[PATCH] power: supply: bq25630: Release battery information after setup

[PATCH] power: supply: bq25630: Release battery information after setup

[Linmao Li]

data->batinfo is only used by bq25630_setup() to program the initial
charge limits, but power_supply_get_battery_info() allocates it on
psy->dev, so it stays around for the lifetime of the device. Nothing
else in the driver uses it. Release it right after setup, as bq256xx
and bq257xx do.

Signed-off-by: Linmao Li <lilinmao@kylinos.cn>
---
 drivers/power/supply/bq25630_charger.c | 5 ++---
 1 file changed, 2 insertions(+), 3 deletions(-)

diff --git a/drivers/power/supply/bq25630_charger.c b/drivers/power/supply/bq25630_charger.c
index 165f8c67b4895..efcdb1f5bc351 100644
--- a/drivers/power/supply/bq25630_charger.c
+++ b/drivers/power/supply/bq25630_charger.c
@@ -1045,10 +1045,9 @@ static int bq25630_probe(struct i2c_client *client)
 		return dev_err_probe(data->dev, ret, "Could not request IRQ\n");
 
 	ret = bq25630_setup(data);
-	if (ret)
-		return ret;
+	power_supply_put_battery_info(data->psy, data->batinfo);
 
-	return 0;
+	return ret;
 }
 
 static const struct of_device_id bq25630_of_match[] = {
-- 
2.25.1

Re: [PATCH] power: supply: bq25630: Release battery information after setup

[Waqar Hameed]

On Mon, Jul 27, 2026 at 17:37 +0800 Linmao Li <lilinmao@kylinos.cn> wrote:

> data->batinfo is only used by bq25630_setup() to program the initial
> charge limits, but power_supply_get_battery_info() allocates it on
> psy->dev, so it stays around for the lifetime of the device. Nothing
> else in the driver uses it. Release it right after setup, as bq256xx
> and bq257xx do.
>
> Signed-off-by: Linmao Li <lilinmao@kylinos.cn>
> ---
>  drivers/power/supply/bq25630_charger.c | 5 ++---
>  1 file changed, 2 insertions(+), 3 deletions(-)
>
> diff --git a/drivers/power/supply/bq25630_charger.c b/drivers/power/supply/bq25630_charger.c
> index 165f8c67b4895..efcdb1f5bc351 100644
> --- a/drivers/power/supply/bq25630_charger.c
> +++ b/drivers/power/supply/bq25630_charger.c
> @@ -1045,10 +1045,9 @@ static int bq25630_probe(struct i2c_client *client)
>  		return dev_err_probe(data->dev, ret, "Could not request IRQ\n");
>  
>  	ret = bq25630_setup(data);
> -	if (ret)
> -		return ret;
> +	power_supply_put_battery_info(data->psy, data->batinfo);
>  
> -	return 0;
> +	return ret;
>  }
>  
>  static const struct of_device_id bq25630_of_match[] = {

The initial thought was that `batinfo` could/would be used elsewhere,
but as you point out, it's currently not. A better design is then to
move the `power_supply_get_battery_info()` into `bq25630_setup()`, in
order to scope the `get/put()`. So let's get `batinfo` there, just
before using it, and have an error/clean-up-path there as well.

Re: [PATCH] power: supply: bq25630: Release battery information after setup

[Sebastian Reichel]

[-- Attachment #1: Type: text/plain, Size: 1950 bytes --]

Hi,

On Mon, Jul 27, 2026 at 02:43:37PM +0200, Waqar Hameed wrote:
> On Mon, Jul 27, 2026 at 17:37 +0800 Linmao Li <lilinmao@kylinos.cn> wrote:
> 
> > data->batinfo is only used by bq25630_setup() to program the initial
> > charge limits, but power_supply_get_battery_info() allocates it on
> > psy->dev, so it stays around for the lifetime of the device. Nothing
> > else in the driver uses it. Release it right after setup, as bq256xx
> > and bq257xx do.
> >
> > Signed-off-by: Linmao Li <lilinmao@kylinos.cn>
> > ---
> >  drivers/power/supply/bq25630_charger.c | 5 ++---
> >  1 file changed, 2 insertions(+), 3 deletions(-)
> >
> > diff --git a/drivers/power/supply/bq25630_charger.c b/drivers/power/supply/bq25630_charger.c
> > index 165f8c67b4895..efcdb1f5bc351 100644
> > --- a/drivers/power/supply/bq25630_charger.c
> > +++ b/drivers/power/supply/bq25630_charger.c
> > @@ -1045,10 +1045,9 @@ static int bq25630_probe(struct i2c_client *client)
> >  		return dev_err_probe(data->dev, ret, "Could not request IRQ\n");
> >  
> >  	ret = bq25630_setup(data);
> > -	if (ret)
> > -		return ret;
> > +	power_supply_put_battery_info(data->psy, data->batinfo);
> >  
> > -	return 0;
> > +	return ret;
> >  }
> >  
> >  static const struct of_device_id bq25630_of_match[] = {
> 
> The initial thought was that `batinfo` could/would be used elsewhere,
> but as you point out, it's currently not. A better design is then to
> move the `power_supply_get_battery_info()` into `bq25630_setup()`, in
> order to scope the `get/put()`. So let's get `batinfo` there, just
> before using it, and have an error/clean-up-path there as well.

Agreed. Keeping an invalid pointer makes no sense.

The setup routine is also a good candidate for the new
.init callback now available in struct power_supply_desc,
so that the registers are initialized before the device
is exposed to the system.

Greetings,

-- Sebastian

[-- Attachment #2: signature.asc --]
[-- Type: application/pgp-signature, Size: 833 bytes --]