From mboxrd@z Thu Jan  1 00:00:00 1970
Return-Path: <linux-kernel-owner@vger.kernel.org>
Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand
        id S1754054AbdKMRW7 (ORCPT <rfc822;w@1wt.eu>);
        Mon, 13 Nov 2017 12:22:59 -0500
Received: from mail-pf0-f181.google.com ([209.85.192.181]:54564 "EHLO
        mail-pf0-f181.google.com" rhost-flags-OK-OK-OK-OK) by vger.kernel.org
        with ESMTP id S1753988AbdKMRW4 (ORCPT
        <rfc822;linux-kernel@vger.kernel.org>);
        Mon, 13 Nov 2017 12:22:56 -0500
X-Google-Smtp-Source: AGs4zMazzXRkylaWJRbmKGRyDn+iQHcmAN/LnJt1VAbUlhRM8u0bCCiRHnqRecscxwo+h13JUWiZsA==
Date: Mon, 13 Nov 2017 09:22:53 -0800
From: Stephen Hemminger <stephen@networkplumber.org>
To: Vitaly Kuznetsov <vkuznets@redhat.com>
Cc: David Miller <davem@davemloft.net>,
        Stephen Hemminger <sthemmin@microsoft.com>, eric.dumazet@gmail.com,
        netdev@vger.kernel.org, haiyangz@microsoft.com,
        linux-kernel@vger.kernel.org, devel@linuxdriverproject.org
Subject: Re: [RFC] hv_netvsc: safer orderly shutdown
Message-ID: <20171113092253.68dd8ae1@xeon-e3>
In-Reply-To: <87ineebg44.fsf@vitty.brq.redhat.com>
References: <20171102103531.5514-1-vkuznets@redhat.com>
        <20171108.103105.142249724388345463.davem@davemloft.net>
        <20171110125342.7aa7943c@shemminger-XPS-13-9360>
        <87k1yyjty6.fsf@vitty.brq.redhat.com>
        <CAOaVG15dBsPcpi1buaKyW5oRGmGUczgaO54De6eYXUqp-1z+gw@mail.gmail.com>
        <87ineebg44.fsf@vitty.brq.redhat.com>
MIME-Version: 1.0
Content-Type: text/plain; charset=US-ASCII
Content-Transfer-Encoding: 7bit
Sender: linux-kernel-owner@vger.kernel.org
List-ID: <linux-kernel.vger.kernel.org>
X-Mailing-List: linux-kernel@vger.kernel.org

On Mon, 13 Nov 2017 11:57:47 +0100
Vitaly Kuznetsov <vkuznets@redhat.com> wrote:

> Stephen Hemminger <stephen@networkplumber.org> writes:
> 
> >
> > The NAPI disable is already handled by rndis close.  
> 
> Sorry, but I'm probably missing something: I can only see
> netif_napi_del() call in netvsc_device_remove() but this happens much
> later. And I don see us doing napi_disable() anywhere on the path.
> But I'm probably missing something.
> 

You need to keep NAPI running to handle transmit completions.
Disabling the Tx and Rx filter should keep spurious activity
away until the halt is done.