On 1/15/2020 12:59 AM, Stefano Garzarella wrote:
> On Tue, Jan 14, 2020 at 5:45 PM Stefan Hajnoczi <stefa...@gmail.com> wrote:
>>
>> On Tue, Jan 14, 2020 at 03:52:29PM +0800, pannengy...@huawei.com wrote:
>>> From: Pan Nengyuan <pannengy...@huawei.com>
>>>
>>> Receive/transmit/event vqs forgot to cleanup in vhost_vsock_unrealize. This
>>> patch save receive/transmit vq pointer in realize() and cleanup vqs
>>> through those vq pointers in unrealize(). The leak stack is as follow:
>>>
>>> Direct leak of 21504 byte(s) in 3 object(s) allocated from:
>>> #0 0x7f86a1356970 (/lib64/libasan.so.5+0xef970) ??:?
>>> #1 0x7f86a09aa49d (/lib64/libglib-2.0.so.0+0x5249d) ??:?
>>> #2 0x5604852f85ca (./x86_64-softmmu/qemu-system-x86_64+0x2c3e5ca)
>>> /mnt/sdb/qemu/hw/virtio/virtio.c:2333
>>> #3 0x560485356208 (./x86_64-softmmu/qemu-system-x86_64+0x2c9c208)
>>> /mnt/sdb/qemu/hw/virtio/vhost-vsock.c:339
>>> #4 0x560485305a17 (./x86_64-softmmu/qemu-system-x86_64+0x2c4ba17)
>>> /mnt/sdb/qemu/hw/virtio/virtio.c:3531
>>> #5 0x5604858e6b65 (./x86_64-softmmu/qemu-system-x86_64+0x322cb65)
>>> /mnt/sdb/qemu/hw/core/qdev.c:865
>>> #6 0x5604861e6c41 (./x86_64-softmmu/qemu-system-x86_64+0x3b2cc41)
>>> /mnt/sdb/qemu/qom/object.c:2102
>>>
>>> Reported-by: Euler Robot <euler.ro...@huawei.com>
>>> Signed-off-by: Pan Nengyuan <pannengy...@huawei.com>
>>> ---
>>> hw/virtio/vhost-vsock.c | 9 +++++++--
>>> include/hw/virtio/vhost-vsock.h | 2 ++
>>> 2 files changed, 9 insertions(+), 2 deletions(-)
>>>
>>> diff --git a/hw/virtio/vhost-vsock.c b/hw/virtio/vhost-vsock.c
>>> index f5744363a8..896c0174c1 100644
>>> --- a/hw/virtio/vhost-vsock.c
>>> +++ b/hw/virtio/vhost-vsock.c
>>> @@ -335,8 +335,10 @@ static void vhost_vsock_device_realize(DeviceState
>>> *dev, Error **errp)
>>> sizeof(struct virtio_vsock_config));
>>>
>>> /* Receive and transmit queues belong to vhost */
>>> - virtio_add_queue(vdev, VHOST_VSOCK_QUEUE_SIZE,
>>> vhost_vsock_handle_output);
>>> - virtio_add_queue(vdev, VHOST_VSOCK_QUEUE_SIZE,
>>> vhost_vsock_handle_output);
>>> + vsock->recv_vq = virtio_add_queue(vdev, VHOST_VSOCK_QUEUE_SIZE,
>>> + vhost_vsock_handle_output);
>>> + vsock->trans_vq = virtio_add_queue(vdev, VHOST_VSOCK_QUEUE_SIZE,
>>> + vhost_vsock_handle_output);
>>>
>>> /* The event queue belongs to QEMU */
>>> vsock->event_vq = virtio_add_queue(vdev, VHOST_VSOCK_QUEUE_SIZE,
>>> @@ -378,6 +380,9 @@ static void vhost_vsock_device_unrealize(DeviceState
>>> *dev, Error **errp)
>>> /* This will stop vhost backend if appropriate. */
>>> vhost_vsock_set_status(vdev, 0);
>>>
>>> + virtio_delete_queue(vsock->recv_vq);
>>> + virtio_delete_queue(vsock->trans_vq);
>>> + virtio_delete_queue(vsock->event_vq);
>>> vhost_dev_cleanup(&vsock->vhost_dev);
>>> virtio_cleanup(vdev);
>>> }
>>
>> Please delete the virtqueues after vhost cleanup (the reverse
>> initialization order). There is currently no reason why it has to be
>> done in reverse initialization order, your patch should work too, but
>> it's a good default for avoiding user-after-free bugs.
>>
>
> Agree!
>
> Since we are here, should we delete the queues also in the error path
> of vhost_vsock_device_realize()?
Yes, I will change the cleanup order and aslo delete in the error path.
Thanks.
>
> Thanks,
> Stefano
>
> .
>