[PATCH v4 0/3] rtc: s5m: driver improvements

[PATCH v4 0/3] rtc: s5m: driver improvements

[Bartosz Golaszewski]

From: Bartosz Golaszewski <bgolaszewski@baylibre.com>

Hopefully this time the kernel bot won't complain anymore...

This is a set of improvements for the rtc-s5m driver. The first patch
is new in v4: I noticed the I2C regmap is not selected by this driver's
Kconfig. Two subsequent patches were already submitted separately.

v1 -> v2:
- remove the remove() callback

v2 -> v3:
- fix an error pointed out by the build robot

v3 -> v4:
- add patch 1/3: ("rtc: s5m: select REGMAP_I2C")
- fix issues raised by the kernel bot

Bartosz Golaszewski (3):
  rtc: s5m: select REGMAP_I2C
  rtc: s5m: check the return value of s5m8767_rtc_init_reg()
  rtc: s5m: use devm_i2c_new_dummy_device()

 drivers/rtc/Kconfig   |  1 +
 drivers/rtc/rtc-s5m.c | 33 +++++++++------------------------
 2 files changed, 10 insertions(+), 24 deletions(-)

-- 
2.29.1

[PATCH v4 1/3] rtc: s5m: select REGMAP_I2C

[Bartosz Golaszewski]

From: Bartosz Golaszewski <bgolaszewski@baylibre.com>

The rtc-s5m uses the I2C regmap but doesn't select it in Kconfig so
depending on the configuration the build may fail. Fix it.

Signed-off-by: Bartosz Golaszewski <bgolaszewski@baylibre.com>
---
 drivers/rtc/Kconfig | 1 +
 1 file changed, 1 insertion(+)

diff --git a/drivers/rtc/Kconfig b/drivers/rtc/Kconfig
index 6123f9f4fbc9..e4bef40831c7 100644
--- a/drivers/rtc/Kconfig
+++ b/drivers/rtc/Kconfig
@@ -692,6 +692,7 @@ config RTC_DRV_S5M
 	tristate "Samsung S2M/S5M series"
 	depends on MFD_SEC_CORE || COMPILE_TEST
 	select REGMAP_IRQ
+	select REGMAP_I2C
 	help
 	  If you say yes here you will get support for the
 	  RTC of Samsung S2MPS14 and S5M PMIC series.
-- 
2.29.1

Re: [PATCH v4 1/3] rtc: s5m: select REGMAP_I2C

[Krzysztof Kozlowski]

On Mon, Jan 11, 2021 at 01:40:25PM +0100, Bartosz Golaszewski wrote:
> From: Bartosz Golaszewski <bgolaszewski@baylibre.com>
> 
> The rtc-s5m uses the I2C regmap but doesn't select it in Kconfig so
> depending on the configuration the build may fail. Fix it.
> 
> Signed-off-by: Bartosz Golaszewski <bgolaszewski@baylibre.com>
> ---
>  drivers/rtc/Kconfig | 1 +
>  1 file changed, 1 insertion(+)
> 

Fixes: 959df7778bbd ("rtc: Enable compile testing for Maxim and Samsung drivers")

Reviewed-by: Krzysztof Kozlowski <krzk@kernel.org>

Best regards,
Krzysztof

[PATCH v4 2/3] rtc: s5m: check the return value of s5m8767_rtc_init_reg()

[Bartosz Golaszewski]

From: Bartosz Golaszewski <bgolaszewski@baylibre.com>

This function can fail if regmap operations fail so check its return
value in probe().

Signed-off-by: Bartosz Golaszewski <bgolaszewski@baylibre.com>
---
 drivers/rtc/rtc-s5m.c | 2 ++
 1 file changed, 2 insertions(+)

diff --git a/drivers/rtc/rtc-s5m.c b/drivers/rtc/rtc-s5m.c
index eb9dde4095a9..e0011d3cf61b 100644
--- a/drivers/rtc/rtc-s5m.c
+++ b/drivers/rtc/rtc-s5m.c
@@ -791,6 +791,8 @@ static int s5m_rtc_probe(struct platform_device *pdev)
 	platform_set_drvdata(pdev, info);
 
 	ret = s5m8767_rtc_init_reg(info);
+	if (ret)
+		return ret;
 
 	device_init_wakeup(&pdev->dev, 1);
 
-- 
2.29.1

Re: [PATCH v4 2/3] rtc: s5m: check the return value of s5m8767_rtc_init_reg()

[Krzysztof Kozlowski]

On Mon, Jan 11, 2021 at 01:40:26PM +0100, Bartosz Golaszewski wrote:
> From: Bartosz Golaszewski <bgolaszewski@baylibre.com>
> 
> This function can fail if regmap operations fail so check its return
> value in probe().
> 
> Signed-off-by: Bartosz Golaszewski <bgolaszewski@baylibre.com>
> ---
>  drivers/rtc/rtc-s5m.c | 2 ++
>  1 file changed, 2 insertions(+)
> 
> diff --git a/drivers/rtc/rtc-s5m.c b/drivers/rtc/rtc-s5m.c
> index eb9dde4095a9..e0011d3cf61b 100644
> --- a/drivers/rtc/rtc-s5m.c
> +++ b/drivers/rtc/rtc-s5m.c
> @@ -791,6 +791,8 @@ static int s5m_rtc_probe(struct platform_device *pdev)
>  	platform_set_drvdata(pdev, info);
>  
>  	ret = s5m8767_rtc_init_reg(info);
> +	if (ret)
> +		return ret;

You leak I2C device.

Best regards,
Krzysztof

Re: [PATCH v4 2/3] rtc: s5m: check the return value of s5m8767_rtc_init_reg()

[Bartosz Golaszewski]

On Mon, Jan 11, 2021 at 2:35 PM Krzysztof Kozlowski <krzk@kernel.org> wrote:
>
> On Mon, Jan 11, 2021 at 01:40:26PM +0100, Bartosz Golaszewski wrote:
> > From: Bartosz Golaszewski <bgolaszewski@baylibre.com>
> >
> > This function can fail if regmap operations fail so check its return
> > value in probe().
> >
> > Signed-off-by: Bartosz Golaszewski <bgolaszewski@baylibre.com>
> > ---
> >  drivers/rtc/rtc-s5m.c | 2 ++
> >  1 file changed, 2 insertions(+)
> >
> > diff --git a/drivers/rtc/rtc-s5m.c b/drivers/rtc/rtc-s5m.c
> > index eb9dde4095a9..e0011d3cf61b 100644
> > --- a/drivers/rtc/rtc-s5m.c
> > +++ b/drivers/rtc/rtc-s5m.c
> > @@ -791,6 +791,8 @@ static int s5m_rtc_probe(struct platform_device *pdev)
> >       platform_set_drvdata(pdev, info);
> >
> >       ret = s5m8767_rtc_init_reg(info);
> > +     if (ret)
> > +             return ret;
>
> You leak I2C device.
>

Yes, the next patch fixes it but I changed the order. Actually this
can be moved after 3/3 with no conflicts when applying.

Bartosz

Re: [PATCH v4 2/3] rtc: s5m: check the return value of s5m8767_rtc_init_reg()

[Krzysztof Kozlowski]

On Mon, Jan 11, 2021 at 03:11:24PM +0100, Bartosz Golaszewski wrote:
> On Mon, Jan 11, 2021 at 2:35 PM Krzysztof Kozlowski <krzk@kernel.org> wrote:
> >
> > On Mon, Jan 11, 2021 at 01:40:26PM +0100, Bartosz Golaszewski wrote:
> > > From: Bartosz Golaszewski <bgolaszewski@baylibre.com>
> > >
> > > This function can fail if regmap operations fail so check its return
> > > value in probe().
> > >
> > > Signed-off-by: Bartosz Golaszewski <bgolaszewski@baylibre.com>
> > > ---
> > >  drivers/rtc/rtc-s5m.c | 2 ++
> > >  1 file changed, 2 insertions(+)
> > >
> > > diff --git a/drivers/rtc/rtc-s5m.c b/drivers/rtc/rtc-s5m.c
> > > index eb9dde4095a9..e0011d3cf61b 100644
> > > --- a/drivers/rtc/rtc-s5m.c
> > > +++ b/drivers/rtc/rtc-s5m.c
> > > @@ -791,6 +791,8 @@ static int s5m_rtc_probe(struct platform_device *pdev)
> > >       platform_set_drvdata(pdev, info);
> > >
> > >       ret = s5m8767_rtc_init_reg(info);
> > > +     if (ret)
> > > +             return ret;
> >
> > You leak I2C device.
> >
> 
> Yes, the next patch fixes it but I changed the order. Actually this
> can be moved after 3/3 with no conflicts when applying.

Yes, but for bisecting and any backporting (e.g. with autosel) the order
is quite important. Please resend with new order.

Best regards,
Krzysztof

[PATCH v4 3/3] rtc: s5m: use devm_i2c_new_dummy_device()

[Bartosz Golaszewski]

From: Bartosz Golaszewski <bgolaszewski@baylibre.com>

Use the managed variant of i2c_new_dummy_device() to shrink code and
remove the goto label. We can drop the remove callback now too.

Signed-off-by: Bartosz Golaszewski <bgolaszewski@baylibre.com>
---
 drivers/rtc/rtc-s5m.c | 31 +++++++------------------------
 1 file changed, 7 insertions(+), 24 deletions(-)

diff --git a/drivers/rtc/rtc-s5m.c b/drivers/rtc/rtc-s5m.c
index e0011d3cf61b..80b66f16db89 100644
--- a/drivers/rtc/rtc-s5m.c
+++ b/drivers/rtc/rtc-s5m.c
@@ -760,7 +760,8 @@ static int s5m_rtc_probe(struct platform_device *pdev)
 		return -ENODEV;
 	}
 
-	info->i2c = i2c_new_dummy_device(s5m87xx->i2c->adapter, RTC_I2C_ADDR);
+	info->i2c = devm_i2c_new_dummy_device(&pdev->dev, s5m87xx->i2c->adapter,
+					      RTC_I2C_ADDR);
 	if (IS_ERR(info->i2c)) {
 		dev_err(&pdev->dev, "Failed to allocate I2C for RTC\n");
 		return PTR_ERR(info->i2c);
@@ -771,7 +772,7 @@ static int s5m_rtc_probe(struct platform_device *pdev)
 		ret = PTR_ERR(info->regmap);
 		dev_err(&pdev->dev, "Failed to allocate RTC register map: %d\n",
 				ret);
-		goto err;
+		return ret;
 	}
 
 	info->dev = &pdev->dev;
@@ -781,10 +782,9 @@ static int s5m_rtc_probe(struct platform_device *pdev)
 	if (s5m87xx->irq_data) {
 		info->irq = regmap_irq_get_virq(s5m87xx->irq_data, alarm_irq);
 		if (info->irq <= 0) {
-			ret = -EINVAL;
 			dev_err(&pdev->dev, "Failed to get virtual IRQ %d\n",
 				alarm_irq);
-			goto err;
+			return -EINVAL;
 		}
 	}
 
@@ -799,10 +799,8 @@ static int s5m_rtc_probe(struct platform_device *pdev)
 	info->rtc_dev = devm_rtc_device_register(&pdev->dev, "s5m-rtc",
 						 &s5m_rtc_ops, THIS_MODULE);
 
-	if (IS_ERR(info->rtc_dev)) {
-		ret = PTR_ERR(info->rtc_dev);
-		goto err;
-	}
+	if (IS_ERR(info->rtc_dev))
+		return PTR_ERR(info->rtc_dev);
 
 	if (!info->irq) {
 		dev_info(&pdev->dev, "Alarm IRQ not available\n");
@@ -815,23 +813,9 @@ static int s5m_rtc_probe(struct platform_device *pdev)
 	if (ret < 0) {
 		dev_err(&pdev->dev, "Failed to request alarm IRQ: %d: %d\n",
 			info->irq, ret);
-		goto err;
+		return ret;
 	}
 
-	return 0;
-
-err:
-	i2c_unregister_device(info->i2c);
-
-	return ret;
-}
-
-static int s5m_rtc_remove(struct platform_device *pdev)
-{
-	struct s5m_rtc_info *info = platform_get_drvdata(pdev);
-
-	i2c_unregister_device(info->i2c);
-
 	return 0;
 }
 
@@ -876,7 +860,6 @@ static struct platform_driver s5m_rtc_driver = {
 		.pm	= &s5m_rtc_pm_ops,
 	},
 	.probe		= s5m_rtc_probe,
-	.remove		= s5m_rtc_remove,
 	.id_table	= s5m_rtc_id,
 };
 
-- 
2.29.1

Re: [PATCH v4 3/3] rtc: s5m: use devm_i2c_new_dummy_device()

[Krzysztof Kozlowski]

On Mon, Jan 11, 2021 at 01:40:27PM +0100, Bartosz Golaszewski wrote:
> From: Bartosz Golaszewski <bgolaszewski@baylibre.com>
> 
> Use the managed variant of i2c_new_dummy_device() to shrink code and
> remove the goto label. We can drop the remove callback now too.
> 
> Signed-off-by: Bartosz Golaszewski <bgolaszewski@baylibre.com>
> ---
>  drivers/rtc/rtc-s5m.c | 31 +++++++------------------------
>  1 file changed, 7 insertions(+), 24 deletions(-)

Reviewed-by: Krzysztof Kozlowski <krzk@kernel.org>

Best regards,
Krzysztof