On Wed, Nov 13, 2019 at 4:23 PM Dmitry Torokhov
[off-list ref] wrote:

On Wed, Nov 13, 2019 at 02:36:56PM +0800, Chuhong Yuan wrote:

quoted

The driver forgets to deal with work and workqueue in remove like what
is done when probe fails.
Add the missed operations to fix it.

Is it really possible for the work to still be pending when fully
registered device is properly unregistered? I thought we'd wait for
successful data acquisition in rmi_f54_buffer_queue() before unregister
can complete.

In fact, I am not familiar with the mechanism here...
I have checked other drivers with video_unregister_device and found none of
them deals with work in remove.
Therefore, I think your opinion should be right and we only need to deal with
the workqueue.

Regards,
Chuhong

quoted

Signed-off-by: Chuhong Yuan <redacted>
---
 drivers/input/rmi4/rmi_f54.c | 3 +++
 1 file changed, 3 insertions(+)

diff --git a/drivers/input/rmi4/rmi_f54.c b/drivers/input/rmi4/rmi_f54.c
index 710b02595486..2c0cde5c775c 100644
--- a/drivers/input/rmi4/rmi_f54.c
+++ b/drivers/input/rmi4/rmi_f54.c

@@ -730,6 +730,9 @@ static void rmi_f54_remove(struct rmi_function *fn)

      video_unregister_device(&f54->vdev);
      v4l2_device_unregister(&f54->v4l2);
+     cancel_delayed_work_sync(&f54->work);
+     flush_workqueue(f54->workqueue);
+     destroy_workqueue(f54->workqueue);
 }

 struct rmi_function_handler rmi_f54_handler = {
--

2.23.0

Thanks.

--
Dmitry