On 09/09/2018 02:07 PM, Hauke Mehrtens wrote: [ ... ]>
+ if (!priv->clk_rate) { + dev_err(dev, "clock has wrong value of 0");This is misleading: It is not known if the clock rate is indeed 0. It is only known that the clock rate is below LTQ_WDT_DIVIDER.Ok I will update this.+ return -EIO;I am not happy with EIO in this case. There is no indication of an I/O error.Is -EINVAL better?
Yes. Thanks, Guenter