Re: [PATCH] net: usb: usbnet: fix use-after-free in race on workqueue

[Paolo Abeni <pabeni@redhat.com>]

On 6/25/25 11:33 AM, Peter GJ. Park wrote:
> When usbnet_disconnect() queued while usbnet_probe() processing,
> it results to free_netdev before kevent gets to run on workqueue,
> thus workqueue does assign_work() with referencing freeed memory address.
> 
> For graceful disconnect and to prevent use-after-free of netdev pointer,
> the fix adds canceling work and timer those are placed by usbnet_probe()
> 
> Signed-off-by: Peter GJ. Park <gyujoon.park@samsung.com>

You should include a suitable fixes tag, and you should have specified
the target tree ('net' in this case) in the prefix subjext

> ---
>  drivers/net/usb/usbnet.c | 3 +++
>  1 file changed, 3 insertions(+)
> 
> diff --git a/drivers/net/usb/usbnet.c b/drivers/net/usb/usbnet.c
> index c04e715a4c2ade3bc5587b0df71643a25cf88c55..3c5d9ba7fa6660273137c80106746103f84f5a37 100644
> --- a/drivers/net/usb/usbnet.c
> +++ b/drivers/net/usb/usbnet.c
> @@ -1660,6 +1660,9 @@ void usbnet_disconnect (struct usb_interface *intf)
>  	usb_free_urb(dev->interrupt);
>  	kfree(dev->padding_pkt);
>  
> +	timer_delete_sync(&dev->delay);
> +	tasklet_kill(&dev->bh);
> +	cancel_work_sync(&dev->kevent);
>  	free_netdev(net);

This happens after unregister_netdev(), which calls usbnet_stop() that
already performs the above cleanup. How the race is supposed to take place?

/P