On Wed, Sep 28, 2022 at 2:49 PM Roger Quadros <rogerq@xxxxxxxxxx> wrote:
>
> Hello Pawel,
>
> On 28/09/2022 09:40, Pawel Laszczak wrote:
> >>
> >> From: Dongliang Mu <mudongliangabcd@xxxxxxxxx>
> >>
> >> Smatch reports the following error:
> >>
> >> drivers/usb/cdns3/cdns3-plat.c:113 cdns3_plat_probe() warn:
> >> platform_get_irq() does not return zero
> >>
> >>From the document, platform_get_irq_byname_optional only returns
> >> non-zero value, and negative value on failure.
> >>
> >> Fix this by removing the zero value checking.
> >>
> >> Signed-off-by: Dongliang Mu <mudongliangabcd@xxxxxxxxx>
> >> ---
> >> drivers/usb/cdns3/cdns3-plat.c | 2 --
> >> 1 file changed, 2 deletions(-)
> >>
> >> diff --git a/drivers/usb/cdns3/cdns3-plat.c b/drivers/usb/cdns3/cdns3-plat.c
> >> index dc068e940ed5..2bc5d094548b 100644
> >> --- a/drivers/usb/cdns3/cdns3-plat.c
> >> +++ b/drivers/usb/cdns3/cdns3-plat.c
> >> @@ -110,8 +110,6 @@ static int cdns3_plat_probe(struct platform_device *pdev)
> >> cdns->wakeup_irq = platform_get_irq_byname_optional(pdev, "wakeup");
> >> if (cdns->wakeup_irq == -EPROBE_DEFER)
> >> return cdns->wakeup_irq;
> >> - else if (cdns->wakeup_irq == 0)
> >> - return -EINVAL;
> >>
> > I think that here we should have:
> > else if (cdns->wakeup_irq == -ENXIO)
> > return -EINVAL;
> > because of function:
> > platform_get_irq_byname_optional -> __platform_get_irq_byname returns
> > irq number (>0), -EPROBE_DEFFER or -ENXIO
>
> But this is changing functionality and should come as a new patch.
I agree. Pawel, you should submit a new patch. This satisfies the rule
of kernel patching.
>
> The original patch is correct as it doesn't change existing code
> functionality.
>
> >
> >
> > thanks
> > Pawel
> >
> >> if (cdns->wakeup_irq < 0) {
> >> dev_dbg(dev, "couldn't get wakeup irq\n");
> >> --
> >> 2.35.1
> >
>
> cheers,
> -roger
