diff mbox series

[2/4] cpuidle: riscv-sbi: Use scoped device node handling to simplify error paths

Message ID 20240816150931.142208-2-krzysztof.kozlowski@linaro.org
State Superseded
Headers show
Series None | expand

Commit Message

Krzysztof Kozlowski Aug. 16, 2024, 3:09 p.m. UTC
Obtain the device node reference with scoped/cleanup.h to reduce error
handling and make the code a bit simpler.

Signed-off-by: Krzysztof Kozlowski <krzysztof.kozlowski@linaro.org>
---
 drivers/cpuidle/cpuidle-riscv-sbi.c | 21 +++++++--------------
 1 file changed, 7 insertions(+), 14 deletions(-)

Comments

Jonathan Cameron Aug. 19, 2024, 4:13 p.m. UTC | #1
On Fri, 16 Aug 2024 17:09:29 +0200
Krzysztof Kozlowski <krzysztof.kozlowski@linaro.org> wrote:

> Obtain the device node reference with scoped/cleanup.h to reduce error
> handling and make the code a bit simpler.
> 
> Signed-off-by: Krzysztof Kozlowski <krzysztof.kozlowski@linaro.org>
The original code looks suspect. See below.

> ---
>  drivers/cpuidle/cpuidle-riscv-sbi.c | 21 +++++++--------------
>  1 file changed, 7 insertions(+), 14 deletions(-)
> 
> diff --git a/drivers/cpuidle/cpuidle-riscv-sbi.c b/drivers/cpuidle/cpuidle-riscv-sbi.c
> index a6e123dfe394..5bb3401220d2 100644
> --- a/drivers/cpuidle/cpuidle-riscv-sbi.c
> +++ b/drivers/cpuidle/cpuidle-riscv-sbi.c
> @@ -8,6 +8,7 @@
>  
>  #define pr_fmt(fmt) "cpuidle-riscv-sbi: " fmt
>  
> +#include <linux/cleanup.h>
>  #include <linux/cpuhotplug.h>
>  #include <linux/cpuidle.h>
>  #include <linux/cpumask.h>
> @@ -236,19 +237,16 @@ static int sbi_cpuidle_dt_init_states(struct device *dev,
>  {
>  	struct sbi_cpuidle_data *data = per_cpu_ptr(&sbi_cpuidle_data, cpu);
>  	struct device_node *state_node;
> -	struct device_node *cpu_node;
>  	u32 *states;
>  	int i, ret;
>  
> -	cpu_node = of_cpu_device_node_get(cpu);
> +	struct device_node *cpu_node __free(device_node) = of_cpu_device_node_get(cpu);
>  	if (!cpu_node)
>  		return -ENODEV;
>  
>  	states = devm_kcalloc(dev, state_count, sizeof(*states), GFP_KERNEL);
> -	if (!states) {
> -		ret = -ENOMEM;
> -		goto fail;
> -	}
> +	if (!states)
> +		return -ENOMEM;
>  
>  	/* Parse SBI specific details from state DT nodes */
>  	for (i = 1; i < state_count; i++) {
> @@ -264,10 +262,8 @@ static int sbi_cpuidle_dt_init_states(struct device *dev,
>  
>  		pr_debug("sbi-state %#x index %d\n", states[i], i);
>  	}
> -	if (i != state_count) {
> -		ret = -ENODEV;
> -		goto fail;
> -	}
> +	if (i != state_count)
> +		return -ENODEV;
>  
>  	/* Initialize optional data, used for the hierarchical topology. */
>  	ret = sbi_dt_cpu_init_topology(drv, data, state_count, cpu);
The handling of error ret from here doesn't free the node.

Bug or something subtle I'm missing?

If it's a bug, then fixes tag.


> @@ -277,10 +273,7 @@ static int sbi_cpuidle_dt_init_states(struct device *dev,
>  	/* Store states in the per-cpu struct. */
>  	data->states = states;
>  
> -fail:
> -	of_node_put(cpu_node);
> -
> -	return ret;
> +	return 0;
>  }
>  
>  static void sbi_cpuidle_deinit_cpu(int cpu)
Jonathan Cameron Aug. 19, 2024, 4:19 p.m. UTC | #2
On Mon, 19 Aug 2024 17:13:13 +0100
Jonathan Cameron <Jonathan.Cameron@Huawei.com> wrote:

> On Fri, 16 Aug 2024 17:09:29 +0200
> Krzysztof Kozlowski <krzysztof.kozlowski@linaro.org> wrote:
> 
> > Obtain the device node reference with scoped/cleanup.h to reduce error
> > handling and make the code a bit simpler.
> > 
> > Signed-off-by: Krzysztof Kozlowski <krzysztof.kozlowski@linaro.org>  
> The original code looks suspect. See below.

Whilst here...  Why not do similar for state_node to avoid
the delayed return check.
Existing code
	{
		state_node = of_get_cpu_state_node(cpu_node, i - 1);
		if (!state_node)
			break;

		ret = sbi_dt_parse_state_node(state_node, &states[i]);
		of_node_put(state_node);

		if (ret)
			//another bug here on holding cpu_node btw.
			return ret;
		pr_debug("sbi-state %#x index %d\n", states[i], i);
	}
//I think only path to this is is early break above.
	if (i != state_count) {
		ret = -ENODEV;
		goto fail;
	}
Can be something like

	{
		struct device_node *state_node __free(device_node) =
			= of_get-cpu_State_nod(cpu_node, i - 1);
	
		if (!state_node)
			return -ENODEV;

		ret = sbi_dt_parse_state_node(state_node, &states[i]);
		if (ret)
			return ret;

		pr_debug("sbi-state %#x index %d\n", states[i], i);
	}
		

> 
> > ---
> >  drivers/cpuidle/cpuidle-riscv-sbi.c | 21 +++++++--------------
> >  1 file changed, 7 insertions(+), 14 deletions(-)
> > 
> > diff --git a/drivers/cpuidle/cpuidle-riscv-sbi.c b/drivers/cpuidle/cpuidle-riscv-sbi.c
> > index a6e123dfe394..5bb3401220d2 100644
> > --- a/drivers/cpuidle/cpuidle-riscv-sbi.c
> > +++ b/drivers/cpuidle/cpuidle-riscv-sbi.c
> > @@ -8,6 +8,7 @@
> >  
> >  #define pr_fmt(fmt) "cpuidle-riscv-sbi: " fmt
> >  
> > +#include <linux/cleanup.h>
> >  #include <linux/cpuhotplug.h>
> >  #include <linux/cpuidle.h>
> >  #include <linux/cpumask.h>
> > @@ -236,19 +237,16 @@ static int sbi_cpuidle_dt_init_states(struct device *dev,
> >  {
> >  	struct sbi_cpuidle_data *data = per_cpu_ptr(&sbi_cpuidle_data, cpu);
> >  	struct device_node *state_node;
> > -	struct device_node *cpu_node;
> >  	u32 *states;
> >  	int i, ret;
> >  
> > -	cpu_node = of_cpu_device_node_get(cpu);
> > +	struct device_node *cpu_node __free(device_node) = of_cpu_device_node_get(cpu);
> >  	if (!cpu_node)
> >  		return -ENODEV;
> >  
> >  	states = devm_kcalloc(dev, state_count, sizeof(*states), GFP_KERNEL);
> > -	if (!states) {
> > -		ret = -ENOMEM;
> > -		goto fail;
> > -	}
> > +	if (!states)
> > +		return -ENOMEM;
> >  
> >  	/* Parse SBI specific details from state DT nodes */
> >  	for (i = 1; i < state_count; i++) {
> > @@ -264,10 +262,8 @@ static int sbi_cpuidle_dt_init_states(struct device *dev,
> >  
> >  		pr_debug("sbi-state %#x index %d\n", states[i], i);
> >  	}
> > -	if (i != state_count) {
> > -		ret = -ENODEV;
> > -		goto fail;
> > -	}
> > +	if (i != state_count)
> > +		return -ENODEV;
> >  
> >  	/* Initialize optional data, used for the hierarchical topology. */
> >  	ret = sbi_dt_cpu_init_topology(drv, data, state_count, cpu);  
> The handling of error ret from here doesn't free the node.
> 
> Bug or something subtle I'm missing?
> 
> If it's a bug, then fixes tag.

> 
> 
> > @@ -277,10 +273,7 @@ static int sbi_cpuidle_dt_init_states(struct device *dev,
> >  	/* Store states in the per-cpu struct. */
> >  	data->states = states;
> >  
> > -fail:
> > -	of_node_put(cpu_node);
> > -
> > -	return ret;
> > +	return 0;
> >  }
> >  
> >  static void sbi_cpuidle_deinit_cpu(int cpu)  
>
Krzysztof Kozlowski Aug. 20, 2024, 9:29 a.m. UTC | #3
On 19/08/2024 18:13, Jonathan Cameron wrote:
>>  
>>  	/* Parse SBI specific details from state DT nodes */
>>  	for (i = 1; i < state_count; i++) {
>> @@ -264,10 +262,8 @@ static int sbi_cpuidle_dt_init_states(struct device *dev,
>>  
>>  		pr_debug("sbi-state %#x index %d\n", states[i], i);
>>  	}
>> -	if (i != state_count) {
>> -		ret = -ENODEV;
>> -		goto fail;
>> -	}
>> +	if (i != state_count)
>> +		return -ENODEV;
>>  
>>  	/* Initialize optional data, used for the hierarchical topology. */
>>  	ret = sbi_dt_cpu_init_topology(drv, data, state_count, cpu);
> The handling of error ret from here doesn't free the node.
> 
> Bug or something subtle I'm missing?
> 
> If it's a bug, then fixes tag.

Yeah, indeed it is a fix.

Best regards,
Krzysztof
Krzysztof Kozlowski Aug. 20, 2024, 9:36 a.m. UTC | #4
On 19/08/2024 18:19, Jonathan Cameron wrote:
> On Mon, 19 Aug 2024 17:13:13 +0100
> Jonathan Cameron <Jonathan.Cameron@Huawei.com> wrote:
> 
>> On Fri, 16 Aug 2024 17:09:29 +0200
>> Krzysztof Kozlowski <krzysztof.kozlowski@linaro.org> wrote:
>>
>>> Obtain the device node reference with scoped/cleanup.h to reduce error
>>> handling and make the code a bit simpler.
>>>
>>> Signed-off-by: Krzysztof Kozlowski <krzysztof.kozlowski@linaro.org>  
>> The original code looks suspect. See below.
> 
> Whilst here...  Why not do similar for state_node to avoid
> the delayed return check.
> Existing code
> 	{
> 		state_node = of_get_cpu_state_node(cpu_node, i - 1);
> 		if (!state_node)
> 			break;

I don't see how __free() helps here. You can return regardless of __free().

> 
> 		ret = sbi_dt_parse_state_node(state_node, &states[i]);
> 		of_node_put(state_node);

... and this code is quite easy to read: you get reference and
immediately release it.

> 
> 		if (ret)
> 			//another bug here on holding cpu_node btw.
> 			return ret;
> 		pr_debug("sbi-state %#x index %d\n", states[i], i);
> 	}
> //I think only path to this is is early break above.
> 	if (i != state_count) {
> 		ret = -ENODEV;
> 		goto fail;
> 	}
> Can be something like
> 
> 	{
> 		struct device_node *state_node __free(device_node) =
> 			= of_get-cpu_State_nod(cpu_node, i - 1);
> 	
> 		if (!state_node)
> 			return -ENODEV;
> 
> 		ret = sbi_dt_parse_state_node(state_node, &states[i]);
> 		if (ret)
> 			return ret;
> 
> 		pr_debug("sbi-state %#x index %d\n", states[i], i);
> 	}
> 		

Maybe I miss something, but I do not see how the __free() simplifies
here anything.

Best regards,
Krzysztof
Jonathan Cameron Aug. 21, 2024, 11:48 a.m. UTC | #5
On Tue, 20 Aug 2024 11:36:32 +0200
Krzysztof Kozlowski <krzysztof.kozlowski@linaro.org> wrote:

> On 19/08/2024 18:19, Jonathan Cameron wrote:
> > On Mon, 19 Aug 2024 17:13:13 +0100
> > Jonathan Cameron <Jonathan.Cameron@Huawei.com> wrote:
> >   
> >> On Fri, 16 Aug 2024 17:09:29 +0200
> >> Krzysztof Kozlowski <krzysztof.kozlowski@linaro.org> wrote:
> >>  
> >>> Obtain the device node reference with scoped/cleanup.h to reduce error
> >>> handling and make the code a bit simpler.
> >>>
> >>> Signed-off-by: Krzysztof Kozlowski <krzysztof.kozlowski@linaro.org>    
> >> The original code looks suspect. See below.  
> > 
> > Whilst here...  Why not do similar for state_node to avoid
> > the delayed return check.
> > Existing code
> > 	{
> > 		state_node = of_get_cpu_state_node(cpu_node, i - 1);
> > 		if (!state_node)
> > 			break;  
> 
> I don't see how __free() helps here. You can return regardless of __free().
> 
> > 
> > 		ret = sbi_dt_parse_state_node(state_node, &states[i]);
> > 		of_node_put(state_node);  
> 
> ... and this code is quite easy to read: you get reference and
> immediately release it.
> 
> > 
> > 		if (ret)
> > 			//another bug here on holding cpu_node btw.
> > 			return ret;
> > 		pr_debug("sbi-state %#x index %d\n", states[i], i);
> > 	}
> > //I think only path to this is is early break above.
> > 	if (i != state_count) {
> > 		ret = -ENODEV;
> > 		goto fail;
> > 	}
> > Can be something like
> > 
> > 	{
> > 		struct device_node *state_node __free(device_node) =
> > 			= of_get-cpu_State_nod(cpu_node, i - 1);
> > 	
> > 		if (!state_node)
> > 			return -ENODEV;
> > 
> > 		ret = sbi_dt_parse_state_node(state_node, &states[i]);
> > 		if (ret)
> > 			return ret;
> > 
> > 		pr_debug("sbi-state %#x index %d\n", states[i], i);
> > 	}
> > 		  
> 
> Maybe I miss something, but I do not see how the __free() simplifies
> here anything.

Personal preference.  To my eyes, it does, but indeed not a huge
advantage.

Jonathan

> 
> Best regards,
> Krzysztof
>
diff mbox series

Patch

diff --git a/drivers/cpuidle/cpuidle-riscv-sbi.c b/drivers/cpuidle/cpuidle-riscv-sbi.c
index a6e123dfe394..5bb3401220d2 100644
--- a/drivers/cpuidle/cpuidle-riscv-sbi.c
+++ b/drivers/cpuidle/cpuidle-riscv-sbi.c
@@ -8,6 +8,7 @@ 
 
 #define pr_fmt(fmt) "cpuidle-riscv-sbi: " fmt
 
+#include <linux/cleanup.h>
 #include <linux/cpuhotplug.h>
 #include <linux/cpuidle.h>
 #include <linux/cpumask.h>
@@ -236,19 +237,16 @@  static int sbi_cpuidle_dt_init_states(struct device *dev,
 {
 	struct sbi_cpuidle_data *data = per_cpu_ptr(&sbi_cpuidle_data, cpu);
 	struct device_node *state_node;
-	struct device_node *cpu_node;
 	u32 *states;
 	int i, ret;
 
-	cpu_node = of_cpu_device_node_get(cpu);
+	struct device_node *cpu_node __free(device_node) = of_cpu_device_node_get(cpu);
 	if (!cpu_node)
 		return -ENODEV;
 
 	states = devm_kcalloc(dev, state_count, sizeof(*states), GFP_KERNEL);
-	if (!states) {
-		ret = -ENOMEM;
-		goto fail;
-	}
+	if (!states)
+		return -ENOMEM;
 
 	/* Parse SBI specific details from state DT nodes */
 	for (i = 1; i < state_count; i++) {
@@ -264,10 +262,8 @@  static int sbi_cpuidle_dt_init_states(struct device *dev,
 
 		pr_debug("sbi-state %#x index %d\n", states[i], i);
 	}
-	if (i != state_count) {
-		ret = -ENODEV;
-		goto fail;
-	}
+	if (i != state_count)
+		return -ENODEV;
 
 	/* Initialize optional data, used for the hierarchical topology. */
 	ret = sbi_dt_cpu_init_topology(drv, data, state_count, cpu);
@@ -277,10 +273,7 @@  static int sbi_cpuidle_dt_init_states(struct device *dev,
 	/* Store states in the per-cpu struct. */
 	data->states = states;
 
-fail:
-	of_node_put(cpu_node);
-
-	return ret;
+	return 0;
 }
 
 static void sbi_cpuidle_deinit_cpu(int cpu)