[PATCH v4 0/6] thermal: scope/cleanup.h improvements

[PATCH v4 0/6] thermal: scope/cleanup.h improvements

[Krzysztof Kozlowski]

Changes in v4:
- Patch 2: rewrite, significant change: kzalloc() also with
  scoped-handling so the entire error handling could be removed.
  Due to above, drop review-tags (Chen-Yu, Jonathan).
- Add Rb tags for other patches.
- Link to v3: https://lore.kernel.org/r/20241008-b4-cleanup-h-of-node-put-thermal-v3-0-825122398f71@linaro.org

Changes in v3:
- Rebase, because there was bigger rework in thermal code.
  This made two patches obsolete, but brought new one:
  1/6: thermal: of: Simplify thermal_of_should_bind with scoped for each OF child
- Link to v2: https://lore.kernel.org/r/20240816-b4-cleanup-h-of-node-put-thermal-v2-0-cee9fc490478@linaro.org

Changes in v2:
- Drop left-over of_node_put in regular exit path (Chen-Yu)
- Link to v1: https://lore.kernel.org/r/20240814-b4-cleanup-h-of-node-put-thermal-v1-0-7a1381e1627e@linaro.org

Few code simplifications with scope/cleanup.h.

Best regards,
Krzysztof

---
Krzysztof Kozlowski (6):
      thermal: of: Simplify thermal_of_should_bind with scoped for each OF child
      thermal: of: Use scoped memory and OF handling to simplify thermal_of_trips_init()
      thermal: of: Use scoped device node handling to simplify of_thermal_zone_find()
      thermal: qcom-spmi-adc-tm5: Simplify with scoped for each OF child loop
      thermal: tegra: Simplify with scoped for each OF child loop
      thermal: sun8i: Use scoped device node handling to simplify error paths

 drivers/thermal/qcom/qcom-spmi-adc-tm5.c |  7 ++---
 drivers/thermal/sun8i_thermal.c          | 11 +++----
 drivers/thermal/tegra/soctherm.c         |  5 ++-
 drivers/thermal/thermal_of.c             | 54 ++++++++++----------------------
 4 files changed, 25 insertions(+), 52 deletions(-)
---
base-commit: 33ce24234fca4c083e6685a18b460a18ebb5d5c1
change-id: 20240814-b4-cleanup-h-of-node-put-thermal-2268440cc6f7

Best regards,
-- 
Krzysztof Kozlowski <krzysztof.kozlowski@linaro.org>

[PATCH v4 1/6] thermal: of: Simplify thermal_of_should_bind with scoped for each OF child

[Krzysztof Kozlowski]

Use scoped for_each_child_of_node_scoped() when iterating over device
nodes to make code a bit simpler.

Reviewed-by: Jonathan Cameron <Jonathan.Cameron@huawei.com>
Signed-off-by: Krzysztof Kozlowski <krzysztof.kozlowski@linaro.org>
---

Changes in v3:
1. New patch
---
 drivers/thermal/thermal_of.c | 5 ++---
 1 file changed, 2 insertions(+), 3 deletions(-)

diff --git a/drivers/thermal/thermal_of.c b/drivers/thermal/thermal_of.c
index a4caf7899f8e4a5f3422e4c02c66502b5050fd6d..f0ffc0e335ba9406f4fd858d6c561f9d23f4b842 100644
--- a/drivers/thermal/thermal_of.c
+++ b/drivers/thermal/thermal_of.c
@@ -301,7 +301,7 @@ static bool thermal_of_should_bind(struct thermal_zone_device *tz,
 				   struct thermal_cooling_device *cdev,
 				   struct cooling_spec *c)
 {
-	struct device_node *tz_np, *cm_np, *child;
+	struct device_node *tz_np, *cm_np;
 	bool result = false;
 
 	tz_np = thermal_of_zone_get_by_name(tz);
@@ -315,7 +315,7 @@ static bool thermal_of_should_bind(struct thermal_zone_device *tz,
 		goto out;
 
 	/* Look up the trip and the cdev in the cooling maps. */
-	for_each_child_of_node(cm_np, child) {
+	for_each_child_of_node_scoped(cm_np, child) {
 		struct device_node *tr_np;
 		int count, i;
 
@@ -334,7 +334,6 @@ static bool thermal_of_should_bind(struct thermal_zone_device *tz,
 				break;
 		}
 
-		of_node_put(child);
 		break;
 	}
 

-- 
2.43.0

[PATCH v4 2/6] thermal: of: Use scoped memory and OF handling to simplify thermal_of_trips_init()

[Krzysztof Kozlowski]

Obtain the device node reference and allocate memory with
scoped/cleanup.h to reduce error handling and make the code a bit
simpler.

The code is not equivalent in one minor aspect: outgoing parameter
"*ntrips" will not be zeroed on errors of memory allocation.  This
difference is not important, because code was already not zeroing it in
case of earlier errors and the only caller does not rely on ntrips being
0 in case of errors.

Signed-off-by: Krzysztof Kozlowski <krzysztof.kozlowski@linaro.org>
---

Cc: Jonathan Cameron <Jonathan.Cameron@huawei.com>
Cc: Chen-Yu Tsai <wenst@chromium.org>

Changes in v4:
1. Significant change: kzalloc() also with scoped-handling so the entire
   error handling could be removed.
2. Due to above, drop review-tags (Chen-Yu, Jonathan).

Changes in v2:
1. Drop left-over of_node_put in regular exit path (Chen-Yu)
---
 drivers/thermal/thermal_of.c | 31 ++++++++-----------------------
 1 file changed, 8 insertions(+), 23 deletions(-)

diff --git a/drivers/thermal/thermal_of.c b/drivers/thermal/thermal_of.c
index f0ffc0e335ba9406f4fd858d6c561f9d23f4b842..37db435b54b124abf25b1d75d6cc4fb75f1c1e5c 100644
--- a/drivers/thermal/thermal_of.c
+++ b/drivers/thermal/thermal_of.c
@@ -95,11 +95,9 @@ static int thermal_of_populate_trip(struct device_node *np,
 
 static struct thermal_trip *thermal_of_trips_init(struct device_node *np, int *ntrips)
 {
-	struct thermal_trip *tt;
-	struct device_node *trips;
 	int ret, count;
 
-	trips = of_get_child_by_name(np, "trips");
+	struct device_node *trips __free(device_node) = of_get_child_by_name(np, "trips");
 	if (!trips) {
 		pr_err("Failed to find 'trips' node\n");
 		return ERR_PTR(-EINVAL);
@@ -108,36 +106,23 @@ static struct thermal_trip *thermal_of_trips_init(struct device_node *np, int *n
 	count = of_get_child_count(trips);
 	if (!count) {
 		pr_err("No trip point defined\n");
-		ret = -EINVAL;
-		goto out_of_node_put;
+		return ERR_PTR(-EINVAL);
 	}
 
-	tt = kzalloc(sizeof(*tt) * count, GFP_KERNEL);
-	if (!tt) {
-		ret = -ENOMEM;
-		goto out_of_node_put;
-	}
-
-	*ntrips = count;
+	struct thermal_trip *tt __free(kfree) = kzalloc(sizeof(*tt) * count, GFP_KERNEL);
+	if (!tt)
+		return ERR_PTR(-ENOMEM);
 
 	count = 0;
 	for_each_child_of_node_scoped(trips, trip) {
 		ret = thermal_of_populate_trip(trip, &tt[count++]);
 		if (ret)
-			goto out_kfree;
+			return ERR_PTR(ret);
 	}
 
-	of_node_put(trips);
+	*ntrips = count;
 
-	return tt;
-
-out_kfree:
-	kfree(tt);
-	*ntrips = 0;
-out_of_node_put:
-	of_node_put(trips);
-
-	return ERR_PTR(ret);
+	return no_free_ptr(tt);
 }
 
 static struct device_node *of_thermal_zone_find(struct device_node *sensor, int id)

-- 
2.43.0

[PATCH v4 3/6] thermal: of: Use scoped device node handling to simplify of_thermal_zone_find()

[Krzysztof Kozlowski]

Obtain the device node reference with scoped/cleanup.h to reduce error
handling and make the code a bit simpler.

Reviewed-by: Chen-Yu Tsai <wenst@chromium.org>
Reviewed-by: Jonathan Cameron <Jonathan.Cameron@huawei.com>
Signed-off-by: Krzysztof Kozlowski <krzysztof.kozlowski@linaro.org>
---
 drivers/thermal/thermal_of.c | 18 ++++++------------
 1 file changed, 6 insertions(+), 12 deletions(-)

diff --git a/drivers/thermal/thermal_of.c b/drivers/thermal/thermal_of.c
index 37db435b54b124abf25b1d75d6cc4fb75f1c1e5c..66514120f460940ed7cfba791140c958b99e056e 100644
--- a/drivers/thermal/thermal_of.c
+++ b/drivers/thermal/thermal_of.c
@@ -127,10 +127,9 @@ static struct thermal_trip *thermal_of_trips_init(struct device_node *np, int *n
 
 static struct device_node *of_thermal_zone_find(struct device_node *sensor, int id)
 {
-	struct device_node *np, *tz;
 	struct of_phandle_args sensor_specs;
 
-	np = of_find_node_by_name(NULL, "thermal-zones");
+	struct device_node *np __free(device_node) = of_find_node_by_name(NULL, "thermal-zones");
 	if (!np) {
 		pr_debug("No thermal zones description\n");
 		return ERR_PTR(-ENODEV);
@@ -148,8 +147,7 @@ static struct device_node *of_thermal_zone_find(struct device_node *sensor, int
 						   "#thermal-sensor-cells");
 		if (count <= 0) {
 			pr_err("%pOFn: missing thermal sensor\n", child);
-			tz = ERR_PTR(-EINVAL);
-			goto out;
+			return ERR_PTR(-EINVAL);
 		}
 
 		for (i = 0; i < count; i++) {
@@ -161,22 +159,18 @@ static struct device_node *of_thermal_zone_find(struct device_node *sensor, int
 							 i, &sensor_specs);
 			if (ret < 0) {
 				pr_err("%pOFn: Failed to read thermal-sensors cells: %d\n", child, ret);
-				tz = ERR_PTR(ret);
-				goto out;
+				return ERR_PTR(ret);
 			}
 
 			if ((sensor == sensor_specs.np) && id == (sensor_specs.args_count ?
 								  sensor_specs.args[0] : 0)) {
 				pr_debug("sensor %pOFn id=%d belongs to %pOFn\n", sensor, id, child);
-				tz = no_free_ptr(child);
-				goto out;
+				return no_free_ptr(child);
 			}
 		}
 	}
-	tz = ERR_PTR(-ENODEV);
-out:
-	of_node_put(np);
-	return tz;
+
+	return ERR_PTR(-ENODEV);
 }
 
 static int thermal_of_monitor_init(struct device_node *np, int *delay, int *pdelay)

-- 
2.43.0

[PATCH v4 4/6] thermal: qcom-spmi-adc-tm5: Simplify with scoped for each OF child loop

[Krzysztof Kozlowski]

Use scoped for_each_available_child_of_node_scoped() when iterating over
device nodes to make code a bit simpler.

Reviewed-by: Chen-Yu Tsai <wenst@chromium.org>
Reviewed-by: Jonathan Cameron <Jonathan.Cameron@huawei.com>
Signed-off-by: Krzysztof Kozlowski <krzysztof.kozlowski@linaro.org>
---
 drivers/thermal/qcom/qcom-spmi-adc-tm5.c | 7 ++-----
 1 file changed, 2 insertions(+), 5 deletions(-)

diff --git a/drivers/thermal/qcom/qcom-spmi-adc-tm5.c b/drivers/thermal/qcom/qcom-spmi-adc-tm5.c
index 5e94a45eba3eef65e436a01b3aa58aabf5f706c2..d7f2e6ca92c2c9fb532ca5d4bcc22b6c2cddbdb0 100644
--- a/drivers/thermal/qcom/qcom-spmi-adc-tm5.c
+++ b/drivers/thermal/qcom/qcom-spmi-adc-tm5.c
@@ -938,7 +938,6 @@ static const struct adc_tm5_data adc_tm5_gen2_data_pmic = {
 static int adc_tm5_get_dt_data(struct adc_tm5_chip *adc_tm, struct device_node *node)
 {
 	struct adc_tm5_channel *channels;
-	struct device_node *child;
 	u32 value;
 	int ret;
 	struct device *dev = adc_tm->dev;
@@ -982,12 +981,10 @@ static int adc_tm5_get_dt_data(struct adc_tm5_chip *adc_tm, struct device_node *
 		adc_tm->avg_samples = VADC_DEF_AVG_SAMPLES;
 	}
 
-	for_each_available_child_of_node(node, child) {
+	for_each_available_child_of_node_scoped(node, child) {
 		ret = adc_tm5_get_dt_channel_data(adc_tm, channels, child);
-		if (ret) {
-			of_node_put(child);
+		if (ret)
 			return ret;
-		}
 
 		channels++;
 	}

-- 
2.43.0

[PATCH v4 5/6] thermal: tegra: Simplify with scoped for each OF child loop

[Krzysztof Kozlowski]

Use scoped for_each_child_of_node_scoped() when iterating over device
nodes to make code a bit simpler.

Reviewed-by: Chen-Yu Tsai <wenst@chromium.org>
Reviewed-by: Jonathan Cameron <Jonathan.Cameron@huawei.com>
Signed-off-by: Krzysztof Kozlowski <krzysztof.kozlowski@linaro.org>
---
 drivers/thermal/tegra/soctherm.c | 5 ++---
 1 file changed, 2 insertions(+), 3 deletions(-)

diff --git a/drivers/thermal/tegra/soctherm.c b/drivers/thermal/tegra/soctherm.c
index a023c948afbddd6d9fc1a00b2fe310c71147bda9..ff4eedb553fb6e78c94b714dd179193c244b6589 100644
--- a/drivers/thermal/tegra/soctherm.c
+++ b/drivers/thermal/tegra/soctherm.c
@@ -1651,7 +1651,7 @@ static void soctherm_init_hw_throt_cdev(struct platform_device *pdev)
 {
 	struct device *dev = &pdev->dev;
 	struct tegra_soctherm *ts = dev_get_drvdata(dev);
-	struct device_node *np_stc, *np_stcc;
+	struct device_node *np_stc;
 	const char *name;
 	int i;
 
@@ -1668,7 +1668,7 @@ static void soctherm_init_hw_throt_cdev(struct platform_device *pdev)
 		return;
 	}
 
-	for_each_child_of_node(np_stc, np_stcc) {
+	for_each_child_of_node_scoped(np_stc, np_stcc) {
 		struct soctherm_throt_cfg *stc;
 		struct thermal_cooling_device *tcd;
 		int err;
@@ -1683,7 +1683,6 @@ static void soctherm_init_hw_throt_cdev(struct platform_device *pdev)
 
 		if (stc->init) {
 			dev_err(dev, "throttle-cfg: %s: redefined!\n", name);
-			of_node_put(np_stcc);
 			break;
 		}
 

-- 
2.43.0

[PATCH v4 6/6] thermal: sun8i: Use scoped device node handling to simplify error paths

[Krzysztof Kozlowski]

Obtain the device node reference with scoped/cleanup.h to reduce error
handling and make the code a bit simpler.

Reviewed-by: Chen-Yu Tsai <wenst@chromium.org>
Reviewed-by: Jonathan Cameron <Jonathan.Cameron@huawei.com>
Signed-off-by: Krzysztof Kozlowski <krzysztof.kozlowski@linaro.org>
---
 drivers/thermal/sun8i_thermal.c | 11 +++++------
 1 file changed, 5 insertions(+), 6 deletions(-)

diff --git a/drivers/thermal/sun8i_thermal.c b/drivers/thermal/sun8i_thermal.c
index 3203d8bd13a8fc2a9e5a59b3547cefc2440542c6..22674790629a7b549d1ce618998ff51f6553613e 100644
--- a/drivers/thermal/sun8i_thermal.c
+++ b/drivers/thermal/sun8i_thermal.c
@@ -9,6 +9,7 @@
  */
 
 #include <linux/bitmap.h>
+#include <linux/cleanup.h>
 #include <linux/clk.h>
 #include <linux/device.h>
 #include <linux/interrupt.h>
@@ -348,19 +349,18 @@ static void sun8i_ths_reset_control_assert(void *data)
 
 static struct regmap *sun8i_ths_get_sram_regmap(struct device_node *node)
 {
-	struct device_node *sram_node;
 	struct platform_device *sram_pdev;
 	struct regmap *regmap = NULL;
 
-	sram_node = of_parse_phandle(node, "allwinner,sram", 0);
+	struct device_node *sram_node __free(device_node) =
+		of_parse_phandle(node, "allwinner,sram", 0);
 	if (!sram_node)
 		return ERR_PTR(-ENODEV);
 
 	sram_pdev = of_find_device_by_node(sram_node);
 	if (!sram_pdev) {
 		/* platform device might not be probed yet */
-		regmap = ERR_PTR(-EPROBE_DEFER);
-		goto out_put_node;
+		return ERR_PTR(-EPROBE_DEFER);
 	}
 
 	/* If no regmap is found then the other device driver is at fault */
@@ -369,8 +369,7 @@ static struct regmap *sun8i_ths_get_sram_regmap(struct device_node *node)
 		regmap = ERR_PTR(-EINVAL);
 
 	platform_device_put(sram_pdev);
-out_put_node:
-	of_node_put(sram_node);
+
 	return regmap;
 }
 

-- 
2.43.0

Re: [PATCH v4 2/6] thermal: of: Use scoped memory and OF handling to simplify thermal_of_trips_init()

[Chen-Yu Tsai]

On Fri, Oct 11, 2024 at 2:06 AM Krzysztof Kozlowski
<krzysztof.kozlowski@linaro.org> wrote:
>
> Obtain the device node reference and allocate memory with
> scoped/cleanup.h to reduce error handling and make the code a bit
> simpler.
>
> The code is not equivalent in one minor aspect: outgoing parameter
> "*ntrips" will not be zeroed on errors of memory allocation.  This
> difference is not important, because code was already not zeroing it in
> case of earlier errors and the only caller does not rely on ntrips being
> 0 in case of errors.
>
> Signed-off-by: Krzysztof Kozlowski <krzysztof.kozlowski@linaro.org>
> ---
>
> Cc: Jonathan Cameron <Jonathan.Cameron@huawei.com>
> Cc: Chen-Yu Tsai <wenst@chromium.org>
>
> Changes in v4:
> 1. Significant change: kzalloc() also with scoped-handling so the entire
>    error handling could be removed.
> 2. Due to above, drop review-tags (Chen-Yu, Jonathan).

The additional changes are the same as what I had done, except that
I used "return_ptr(tt)" instead of "return no_free_ptr(tt)", and I
had reset *ntrips to 0 at the beginning.

So,

Reviewed-by: Chen-Yu Tsai <wenst@chromium.org>

> Changes in v2:
> 1. Drop left-over of_node_put in regular exit path (Chen-Yu)
> ---
>  drivers/thermal/thermal_of.c | 31 ++++++++-----------------------
>  1 file changed, 8 insertions(+), 23 deletions(-)
>
> diff --git a/drivers/thermal/thermal_of.c b/drivers/thermal/thermal_of.c
> index f0ffc0e335ba9406f4fd858d6c561f9d23f4b842..37db435b54b124abf25b1d75d6cc4fb75f1c1e5c 100644
> --- a/drivers/thermal/thermal_of.c
> +++ b/drivers/thermal/thermal_of.c
> @@ -95,11 +95,9 @@ static int thermal_of_populate_trip(struct device_node *np,
>
>  static struct thermal_trip *thermal_of_trips_init(struct device_node *np, int *ntrips)
>  {
> -       struct thermal_trip *tt;
> -       struct device_node *trips;
>         int ret, count;
>
> -       trips = of_get_child_by_name(np, "trips");
> +       struct device_node *trips __free(device_node) = of_get_child_by_name(np, "trips");
>         if (!trips) {
>                 pr_err("Failed to find 'trips' node\n");
>                 return ERR_PTR(-EINVAL);
> @@ -108,36 +106,23 @@ static struct thermal_trip *thermal_of_trips_init(struct device_node *np, int *n
>         count = of_get_child_count(trips);
>         if (!count) {
>                 pr_err("No trip point defined\n");
> -               ret = -EINVAL;
> -               goto out_of_node_put;
> +               return ERR_PTR(-EINVAL);
>         }
>
> -       tt = kzalloc(sizeof(*tt) * count, GFP_KERNEL);
> -       if (!tt) {
> -               ret = -ENOMEM;
> -               goto out_of_node_put;
> -       }
> -
> -       *ntrips = count;
> +       struct thermal_trip *tt __free(kfree) = kzalloc(sizeof(*tt) * count, GFP_KERNEL);
> +       if (!tt)
> +               return ERR_PTR(-ENOMEM);
>
>         count = 0;
>         for_each_child_of_node_scoped(trips, trip) {
>                 ret = thermal_of_populate_trip(trip, &tt[count++]);
>                 if (ret)
> -                       goto out_kfree;
> +                       return ERR_PTR(ret);
>         }
>
> -       of_node_put(trips);
> +       *ntrips = count;
>
> -       return tt;
> -
> -out_kfree:
> -       kfree(tt);
> -       *ntrips = 0;
> -out_of_node_put:
> -       of_node_put(trips);
> -
> -       return ERR_PTR(ret);
> +       return no_free_ptr(tt);
>  }
>
>  static struct device_node *of_thermal_zone_find(struct device_node *sensor, int id)
>
> --
> 2.43.0
>

Re: [PATCH v4 0/6] thermal: scope/cleanup.h improvements

[Chen-Yu Tsai]

On Fri, Oct 11, 2024 at 2:06 AM Krzysztof Kozlowski
<krzysztof.kozlowski@linaro.org> wrote:
>
> Changes in v4:
> - Patch 2: rewrite, significant change: kzalloc() also with
>   scoped-handling so the entire error handling could be removed.
>   Due to above, drop review-tags (Chen-Yu, Jonathan).
> - Add Rb tags for other patches.
> - Link to v3: https://lore.kernel.org/r/20241008-b4-cleanup-h-of-node-put-thermal-v3-0-825122398f71@linaro.org
>
> Changes in v3:
> - Rebase, because there was bigger rework in thermal code.
>   This made two patches obsolete, but brought new one:
>   1/6: thermal: of: Simplify thermal_of_should_bind with scoped for each OF child
> - Link to v2: https://lore.kernel.org/r/20240816-b4-cleanup-h-of-node-put-thermal-v2-0-cee9fc490478@linaro.org
>
> Changes in v2:
> - Drop left-over of_node_put in regular exit path (Chen-Yu)
> - Link to v1: https://lore.kernel.org/r/20240814-b4-cleanup-h-of-node-put-thermal-v1-0-7a1381e1627e@linaro.org
>
> Few code simplifications with scope/cleanup.h.
>
> Best regards,
> Krzysztof
>
> ---
> Krzysztof Kozlowski (6):
>       thermal: of: Simplify thermal_of_should_bind with scoped for each OF child

I couldn't find this in my inbox. But since I already reviewed all the other
patches, and I looked at this one on lore, consider the whole series is now

Reviewed-by: Chen-Yu Tsai <wenst@chromium.org>

>       thermal: of: Use scoped memory and OF handling to simplify thermal_of_trips_init()
>       thermal: of: Use scoped device node handling to simplify of_thermal_zone_find()
>       thermal: qcom-spmi-adc-tm5: Simplify with scoped for each OF child loop
>       thermal: tegra: Simplify with scoped for each OF child loop
>       thermal: sun8i: Use scoped device node handling to simplify error paths
>
>  drivers/thermal/qcom/qcom-spmi-adc-tm5.c |  7 ++---
>  drivers/thermal/sun8i_thermal.c          | 11 +++----
>  drivers/thermal/tegra/soctherm.c         |  5 ++-
>  drivers/thermal/thermal_of.c             | 54 ++++++++++----------------------
>  4 files changed, 25 insertions(+), 52 deletions(-)
> ---
> base-commit: 33ce24234fca4c083e6685a18b460a18ebb5d5c1
> change-id: 20240814-b4-cleanup-h-of-node-put-thermal-2268440cc6f7
>
> Best regards,
> --
> Krzysztof Kozlowski <krzysztof.kozlowski@linaro.org>
>

Re: [PATCH v4 0/6] thermal: scope/cleanup.h improvements

[Krzysztof Kozlowski]

On 14/10/2024 10:32, Chen-Yu Tsai wrote:
> On Fri, Oct 11, 2024 at 2:06 AM Krzysztof Kozlowski
> <krzysztof.kozlowski@linaro.org> wrote:
>>
>> Changes in v4:
>> - Patch 2: rewrite, significant change: kzalloc() also with
>>   scoped-handling so the entire error handling could be removed.
>>   Due to above, drop review-tags (Chen-Yu, Jonathan).
>> - Add Rb tags for other patches.
>> - Link to v3: https://lore.kernel.org/r/20241008-b4-cleanup-h-of-node-put-thermal-v3-0-825122398f71@linaro.org
>>
>> Changes in v3:
>> - Rebase, because there was bigger rework in thermal code.
>>   This made two patches obsolete, but brought new one:
>>   1/6: thermal: of: Simplify thermal_of_should_bind with scoped for each OF child
>> - Link to v2: https://lore.kernel.org/r/20240816-b4-cleanup-h-of-node-put-thermal-v2-0-cee9fc490478@linaro.org
>>
>> Changes in v2:
>> - Drop left-over of_node_put in regular exit path (Chen-Yu)
>> - Link to v1: https://lore.kernel.org/r/20240814-b4-cleanup-h-of-node-put-thermal-v1-0-7a1381e1627e@linaro.org
>>
>> Few code simplifications with scope/cleanup.h.
>>
>> Best regards,
>> Krzysztof
>>
>> ---
>> Krzysztof Kozlowski (6):
>>       thermal: of: Simplify thermal_of_should_bind with scoped for each OF child
> 
> I couldn't find this in my inbox. But since I already reviewed all the other
> patches, and I looked at this one on lore, consider the whole series is now

Sorry for that. Your wens@csie.org was cc-ed, but not the chromium. If I
respin, I will add both on Cc.

> 
> Reviewed-by: Chen-Yu Tsai <wenst@chromium.org>


Best regards,
Krzysztof

Re: [PATCH v4 0/6] thermal: scope/cleanup.h improvements

[Chen-Yu Tsai]

On Mon, Oct 14, 2024 at 4:41 PM Krzysztof Kozlowski
<krzysztof.kozlowski@linaro.org> wrote:
>
> On 14/10/2024 10:32, Chen-Yu Tsai wrote:
> > On Fri, Oct 11, 2024 at 2:06 AM Krzysztof Kozlowski
> > <krzysztof.kozlowski@linaro.org> wrote:
> >>
> >> Changes in v4:
> >> - Patch 2: rewrite, significant change: kzalloc() also with
> >>   scoped-handling so the entire error handling could be removed.
> >>   Due to above, drop review-tags (Chen-Yu, Jonathan).
> >> - Add Rb tags for other patches.
> >> - Link to v3: https://lore.kernel.org/r/20241008-b4-cleanup-h-of-node-put-thermal-v3-0-825122398f71@linaro.org
> >>
> >> Changes in v3:
> >> - Rebase, because there was bigger rework in thermal code.
> >>   This made two patches obsolete, but brought new one:
> >>   1/6: thermal: of: Simplify thermal_of_should_bind with scoped for each OF child
> >> - Link to v2: https://lore.kernel.org/r/20240816-b4-cleanup-h-of-node-put-thermal-v2-0-cee9fc490478@linaro.org
> >>
> >> Changes in v2:
> >> - Drop left-over of_node_put in regular exit path (Chen-Yu)
> >> - Link to v1: https://lore.kernel.org/r/20240814-b4-cleanup-h-of-node-put-thermal-v1-0-7a1381e1627e@linaro.org
> >>
> >> Few code simplifications with scope/cleanup.h.
> >>
> >> Best regards,
> >> Krzysztof
> >>
> >> ---
> >> Krzysztof Kozlowski (6):
> >>       thermal: of: Simplify thermal_of_should_bind with scoped for each OF child
> >
> > I couldn't find this in my inbox. But since I already reviewed all the other
> > patches, and I looked at this one on lore, consider the whole series is now
>
> Sorry for that. Your wens@csie.org was cc-ed, but not the chromium. If I
> respin, I will add both on Cc.

That's OK. I should've gotten a copy through LAKML but I could find no
trace of it in my chromium email, which is what I found weird.

ChenYu


> >
> > Reviewed-by: Chen-Yu Tsai <wenst@chromium.org>
>
>
> Best regards,
> Krzysztof
>

Re: [PATCH v4 2/6] thermal: of: Use scoped memory and OF handling to simplify thermal_of_trips_init()

[Jonathan Cameron]

On Thu, 10 Oct 2024 20:06:18 +0200
Krzysztof Kozlowski <krzysztof.kozlowski@linaro.org> wrote:

> Obtain the device node reference and allocate memory with
> scoped/cleanup.h to reduce error handling and make the code a bit
> simpler.
> 
> The code is not equivalent in one minor aspect: outgoing parameter
> "*ntrips" will not be zeroed on errors of memory allocation.  This
> difference is not important, because code was already not zeroing it in
> case of earlier errors and the only caller does not rely on ntrips being
> 0 in case of errors.
> 
> Signed-off-by: Krzysztof Kozlowski <krzysztof.kozlowski@linaro.org>
Trivial unrelated comment inline + maybe return_ptr() is the way to go as
Chen-Yu mentioned.

Reviewed-by: Jonathan Cameron <Jonathan.Cameron@huawei.com>

> ---
> 
> Cc: Jonathan Cameron <Jonathan.Cameron@huawei.com>
> Cc: Chen-Yu Tsai <wenst@chromium.org>
> 
> Changes in v4:
> 1. Significant change: kzalloc() also with scoped-handling so the entire
>    error handling could be removed.
> 2. Due to above, drop review-tags (Chen-Yu, Jonathan).
> 
> Changes in v2:
> 1. Drop left-over of_node_put in regular exit path (Chen-Yu)
> ---
>  drivers/thermal/thermal_of.c | 31 ++++++++-----------------------
>  1 file changed, 8 insertions(+), 23 deletions(-)
> 
> diff --git a/drivers/thermal/thermal_of.c b/drivers/thermal/thermal_of.c
> index f0ffc0e335ba9406f4fd858d6c561f9d23f4b842..37db435b54b124abf25b1d75d6cc4fb75f1c1e5c 100644
> --- a/drivers/thermal/thermal_of.c
> +++ b/drivers/thermal/thermal_of.c
> @@ -95,11 +95,9 @@ static int thermal_of_populate_trip(struct device_node *np,
>  
>  static struct thermal_trip *thermal_of_trips_init(struct device_node *np, int *ntrips)
>  {
> -	struct thermal_trip *tt;
> -	struct device_node *trips;
>  	int ret, count;
>  
> -	trips = of_get_child_by_name(np, "trips");
> +	struct device_node *trips __free(device_node) = of_get_child_by_name(np, "trips");
>  	if (!trips) {
>  		pr_err("Failed to find 'trips' node\n");
>  		return ERR_PTR(-EINVAL);
> @@ -108,36 +106,23 @@ static struct thermal_trip *thermal_of_trips_init(struct device_node *np, int *n
>  	count = of_get_child_count(trips);
>  	if (!count) {
>  		pr_err("No trip point defined\n");
> -		ret = -EINVAL;
> -		goto out_of_node_put;
> +		return ERR_PTR(-EINVAL);
>  	}
>  
> -	tt = kzalloc(sizeof(*tt) * count, GFP_KERNEL);
> -	if (!tt) {
> -		ret = -ENOMEM;
> -		goto out_of_node_put;
> -	}
> -
> -	*ntrips = count;
> +	struct thermal_trip *tt __free(kfree) = kzalloc(sizeof(*tt) * count, GFP_KERNEL);

Trivial and unrelated, but maybe kcalloc(count, sizeof(tt), GFP_KERNEL);

> +	if (!tt)
> +		return ERR_PTR(-ENOMEM);
>  
>  	count = 0;
>  	for_each_child_of_node_scoped(trips, trip) {
>  		ret = thermal_of_populate_trip(trip, &tt[count++]);
>  		if (ret)
> -			goto out_kfree;
> +			return ERR_PTR(ret);
>  	}
>  
> -	of_node_put(trips);
> +	*ntrips = count;
>  
> -	return tt;
> -
> -out_kfree:
> -	kfree(tt);
> -	*ntrips = 0;
> -out_of_node_put:
> -	of_node_put(trips);
> -
> -	return ERR_PTR(ret);
> +	return no_free_ptr(tt);
>  }
>  
>  static struct device_node *of_thermal_zone_find(struct device_node *sensor, int id)
> 

Re: [PATCH v4 0/6] thermal: scope/cleanup.h improvements

[Rafael J. Wysocki]

Hi Daniel,

On Thu, Oct 10, 2024 at 8:06 PM Krzysztof Kozlowski
<krzysztof.kozlowski@linaro.org> wrote:
>
> Changes in v4:
> - Patch 2: rewrite, significant change: kzalloc() also with
>   scoped-handling so the entire error handling could be removed.
>   Due to above, drop review-tags (Chen-Yu, Jonathan).
> - Add Rb tags for other patches.
> - Link to v3: https://lore.kernel.org/r/20241008-b4-cleanup-h-of-node-put-thermal-v3-0-825122398f71@linaro.org
>
> Changes in v3:
> - Rebase, because there was bigger rework in thermal code.
>   This made two patches obsolete, but brought new one:
>   1/6: thermal: of: Simplify thermal_of_should_bind with scoped for each OF child
> - Link to v2: https://lore.kernel.org/r/20240816-b4-cleanup-h-of-node-put-thermal-v2-0-cee9fc490478@linaro.org
>
> Changes in v2:
> - Drop left-over of_node_put in regular exit path (Chen-Yu)
> - Link to v1: https://lore.kernel.org/r/20240814-b4-cleanup-h-of-node-put-thermal-v1-0-7a1381e1627e@linaro.org
>
> Few code simplifications with scope/cleanup.h.
>
> Best regards,
> Krzysztof
>
> ---
> Krzysztof Kozlowski (6):
>       thermal: of: Simplify thermal_of_should_bind with scoped for each OF child
>       thermal: of: Use scoped memory and OF handling to simplify thermal_of_trips_init()
>       thermal: of: Use scoped device node handling to simplify of_thermal_zone_find()
>       thermal: qcom-spmi-adc-tm5: Simplify with scoped for each OF child loop
>       thermal: tegra: Simplify with scoped for each OF child loop
>       thermal: sun8i: Use scoped device node handling to simplify error paths
>
>  drivers/thermal/qcom/qcom-spmi-adc-tm5.c |  7 ++---
>  drivers/thermal/sun8i_thermal.c          | 11 +++----
>  drivers/thermal/tegra/soctherm.c         |  5 ++-
>  drivers/thermal/thermal_of.c             | 54 ++++++++++----------------------
>  4 files changed, 25 insertions(+), 52 deletions(-)
> ---

This seems to have fallen through cracks.

Since all of the other 6.13 thermal changes have been integrated now,
I'm going to apply it and push next week unless you have concerns (in
which case please let me know).

Thanks!