On Mon, Aug 10, 2026 at 09:37:21PM +0300, Andy Shevchenko wrote: > > - if (chan->type != IIO_LIGHT && chan->type != IIO_PROXIMITY) > > + if (chan->type != IIO_LIGHT && chan->type != IIO_PROXIMITY && > > + chan->type != IIO_INTENSITY) > > return -EINVAL; > > Is this now a duplicating a bit with the below choices?
A bit, in the int_time and scale arms. The check at the top is pre-existing (this patch only adds IIO_INTENSITY to it) and is the only type filter for the RAW case, which serves all three types through chan->address with no per-type branching of its own. > > case IIO_CHAN_INFO_INT_TIME: > > if (chan->type == IIO_LIGHT) > > ret = regmap_field_read(data->reg_als_it, &index); > > - else > > + else if (chan->type == IIO_PROXIMITY) > > ret = regmap_field_read(data->reg_ps_it, &index); > > + else > > + return -EINVAL; > > If you want to keep a single point of return, use > > ret = -EINVAL; > > here and below in _read_raw(). Neither of the new else branches is reachable today: the colour channels declare RAW only, so no int_time or scale attribute exists for them, and in-kernel readers are gated by iio_channel_has_info(). I would rather keep them than have a bare else read as proximity for a colour channel. v2 writes them with ret = -EINVAL as you suggest, falling into the existing if (ret < 0) return ret. Thanks, Jorijn
