From mboxrd@z Thu Jan  1 00:00:00 1970
From: Jakub Kicinski <kubakici@wp.pl>
Subject: Re: [patch net-next 3/5] net: sch: prio: Add offload ability to
 PRIO qdisc
Date: Fri, 12 Jan 2018 01:42:16 -0800
Message-ID: <20180112014216.396904f3@laptop>
References: <20180111102102.4310-1-jiri@resnulli.us>
        <20180111102102.4310-4-jiri@resnulli.us>
        <20180111152537.3a90c722@cakuba.netronome.com>
        <VI1PR05MB35206C809B87495A33F2C8CABF160@VI1PR05MB3520.eurprd05.prod.outlook.com>
        <20180111160013.6f523996@cakuba.netronome.com>
        <VI1PR05MB35209D291846A0D2FA9AFD08BF170@VI1PR05MB3520.eurprd05.prod.outlook.com>
        <DB5PR05MB18955C7EA9DC6CD7D00C4764AC170@DB5PR05MB1895.eurprd05.prod.outlook.com>
Mime-Version: 1.0
Content-Type: text/plain; charset=US-ASCII
Content-Transfer-Encoding: 7bit
Cc: Yuval Mintz <yuvalm@mellanox.com>, Jiri Pirko <jiri@resnulli.us>,
        "netdev@vger.kernel.org" <netdev@vger.kernel.org>,
        "davem@davemloft.net" <davem@davemloft.net>,
        "Ido Schimmel" <idosch@mellanox.com>, mlxsw <mlxsw@mellanox.com>,
        "jhs@mojatatu.com" <jhs@mojatatu.com>,
        "xiyou.wangcong@gmail.com" <xiyou.wangcong@gmail.com>
To: Nogah Frankel <nogahf@mellanox.com>
Return-path: <netdev-owner@vger.kernel.org>
Received: from mx3.wp.pl ([212.77.101.10]:23151 "EHLO mx3.wp.pl"
        rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP
        id S1754430AbeALJm0 (ORCPT <rfc822;netdev@vger.kernel.org>);
        Fri, 12 Jan 2018 04:42:26 -0500
In-Reply-To: <DB5PR05MB18955C7EA9DC6CD7D00C4764AC170@DB5PR05MB1895.eurprd05.prod.outlook.com>
Sender: netdev-owner@vger.kernel.org
List-ID: <netdev.vger.kernel.org>

On Fri, 12 Jan 2018 09:20:46 +0000
Nogah Frankel <nogahf@mellanox.com> wrote:

> > > > > Could we please pass the full qstats on replace and destroy.
> > > > > This simplifies the driver code and allows handling the qlen
> > > > > as well as backlog.  Please see the 2 patch series I sent
> > > > > earlier yesterday.  
> 
> On replace - no problem.
> On destroy, I think it is redundant because the qdisc is going to be
> deleted anyway.

Right, the hope was it wouldn't matter and could make the code easier
to follow, but the replace with new qdisc complicates things.

> (To make sure that we are on the same page, if replace in the HW
> fails, we "rollback" only the backlog and qlen. The rest of the
> counters like TX count should reflect the number of the packets that
> passed in the HW while this qdisc was offloaded)

Agreed!