
>
>rcu_access_pointer() is fine in this context, this makes sparse happy
>mostly.

Fair enough.

Acked-by: Veaceslav Falico <vfalico@gmail.com>