On Tue, Nov 29, 2016 at 11:55:47AM +0100, Jacek Anaszewski wrote:
> On 11/29/2016 11:21 AM, Tin Huynh wrote:
> > This patch enables ACPI support for leds-pca963x driver.
> >
> > Signed-off-by: Tin Huynh <tnhu...@apm.com>
> > ---
> > drivers/leds/leds-pca963x.c | 25 ++++++++++++++++++++++++-
> > 1 files changed, 24 insertions(+), 1 deletions(-)
> >
> > Change from V1:
> > -Add CONFIG_ACPI.
> >
> > diff --git a/drivers/leds/leds-pca963x.c b/drivers/leds/leds-pca963x.c
> > index 407eba1..57f11e3 100644
> > --- a/drivers/leds/leds-pca963x.c
> > +++ b/drivers/leds/leds-pca963x.c
> > @@ -25,6 +25,7 @@
> > * or by adding the 'nxp,hw-blink' property to the DTS.
> > */
> >
> > +#include <linux/acpi.h>
> > #include <linux/module.h>
> > #include <linux/delay.h>
> > #include <linux/string.h>
> > @@ -95,6 +96,17 @@ struct pca963x_chipdef {
> > };
> > MODULE_DEVICE_TABLE(i2c, pca963x_id);
> >
> > +#ifdef CONFIG_ACPI
> > +static const struct acpi_device_id pca963x_acpi_ids[] = {
> > + { "PCA9632", pca9633 },
> > + { "PCA9633", pca9633 },
> > + { "PCA9634", pca9634 },
> > + { "PCA9635", pca9635 },
> > + { }
> > +};
> > +MODULE_DEVICE_TABLE(acpi, pca963x_acpi_ids);
> > +#endif
> > +
> > struct pca963x_led;
> >
> > struct pca963x {
> > @@ -322,7 +334,17 @@ static int pca963x_probe(struct i2c_client *client,
> > struct pca963x_chipdef *chip;
> > int i, err;
> >
> > - chip = &pca963x_chipdefs[id->driver_data];
> > + if (id) {
> > + chip = &pca963x_chipdefs[id->driver_data];
> > + } else {
> > + const struct acpi_device_id *acpi_id;
> > +
> > + acpi_id = acpi_match_device(ACPI_PTR(pca963x_acpi_ids),
> > + &client->dev);
>
> What kind of problem did you get while compiling without ACPI_PTR here,
> when CONFIG_ACPI is disabled? I also tried this configuration but
> nothing wrong happened. Also at first glance I don't see why lack of
> ACPI_PTR macro could cause problems.
>
> Grep also doesn't show any call to acpi_match_device
> with ACPI_PTR as the first argument in the existing drivers.
Indeed, that is not needed at all.
