Netdev List
 help / color / mirror / Atom feed
From: Alexandru Matei <alexandru.matei@uipath.com>
To: Stefano Garzarella <sgarzare@redhat.com>
Cc: Stefan Hajnoczi <stefanha@redhat.com>,
	"David S . Miller" <davem@davemloft.net>,
	Eric Dumazet <edumazet@google.com>,
	Jakub Kicinski <kuba@kernel.org>, Paolo Abeni <pabeni@redhat.com>,
	kvm@vger.kernel.org, virtualization@lists.linux-foundation.org,
	netdev@vger.kernel.org, linux-kernel@vger.kernel.org,
	Mihai Petrisor <mihai.petrisor@uipath.com>,
	Viorel Canja <viorel.canja@uipath.com>,
	Alexandru Matei <alexandru.matei@uipath.com>
Subject: Re: [PATCH v2] vsock/virtio: initialize the_virtio_vsock before using VQs
Date: Mon, 23 Oct 2023 17:59:45 +0300	[thread overview]
Message-ID: <632465d0-e04c-4e10-abb9-a740d6e3dc30@uipath.com> (raw)
In-Reply-To: <0624137c-85cf-4086-8256-af2b8405f434@uipath.com>

On 10/23/2023 5:52 PM, Alexandru Matei wrote:
> On 10/23/2023 5:29 PM, Stefano Garzarella wrote:
>> On Mon, Oct 23, 2023 at 05:08:33PM +0300, Alexandru Matei wrote:
>>> Once VQs are filled with empty buffers and we kick the host,
>>> it can send connection requests.  If 'the_virtio_vsock' is not
>>> initialized before, replies are silently dropped and do not reach the host.
>>>
>>> Fixes: 0deab087b16a ("vsock/virtio: use RCU to avoid use-after-free on the_virtio_vsock")
>>> Signed-off-by: Alexandru Matei <alexandru.matei@uipath.com>
>>> ---
>>> v2:
>>> - split virtio_vsock_vqs_init in vqs_init and vqs_fill and moved
>>>  the_virtio_vsock initialization after vqs_init
>>>
>>> net/vmw_vsock/virtio_transport.c | 9 +++++++--
>>> 1 file changed, 7 insertions(+), 2 deletions(-)
>>>
>>> diff --git a/net/vmw_vsock/virtio_transport.c b/net/vmw_vsock/virtio_transport.c
>>> index e95df847176b..92738d1697c1 100644
>>> --- a/net/vmw_vsock/virtio_transport.c
>>> +++ b/net/vmw_vsock/virtio_transport.c
>>> @@ -559,6 +559,11 @@ static int virtio_vsock_vqs_init(struct virtio_vsock *vsock)
>>>     vsock->tx_run = true;
>>>     mutex_unlock(&vsock->tx_lock);
>>>
>>> +    return 0;
>>> +}
>>> +
>>> +static void virtio_vsock_vqs_fill(struct virtio_vsock *vsock)
>>
>> What about renaming this function in virtio_vsock_vqs_start() and move also the setting of `tx_run` here?
> 
> It works but in this case we also need to move rcu_assign_pointer in virtio_vsock_vqs_start(), 
> the assignment needs to be right after setting tx_run to true and before filling the VQs.
> 

And if we move rcu_assign_pointer then there is no need to split the function in two,
We can move rcu_assign_pointer() directly inside virtio_vsock_vqs_init() after setting tx_run.

>>
>> Thanks,
>> Stefano
>>
>>> +{
>>>     mutex_lock(&vsock->rx_lock);
>>>     virtio_vsock_rx_fill(vsock);
>>>     vsock->rx_run = true;
>>> @@ -568,8 +573,6 @@ static int virtio_vsock_vqs_init(struct virtio_vsock *vsock)
>>>     virtio_vsock_event_fill(vsock);
>>>     vsock->event_run = true;
>>>     mutex_unlock(&vsock->event_lock);
>>> -
>>> -    return 0;
>>> }
>>>
>>> static void virtio_vsock_vqs_del(struct virtio_vsock *vsock)
>>> @@ -664,6 +667,7 @@ static int virtio_vsock_probe(struct virtio_device *vdev)
>>>         goto out;
>>>
>>>     rcu_assign_pointer(the_virtio_vsock, vsock);
>>> +    virtio_vsock_vqs_fill(vsock);
>>>
>>>     mutex_unlock(&the_virtio_vsock_mutex);
>>>
>>> @@ -736,6 +740,7 @@ static int virtio_vsock_restore(struct virtio_device *vdev)
>>>         goto out;
>>>
>>>     rcu_assign_pointer(the_virtio_vsock, vsock);
>>> +    virtio_vsock_vqs_fill(vsock);
>>>
>>> out:
>>>     mutex_unlock(&the_virtio_vsock_mutex);
>>> -- 
>>> 2.34.1
>>>
>>

  reply	other threads:[~2023-10-23 14:59 UTC|newest]

Thread overview: 8+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2023-10-23 14:08 [PATCH v2] vsock/virtio: initialize the_virtio_vsock before using VQs Alexandru Matei
2023-10-23 14:29 ` Stefano Garzarella
2023-10-23 14:52   ` Alexandru Matei
2023-10-23 14:59     ` Alexandru Matei [this message]
2023-10-23 15:13       ` Stefano Garzarella
2023-10-23 15:36         ` Alexandru Matei
2023-10-23 16:10           ` Stefano Garzarella
2023-10-23 16:28             ` Alexandru Matei

Reply instructions:

You may reply publicly to this message via plain-text email
using any one of the following methods:

* Save the following mbox file, import it into your mail client,
  and reply-to-all from there: mbox

  Avoid top-posting and favor interleaved quoting:
  https://en.wikipedia.org/wiki/Posting_style#Interleaved_style

* Reply using the --to, --cc, and --in-reply-to
  switches of git-send-email(1):

  git send-email \
    --in-reply-to=632465d0-e04c-4e10-abb9-a740d6e3dc30@uipath.com \
    --to=alexandru.matei@uipath.com \
    --cc=davem@davemloft.net \
    --cc=edumazet@google.com \
    --cc=kuba@kernel.org \
    --cc=kvm@vger.kernel.org \
    --cc=linux-kernel@vger.kernel.org \
    --cc=mihai.petrisor@uipath.com \
    --cc=netdev@vger.kernel.org \
    --cc=pabeni@redhat.com \
    --cc=sgarzare@redhat.com \
    --cc=stefanha@redhat.com \
    --cc=viorel.canja@uipath.com \
    --cc=virtualization@lists.linux-foundation.org \
    /path/to/YOUR_REPLY

  https://kernel.org/pub/software/scm/git/docs/git-send-email.html

* If your mail client supports setting the In-Reply-To header
  via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line before the message body.
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox