From mboxrd@z Thu Jan 1 00:00:00 1970 From: Jason Wang Subject: Re: [PATCH v2] virtio_net: avoid using netif_tx_disable() for serializing tx routine Date: Wed, 17 Oct 2018 17:02:35 +0800 Message-ID: <276ac5bb-90f7-5fb6-a826-0bb05ecfa069@redhat.com> References: <20181017075918.7043-1-ake@igel.co.jp> Mime-Version: 1.0 Content-Type: text/plain; charset=utf-8; format=flowed Content-Transfer-Encoding: 8bit Cc: "Michael S. Tsirkin" , "David S. Miller" , virtualization@lists.linux-foundation.org, netdev@vger.kernel.org, linux-kernel@vger.kernel.org To: Ake Koomsin Return-path: In-Reply-To: <20181017075918.7043-1-ake@igel.co.jp> Content-Language: en-US Sender: linux-kernel-owner@vger.kernel.org List-Id: netdev.vger.kernel.org On 2018/10/17 下午3:59, Ake Koomsin wrote: > Commit 713a98d90c5e ("virtio-net: serialize tx routine during reset") > introduces netif_tx_disable() after netif_device_detach() in order to > avoid use-after-free of tx queues. However, there are two issues. > > 1) Its operation is redundant with netif_device_detach() if the interface > is running. > 2) In case of the interface is not running before suspending and > resuming, the tx does not get resumed by netif_device_attach(). > This results in losing network connectivity. > > It is better to use netif_tx_lock()/netif_tx_unlock() instead for > serializing tx routine during reset. This also preserves the symmetry > of netif_device_detach() and netif_device_attach(). > > Fixes commit 713a98d90c5e ("virtio-net: serialize tx routine during reset") > Signed-off-by: Ake Koomsin > --- > drivers/net/virtio_net.c | 5 ++++- > 1 file changed, 4 insertions(+), 1 deletion(-) > > diff --git a/drivers/net/virtio_net.c b/drivers/net/virtio_net.c > index 3f5aa59c37b7..41ccf9c994a4 100644 > --- a/drivers/net/virtio_net.c > +++ b/drivers/net/virtio_net.c > @@ -2267,8 +2267,9 @@ static void virtnet_freeze_down(struct virtio_device *vdev) > /* Make sure no work handler is accessing the device */ > flush_work(&vi->config_work); > > + netif_tx_lock(vi->dev); > netif_device_detach(vi->dev); > - netif_tx_disable(vi->dev); > + netif_tx_unlock(vi->dev); Sorry for not finding this earlier. I think we should use netif_tx_lock_bh() to prevent start_xmit() to run under bh. Thanks > cancel_delayed_work_sync(&vi->refill); > > if (netif_running(vi->dev)) { > @@ -2304,7 +2305,9 @@ static int virtnet_restore_up(struct virtio_device *vdev) > } > } > > + netif_tx_lock(vi->dev); > netif_device_attach(vi->dev); > + netif_tx_unlock(vi->dev); > return err; > } >