On 03 July 2020 10:33, Adam Thomson wrote
> On 01 July 2020 19:24, Pierre-Louis Bossart wrote:
>
> > fix W=1 warning
> >
> > sound/soc/codecs/da7219.c:1711:36: warning: 'da7219_acpi_match'
> > defined but not used [-Wunused-const-variable=]
> > 1711 | static const struct acpi_device_id da7219_acpi_match[] = {
> > |^
> >
> > Signed-off-by: Pierre-Louis Bossart
> > ---
> > sound/soc/codecs/da7219.c | 2 ++
> > 1 file changed, 2 insertions(+)
> >
> > diff --git a/sound/soc/codecs/da7219.c b/sound/soc/codecs/da7219.c
> > index f2520a6c7875..153ea30b5a8f 100644
> > --- a/sound/soc/codecs/da7219.c
> > +++ b/sound/soc/codecs/da7219.c
> > @@ -1708,11 +1708,13 @@ static const struct of_device_id da7219_of_match[]
> =
> > {
> > };
> > MODULE_DEVICE_TABLE(of, da7219_of_match);
> >
> > +#ifdef CONFIG_ACPI
> > static const struct acpi_device_id da7219_acpi_match[] = {
> > { .id = "DLGS7219", },
> > { }
> > };
> > MODULE_DEVICE_TABLE(acpi, da7219_acpi_match);
> > +#endif
>
> I think this will break non-ACPI builds as this symbol is used in the
> declaration of 'da7219_i2c_driver', without conditional compilation
> surrounding
> it. Unless of course I'm missing something. Could we instead use
> '__maybe_unused' to avoid this warning?
Obviously a slow start to the day for my brain. You've obviously built test this
given the intention behind this is for non-ACPI builds. Will get more coffee
shortly.
Still wonder if '__maybe_unused' might be nicer as per suspend/resume functions
on platforms which don't include PM_OPS. Either way though:
Reviewed-by: Adam Thomson
>
> >
> > static enum da7219_micbias_voltage
> > da7219_fw_micbias_lvl(struct device *dev, u32 val)
> > --
> > 2.25.1
