On 21-08-05 11:58:59, Greg Kroah-Hartman wrote:
> On Wed, Aug 04, 2021 at 09:38:09AM +0800, Peter Chen wrote:
> > On 21-08-02 22:52:12, Salah Triki wrote:
> > > Based on the documentation of usb_[disable|enable]_autosuspend(), the
> > > caller must hold udev's device lock.
> > >
> > > Signed-off-by: Salah Triki <salah.triki@xxxxxxxxx>
> > > ---
> > > drivers/usb/chipidea/otg_fsm.c | 2 ++
> > > 1 file changed, 2 insertions(+)
> > >
> > > diff --git a/drivers/usb/chipidea/otg_fsm.c b/drivers/usb/chipidea/otg_fsm.c
> > > index 6ed4b00dba96..2d4174250432 100644
> > > --- a/drivers/usb/chipidea/otg_fsm.c
> > > +++ b/drivers/usb/chipidea/otg_fsm.c
> > > @@ -518,12 +518,14 @@ static void ci_otg_loc_sof(struct otg_fsm *fsm, int on)
> > > if (!udev)
> > > return;
> > >
> > > + usb_lock_device(udev);
> > > if (on) {
> > > usb_disable_autosuspend(udev);
> > > } else {
> > > pm_runtime_set_autosuspend_delay(&udev->dev, 0);
> > > usb_enable_autosuspend(udev);
> > > }
> > > + usb_unlock_device(udev);
> > > }
> > >
> >
> > Acked-by: Peter Chen <peter.chen@xxxxxxxxxx>
>
> I think this is not ok, see the other threads...
Thanks for checking it, I read Alan's comments for the similar patches.
Yes, it doesn't need since all callers have already holds fsm->lock.
--
Thanks,
Peter Chen
