Re: [PATCH v3 2/2] iio: temperature: add support for MCP998X

[Andy Shevchenko]

On Thu, Jun 19, 2025 at 10:22 AM <Victor.Duicu@xxxxxxxxxxxxx> wrote:
> On Sat, 2025-06-14 at 00:50 +0300, Andy Shevchenko wrote:
> > On Fri, Jun 13, 2025 at 4:02 PM <victor.duicu@xxxxxxxxxxxxx> wrote:

...

> > > +MICROCHIP MCP9982 TEMPERATURE DRIVER
> > > +M:     Victor Duicu <victor.duicu@xxxxxxxxxxxxx>
> > > +L:     linux-iio@xxxxxxxxxxxxxxx
> > > +S:     Supported
> > > +F:
> > > Documentation/devicetree/bindings/iio/temperature/microchip,mcp9982
> > > .yaml
> > > +F:     drivers/iio/temperature/mcp9982.c
> >
> > So, with the first patch only the dangling file will be present
> > without record in MAINTAINERS. Please, make sure that your DT schema
> > file is in MAINTAINERS.
>
> Are you referring here to the file sysfs-bus-iio-temperature-mcp9982?
> This file was in v2 where there were a few custom attributes. In v3
> I removed them, so the driver currently doesn't have custom attributes.
> Should I had added it to the files in MAINTAINERS?

You should have added the file to the MAINTAINERS in the same patch it
appears. Not in some arbitrary change afterwards.

> Isn't the yaml file sufficient to describe the devicetree? Should I
> also add a dts file?

No, this is not the point.

...

> > > +#define MCP9982_CHAN(index, si, __address)
> > > ({                                          \
> > > +       struct iio_chan_spec __chan =
> > > {                                                 \
> >
> > Why not compound literal?
> >
> In v2 I used compound literal, but Jonathan suggested to add
> the struct in the macro. After describing the reasoning, we
> agreed to code it like this.

Neither of the versions has a compound literal.

> > > };
> > >              \
> > > +
> > > __chan;
> > >              \
> > > +})


https://gcc.gnu.org/onlinedocs/gcc/Compound-Literals.html

-- 
With Best Regards,
Andy Shevchenko