Hi Geert, Mark > > ret = of_property_read_u32(np, "dai-tdm-slot-num", &val); > > - if (ret && ret != -EINVAL) > > - return ret; > > if (!ret && slots) > > *slots = val; (snip) > That would be a change in behavior. Probably it would be fine for > existing users, though, as no existing DTS should cause these errors, > else sound wouldn't work. For a new DTS, it would silently ignore errors. > You are in a better position to make that decision, though. Thanks I will post that patch if Gerrt's patch was applied. > BTW, is there any specific reason the code always checks for the > presence of "dai-tdm-slot-num", even if slots is NULL, and the result > sn't used? I.e. would > > if (slots) { > ret = of_property_read_u32(np, "dai-tdm-slot-num", &val); > if (!ret) > *slots = val; > else if (ret != -EINVAL) > return ret; > } I'm not 100% sure, but -num is used by the driver which can select the number of TDM. Some driver uses default / fixed number. And -width too. So I think these can be independent. Thank you for your help !! Best regards --- Kuninori Morimoto