On 08.07.22 07:53, Ahmad Fatoum wrote: > Regulator core already handles always-on property, so remove handling in > fixed-regulator driver. For this to have no functional change, all > regulator uses must be balanced as regulator code implements always-on > by initializing use count to baseline of 1, while fixed-regulator > prevents disabling altogether. As mismatched regulator enable/disable > is a bug in need of fixing anyway, this is tolerated.
Rethinking it, I guess we should just move this into regulator core instead. That's what Linux does as well. > > Signed-off-by: Ahmad Fatoum <[email protected]> > --- > drivers/regulator/fixed.c | 10 ---------- > 1 file changed, 10 deletions(-) > > diff --git a/drivers/regulator/fixed.c b/drivers/regulator/fixed.c > index c5f97d78df40..ec64f39b86c7 100644 > --- a/drivers/regulator/fixed.c > +++ b/drivers/regulator/fixed.c > @@ -15,7 +15,6 @@ > > struct regulator_fixed { > int gpio; > - int always_on; > struct regulator_dev rdev; > struct regulator_desc rdesc; > }; > @@ -34,9 +33,6 @@ static int regulator_fixed_disable(struct regulator_dev > *rdev) > { > struct regulator_fixed *fix = container_of(rdev, struct > regulator_fixed, rdev); > > - if (fix->always_on) > - return 0; > - > if (!gpio_is_valid(fix->gpio)) > return 0; > > @@ -76,12 +72,6 @@ static int regulator_fixed_probe(struct device_d *dev) > if (!of_property_read_u32(np, "off-on-delay-us", &delay)) > fix->rdesc.off_on_delay = delay; > > - if (of_find_property(np, "regulator-always-on", NULL) || > - of_find_property(np, "regulator-boot-on", NULL)) { > - fix->always_on = 1; > - regulator_fixed_enable(&fix->rdev); > - } > - > ret = of_regulator_register(&fix->rdev, np); > if (ret) > goto err; -- Pengutronix e.K. | | Steuerwalder Str. 21 | http://www.pengutronix.de/ | 31137 Hildesheim, Germany | Phone: +49-5121-206917-0 | Amtsgericht Hildesheim, HRA 2686 | Fax: +49-5121-206917-5555 |
