On Mon, Jan 31, 2022 at 03:23:13PM +0200, Andy Shevchenko wrote:
> On Mon, Jan 31, 2022 at 01:08:32PM +0100, Javier Martinez Canillas wrote:
> > On 1/31/22 12:36, Andy Shevchenko wrote:
...
> > I actually added this dependency deliberative. It's true that the driver is using
> > the device properties API and so there isn't anything from the properties parsing
> > point of view that depends on OF. And the original driver didn't depend on OF.
> >
> > But the original driver also only would had worked with Device Trees since the
> > of_device_id table is the only one that contains the device specific data info.
> >
> > The i2c_device_id table only listed the devices supported to match, but then it
> > would only had worked with the default values that are set by the driver.
> >
> > So in practice it *does* depend on OF. I'll be happy to drop that dependency if
> > you provide an acpi_device_id table to match.
>
> The code is deceptive and you become to a wrong conclusion. No, the driver
> does NOT depend on OF as a matter of fact. The tricky part is the PRP0001
> ACPI PNP ID that allows to reuse it on ACPI-based platforms.
>
> That said, please drop OF dependency.
Side note: 72915994e028 ("video: ssd1307fb: Make use of device properties")
--
With Best Regards,
Andy Shevchenko
