On Thursday, October 28, 2021 11:40:38 AM CEST Yang Yingliang wrote:
> Add the missing unlock before return from rtw_resume().
>
> Reported-by: Hulk Robot <hulkci@xxxxxxxxxx>
> Fixes: 2b42bd58b321 ("staging: r8188eu: introduce new os_dep dir for
RTL8188eu driver")
> Signed-off-by: Yang Yingliang <yangyingliang@xxxxxxxxxx>
> ---
> drivers/staging/r8188eu/os_dep/usb_intf.c | 4 +++-
> 1 file changed, 3 insertions(+), 1 deletion(-)
>
> diff --git a/drivers/staging/r8188eu/os_dep/usb_intf.c b/drivers/staging/
r8188eu/os_dep/usb_intf.c
> index 6b93830130e7..5a35d9fe3fc9 100644
> --- a/drivers/staging/r8188eu/os_dep/usb_intf.c
> +++ b/drivers/staging/r8188eu/os_dep/usb_intf.c
> @@ -298,8 +298,10 @@ static int rtw_resume(struct usb_interface *pusb_intf)
> pwrpriv->bkeepfwalive = false;
>
> DBG_88E("bkeepfwalive(%x)\n", pwrpriv->bkeepfwalive);
> - if (pm_netdev_open(pnetdev, true) != 0)
> + if (pm_netdev_open(pnetdev, true) != 0) {
> + mutex_unlock(&pwrpriv->lock);
> goto exit;
> + }
>
> netif_device_attach(pnetdev);
> netif_carrier_on(pnetdev);
> --
> 2.25.1
>
Good catch, well done!
I recently replaced semaphores with mutexes in this and other r8188eu driver
functions (commit 7e4c7947b42c "staging: r8188eu: Use a Mutex instead of a
binary semaphore") but, while making those changes, I didn't notice that the
'if' in rtw_resume () can lead to a jump to the "exit" label without properly
unlocking the "pwrpriv-> lock" mutex.
Acked-by: Fabio M. De Francesco <fmdefrancesco@xxxxxxxxx>
Thank you,
Fabio
