On Wed, Nov 18, 2020 at 12:09 PM Lee Jones <lee.jones@xxxxxxxxxx> wrote: > On Wed, 18 Nov 2020, Andy Shevchenko wrote: > > On Wed, Nov 18, 2020 at 11:56 AM Andy Shevchenko > > <andy.shevchenko@xxxxxxxxx> wrote: > > > On Wed, Nov 18, 2020 at 11:54 AM Uwe Kleine-König > > > <u.kleine-koenig@xxxxxxxxxxxxxx> wrote: > > > > On Tue, Nov 17, 2020 at 06:45:37PM +0200, Andy Shevchenko wrote: > > > > ... > > > > > > So a system without CONFIG_OF might still make use of .of_match_table? > > > > > > Yep! > > > > If you are curious: > > https://elixir.bootlin.com/linux/latest/source/drivers/acpi/bus.c#L615 > > This still doesn't sit quite well with me. > > Not the process of ACPI using OF artifacts or the possibilty to embed > OF type structures into ACPI tables, but the fact that an OF-only user > (or so they think) should have to care about that special use-case. OF-only users shouldn't take care about this. If you are sure that the driver may not be used on a non-OF platform. > What if we make ACPI select OF? Or handle the ACPI case explicitly in > the of_match_ptr() declaration? I guess you are a bit late, but you always may propose patches to ACPI maintainers. -- With Best Regards, Andy Shevchenko
