On 20/05/2018 20:40:01+0200, Uwe Kleine-König wrote:
> Hello Alexandre,
>
> On Sun, May 20, 2018 at 08:05:20PM +0200, Alexandre Belloni wrote:
> > On 20/05/2018 15:37:23+0200, Uwe Kleine-König wrote:
> > > @@ -200,8 +242,21 @@ static int pcf2127_probe(struct device *dev, struct regmap *regmap,
> > >
> > > pcf2127->rtc = devm_rtc_device_register(dev, name, &pcf2127_rtc_ops,
> > > THIS_MODULE);
> > > + if (IS_ERR(pcf2127->rtc))
> > > + return PTR_ERR(pcf2127->rtc);
> > > +
> > > + if (has_nvmem) {
> > > + struct nvmem_config nvmem_cfg = {
> > > + .priv = pcf2127,
> > > + .reg_read = pcf2127_nvmem_read,
> > > + .reg_write = pcf2127_nvmem_write,
> > > + .size = 512,
> > > + };
> > > +
> > > + ret = rtc_nvmem_register(pcf2127->rtc, &nvmem_cfg);
> > > + }
> > >
> > > - return PTR_ERR_OR_ZERO(pcf2127->rtc);
> > > + return ret;
> >
> > You must not return an error here once devm_rtc_device_register has
> > succeeded.
>
> Why? Sounds like something that should be fixed.
>
This is the explanation:
http://patchwork.ozlabs.org/patch/324263/
--
Alexandre Belloni, Bootlin (formerly Free Electrons)
Embedded Linux and Kernel engineering
https://bootlin.com
