From mboxrd@z Thu Jan  1 00:00:00 1970
Received: from eggs.gnu.org ([2001:4830:134:3::10]:42201)
	by lists.gnu.org with esmtp (Exim 4.71)
	(envelope-from <zhangjie14@huawei.com>) id 1XKMxG-0004Nu-K1
	for qemu-devel@nongnu.org; Thu, 21 Aug 2014 03:38:48 -0400
Received: from Debian-exim by eggs.gnu.org with spam-scanned (Exim 4.71)
	(envelope-from <zhangjie14@huawei.com>) id 1XKMxA-0007F6-Ep
	for qemu-devel@nongnu.org; Thu, 21 Aug 2014 03:38:42 -0400
Received: from szxga02-in.huawei.com ([119.145.14.65]:36441)
	by eggs.gnu.org with esmtp (Exim 4.71)
	(envelope-from <zhangjie14@huawei.com>) id 1XKMx9-0007Cb-M2
	for qemu-devel@nongnu.org; Thu, 21 Aug 2014 03:38:36 -0400
Message-ID: <53F5A1C0.9000907@huawei.com>
Date: Thu, 21 Aug 2014 15:37:36 +0800
From: "Zhangjie (HZ)" <zhangjie14@huawei.com>
MIME-Version: 1.0
References: <1408424189-10510-1-git-send-email-jasowang@redhat.com>
	<53F46909.2040101@huawei.com> <20140820101833.GA17371@redhat.com>
	<20140820212153.GA25585@redhat.com>
In-Reply-To: <20140820212153.GA25585@redhat.com>
Content-Type: text/plain; charset="ISO-8859-1"
Content-Transfer-Encoding: 7bit
Subject: Re: [Qemu-devel] [PATCH V3] vhost_net: start/stop guest notifiers
	properly
List-Id: <qemu-devel.nongnu.org>
List-Unsubscribe: <https://lists.nongnu.org/mailman/options/qemu-devel>,
	<mailto:qemu-devel-request@nongnu.org?subject=unsubscribe>
List-Archive: <http://lists.nongnu.org/archive/html/qemu-devel>
List-Post: <mailto:qemu-devel@nongnu.org>
List-Help: <mailto:qemu-devel-request@nongnu.org?subject=help>
List-Subscribe: <https://lists.nongnu.org/mailman/listinfo/qemu-devel>,
	<mailto:qemu-devel-request@nongnu.org?subject=subscribe>
To: "Michael S. Tsirkin" <mst@redhat.com>
Cc: Jason Wang <jasowang@redhat.com>, qemu-devel@nongnu.org, William Dauchy <wdauchy@gmail.com>

On 2014/8/21 5:21, Michael S. Tsirkin wrote:
> On Wed, Aug 20, 2014 at 12:18:33PM +0200, Michael S. Tsirkin wrote:
>>> Because of vhost_net_set_vq_index, VM can be start successfully.
>>> But, after about 80 times of migration under my environment, virtual nic became unreachable again.
>>> When I use jprobe to notify tap, the virtual nic becomes reachable again. This shows that interrupts missing causes
>>> the problem.
>>
>> Could you please clarify what do you mean by "notify tap" here?
>> Thanks!
>  
> Or just post your jprobe script.
> 
Ok, thanks, this is the key function.
static int cont =1;
static int virtnet_poll_jprobe(struct napi_struct *napi, int budget)
{
        struct receive_queue *rq =
                container_of(napi, struct receive_queue, napi);
        struct virtnet_info *vi = rq->vq->vdev->priv;
        struct send_queue *sq;
        struct vring_virtqueue *vq;
        int i;

        if (cont == 1) {
                for (i = 0; i < curr_queue_pairs; i++) {
                        sq = &vi->sq[i];
                        vq = to_vvq(sq->vq);
                       vq->notify(sq->vq);
                }
                cont = 0;
        }
        jprobe_return();
        return 0;
}
When the problem occurs, one queue pair(each nic has two queue pairs) is active, so virtual nic can still receive skbs.
And, only one queue pair misses a interrupt.
To run the jprobe, some definition of function and structure should be copied to jprobe.c.

-- 
Best Wishes!
Zhang Jie