Re: [RFC PATCH 1/5] usb: core: hcd: Modularize HCD stop configuration in usb_stop_hcd()

[Greg Kroah-Hartman <gregkh@linuxfoundation.org>]

On Tue, Aug 24, 2021 at 08:48:45PM +0530, Kishon Vijay Abraham I wrote:
> Hi Greg,
> 
> On 24/08/21 6:36 pm, Greg Kroah-Hartman wrote:
> > On Tue, Aug 24, 2021 at 04:22:58PM +0530, Kishon Vijay Abraham I wrote:
> >> No functional change. Since configuration to stop HCD is invoked from
> >> multiple places, group all of them in usb_stop_hcd().
> >>
> >> Signed-off-by: Kishon Vijay Abraham I <kishon@ti.com>
> >> ---
> >>  drivers/usb/core/hcd.c | 42 +++++++++++++++++++++++++-----------------
> >>  1 file changed, 25 insertions(+), 17 deletions(-)
> >>
> >> diff --git a/drivers/usb/core/hcd.c b/drivers/usb/core/hcd.c
> >> index 0f8b7c93310e..c036ba5311b3 100644
> >> --- a/drivers/usb/core/hcd.c
> >> +++ b/drivers/usb/core/hcd.c
> >> @@ -2760,6 +2760,29 @@ static void usb_put_invalidate_rhdev(struct usb_hcd *hcd)
> >>  	usb_put_dev(rhdev);
> >>  }
> >>  
> >> +/**
> >> + * usb_stop_hcd - Halt the HCD
> >> + * @hcd: the usb_hcd that has to be halted
> >> + *
> >> + * Stop the timer and invoke ->stop() callback on the HCD
> >> + */
> >> +static void usb_stop_hcd(struct usb_hcd *hcd)
> >> +{
> >> +	if (!hcd)
> >> +		return;
> > 
> > That's impossible to hit, so no need to check for it, right?
> 
> Patch 3 of this series adds support for registering roothub of shared
> HCD. So after that patch there can be a case where shared_hcd is NULL.
> The other option would be to check for non-null value in hcd and then
> invoke usb_stop_hcd().

Then add the check when you need it please.

thanks,

greg k-h