From mboxrd@z Thu Jan 1 00:00:00 1970 From: roopa Subject: Re: [PATCH net-next] bridge: fix setlink/dellink notifications Date: Wed, 14 Jan 2015 12:57:12 -0800 Message-ID: <54B6D828.3030204@cumulusnetworks.com> References: <1421218123-18346-1-git-send-email-roopa@cumulusnetworks.com> <20150114194137.GC2105@casper.infradead.org> Mime-Version: 1.0 Content-Type: text/plain; charset=ISO-8859-1; format=flowed Content-Transfer-Encoding: 7bit Cc: netdev@vger.kernel.org, shemminger@vyatta.com, vyasevic@redhat.com, john.fastabend@gmail.com, jhs@mojatatu.com, sfeldma@gmail.com, jiri@resnulli.us, wkok@cumulusnetworks.com To: Thomas Graf Return-path: Received: from mail-pa0-f54.google.com ([209.85.220.54]:52037 "EHLO mail-pa0-f54.google.com" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1750719AbbANU5O (ORCPT ); Wed, 14 Jan 2015 15:57:14 -0500 Received: by mail-pa0-f54.google.com with SMTP id fb1so12790291pad.13 for ; Wed, 14 Jan 2015 12:57:14 -0800 (PST) In-Reply-To: <20150114194137.GC2105@casper.infradead.org> Sender: netdev-owner@vger.kernel.org List-ID: On 1/14/15, 11:41 AM, Thomas Graf wrote: > On 01/13/15 at 10:48pm, roopa@cumulusnetworks.com wrote: >> 2) Generate one notification from master and the other notification from >> self (This seems to be ideal): >> For master: the master driver will send notification (bridge in this >> example) >> For self: the self driver will send notification (rocker in the above >> example. It can use helpers from rtnetlink.c to do so. Like the >> ndo_dflt_bridge_getlink api). >> >> This patch implements 2) (leaving the 'rtnl_bridge_notify' around to be used >> with 'self'). >> >> CC'ing others who might be affected by this change for review. >> >> Signed-off-by: Roopa Prabhu > I haven't digested this line by line yet but I agree that what you > describe above would be a good end state. > > If I read the patch correctly then we would omit one notification > for the master case. yes, correct. > Were both notifications exactly identical > previously? > This has the chance of breaking existing users terribly. yes AFAICT, because rtnl_bridge_notify() from rtnetlink.c for bridge results in calling br_fill_ifinfo() via ndo_bridge_getlink() and the notification from bridge driver br_ifinfo_notify() also ends up calling br_fill_ifinfo(). Thanks, Roopa