Subject: Re: [PATCH 1/6] vhost: get rid of vhost_poll_flush() wrapper
To:     Stefano Garzarella <sgarzare@redhat.com>
Cc:     "Michael S. Tsirkin" <mst@redhat.com>,
        Jason Wang <jasowang@redhat.com>,
        Stefan Hajnoczi <stefanha@redhat.com>, kvm@vger.kernel.org,
        virtualization@lists.linux-foundation.org, netdev@vger.kernel.org,
        linux-kernel@vger.kernel.org
References: <20211115153003.9140-1-arbn@yandex-team.com>
 <20211116144119.56ph52twuyc4jtdr@steredhat>
From:   Andrey Ryabinin <arbn@yandex-team.com>
Message-ID: <aff42feb-6ca5-1d9a-6ecf-74c213abfaa6@yandex-team.com>
Date:   Fri, 19 Nov 2021 13:30:10 +0300
User-Agent: Mozilla/5.0 (X11; Linux x86_64; rv:78.0) Gecko/20100101
 Thunderbird/78.14.0
MIME-Version: 1.0
In-Reply-To: <20211116144119.56ph52twuyc4jtdr@steredhat>
Content-Type: text/plain; charset=utf-8
Content-Language: en-US
Content-Transfer-Encoding: 8bit
Precedence: bulk


On 11/16/21 5:41 PM, Stefano Garzarella wrote:
> On Mon, Nov 15, 2021 at 06:29:58PM +0300, Andrey Ryabinin wrote:

>> void vhost_work_queue(struct vhost_dev *dev, struct vhost_work *work)
>> {
>>     if (!dev->worker)
>> @@ -663,7 +655,7 @@ void vhost_dev_stop(struct vhost_dev *dev)
>>     for (i = 0; i < dev->nvqs; ++i) {
>>         if (dev->vqs[i]->kick && dev->vqs[i]->handle_kick) {
>>             vhost_poll_stop(&dev->vqs[i]->poll);
>> -            vhost_poll_flush(&dev->vqs[i]->poll);
>> +            vhost_work_dev_flush(dev->vqs[i]->poll.dev);
> 
> Not related to this patch, but while looking at vhost-vsock I'm wondering if we can do the same here in vhost_dev_stop(), I mean move vhost_work_dev_flush() outside the loop and and call it once. (In another patch eventually)
> 

Yeah, seems reasonable. I can't see any reason why would subsequent vhost_poll_stop() require the vhost_work_dev_flush() in between.