[PATCH] tracing: Cleanup event_enable_trigger_parse() by using __free()

[PATCH] tracing: Cleanup event_enable_trigger_parse() by using __free()

[Steven Rostedt]

From: Steven Rostedt <rostedt@goodmis.org>

The enable_data variable gets freed on most error paths in
event_enable_trigger_parse(). Use free() to free it and just before
returning normally, call no_free_ptr(enable_data) to keep it from being
freed.

Signed-off-by: Steven Rostedt <rostedt@goodmis.org>
---
 kernel/trace/trace_events_trigger.c | 12 ++++--------
 1 file changed, 4 insertions(+), 8 deletions(-)

diff --git a/kernel/trace/trace_events_trigger.c b/kernel/trace/trace_events_trigger.c
index 655db2e82513..fe5ca6ac7c8d 100644
--- a/kernel/trace/trace_events_trigger.c
+++ b/kernel/trace/trace_events_trigger.c
@@ -1739,7 +1739,7 @@ int event_enable_trigger_parse(struct event_command *cmd_ops,
 			       char *glob, char *cmd, char *param_and_filter)
 {
 	struct trace_event_file *event_enable_file;
-	struct enable_trigger_data *enable_data;
+	struct enable_trigger_data *enable_data __free(kfree) = NULL;
 	struct event_trigger_data *trigger_data;
 	struct trace_array *tr = file->tr;
 	char *param, *filter;
@@ -1789,17 +1789,13 @@ int event_enable_trigger_parse(struct event_command *cmd_ops,
 	enable_data->file = event_enable_file;
 
 	trigger_data = trigger_data_alloc(cmd_ops, cmd, param, enable_data);
-	if (!trigger_data) {
-		kfree(enable_data);
+	if (!trigger_data)
 		return ret;
-	}
 
 	if (remove) {
 		event_trigger_unregister(cmd_ops, file, glob+1, trigger_data);
 		kfree(trigger_data);
-		kfree(enable_data);
-		ret = 0;
-		return ret;
+		return 0;
 	}
 
 	/* Up the trigger_data count to make sure nothing frees it on failure */
@@ -1829,6 +1825,7 @@ int event_enable_trigger_parse(struct event_command *cmd_ops,
 		goto out_disable;
 
 	event_trigger_free(trigger_data);
+	no_free_ptr(enable_data);
 	return ret;
  out_disable:
 	trace_event_enable_disable(event_enable_file, 0, 1);
@@ -1837,7 +1834,6 @@ int event_enable_trigger_parse(struct event_command *cmd_ops,
  out_free:
 	event_trigger_reset_filter(cmd_ops, trigger_data);
 	event_trigger_free(trigger_data);
-	kfree(enable_data);
 
 	return ret;
 }
-- 
2.53.0

Re: [PATCH] tracing: Cleanup event_enable_trigger_parse() by using __free()

[Masami Hiramatsu]

On Wed, 1 Jul 2026 17:48:29 -0400
Steven Rostedt <rostedt@goodmis.org> wrote:

> From: Steven Rostedt <rostedt@goodmis.org>
> 
> The enable_data variable gets freed on most error paths in
> event_enable_trigger_parse(). Use free() to free it and just before
> returning normally, call no_free_ptr(enable_data) to keep it from being
> freed.
> 

This looks good to me.

Reviewed-by: Masami Hiramatsu (Google) <mhiramat@kernel.org>

Thank you,

> Signed-off-by: Steven Rostedt <rostedt@goodmis.org>
> ---
>  kernel/trace/trace_events_trigger.c | 12 ++++--------
>  1 file changed, 4 insertions(+), 8 deletions(-)
> 
> diff --git a/kernel/trace/trace_events_trigger.c b/kernel/trace/trace_events_trigger.c
> index 655db2e82513..fe5ca6ac7c8d 100644
> --- a/kernel/trace/trace_events_trigger.c
> +++ b/kernel/trace/trace_events_trigger.c
> @@ -1739,7 +1739,7 @@ int event_enable_trigger_parse(struct event_command *cmd_ops,
>  			       char *glob, char *cmd, char *param_and_filter)
>  {
>  	struct trace_event_file *event_enable_file;
> -	struct enable_trigger_data *enable_data;
> +	struct enable_trigger_data *enable_data __free(kfree) = NULL;
>  	struct event_trigger_data *trigger_data;
>  	struct trace_array *tr = file->tr;
>  	char *param, *filter;
> @@ -1789,17 +1789,13 @@ int event_enable_trigger_parse(struct event_command *cmd_ops,
>  	enable_data->file = event_enable_file;
>  
>  	trigger_data = trigger_data_alloc(cmd_ops, cmd, param, enable_data);
> -	if (!trigger_data) {
> -		kfree(enable_data);
> +	if (!trigger_data)
>  		return ret;
> -	}
>  
>  	if (remove) {
>  		event_trigger_unregister(cmd_ops, file, glob+1, trigger_data);
>  		kfree(trigger_data);
> -		kfree(enable_data);
> -		ret = 0;
> -		return ret;
> +		return 0;
>  	}
>  
>  	/* Up the trigger_data count to make sure nothing frees it on failure */
> @@ -1829,6 +1825,7 @@ int event_enable_trigger_parse(struct event_command *cmd_ops,
>  		goto out_disable;
>  
>  	event_trigger_free(trigger_data);
> +	no_free_ptr(enable_data);
>  	return ret;
>   out_disable:
>  	trace_event_enable_disable(event_enable_file, 0, 1);
> @@ -1837,7 +1834,6 @@ int event_enable_trigger_parse(struct event_command *cmd_ops,
>   out_free:
>  	event_trigger_reset_filter(cmd_ops, trigger_data);
>  	event_trigger_free(trigger_data);
> -	kfree(enable_data);
>  
>  	return ret;
>  }
> -- 
> 2.53.0
> 


-- 
Masami Hiramatsu (Google) <mhiramat@kernel.org>

Re: [PATCH] tracing: Cleanup event_enable_trigger_parse() by using __free()

[Markus Elfring]

…
> +++ b/kernel/trace/trace_events_trigger.c
> @@ -1739,7 +1739,7 @@ int event_enable_trigger_parse(struct event_command *cmd_ops,
>  			       char *glob, char *cmd, char *param_and_filter)
>  {
>  	struct trace_event_file *event_enable_file;
> -	struct enable_trigger_data *enable_data;
> +	struct enable_trigger_data *enable_data __free(kfree) = NULL;
…

How do you think about to reduce the scope for this local variable?

Regards,
Markus

Re: [PATCH] tracing: Cleanup event_enable_trigger_parse() by using __free()

[Steven Rostedt]

On Thu, 2 Jul 2026 11:58:02 +0200
Markus Elfring <Markus.Elfring@web.de> wrote:

> …
> > +++ b/kernel/trace/trace_events_trigger.c
> > @@ -1739,7 +1739,7 @@ int event_enable_trigger_parse(struct event_command *cmd_ops,
> >  			       char *glob, char *cmd, char *param_and_filter)
> >  {
> >  	struct trace_event_file *event_enable_file;
> > -	struct enable_trigger_data *enable_data;
> > +	struct enable_trigger_data *enable_data __free(kfree) = NULL;  
> …
> 
> How do you think about to reduce the scope for this local variable?
>

No, it's fine as is.

-- Steve

Re: [PATCH] tracing: Cleanup event_enable_trigger_parse() by using __free()

[Markus Elfring]

>> …
>>> +++ b/kernel/trace/trace_events_trigger.c
>>> @@ -1739,7 +1739,7 @@ int event_enable_trigger_parse(struct event_command *cmd_ops,
>>>  			       char *glob, char *cmd, char *param_and_filter)
>>>  {
>>>  	struct trace_event_file *event_enable_file;
>>> -	struct enable_trigger_data *enable_data;
>>> +	struct enable_trigger_data *enable_data __free(kfree) = NULL;  
>> …
>>
>> How do you think about to reduce the scope for this local variable?
>>
> 
> No, it's fine as is.
Does this feedback mean that you disagree to an essential development requirement
according to such a programming interface?
https://elixir.bootlin.com/linux/v7.1.2/source/include/linux/cleanup.h#L142-L153

Regards,
Markus

Re: [PATCH] tracing: Cleanup event_enable_trigger_parse() by using __free()

[Steven Rostedt]

On Fri, 3 Jul 2026 07:00:10 +0200
Markus Elfring <Markus.Elfring@web.de> wrote:

> > No, it's fine as is.  
> Does this feedback mean that you disagree to an essential development requirement
> according to such a programming interface?
> https://elixir.bootlin.com/linux/v7.1.2/source/include/linux/cleanup.h#L142-L153

>  * Given that the "__free(...) = NULL" pattern for variables defined at
>  * the top of the function poses this potential interdependency problem
>  * the recommendation is to always define and assign variables in one
>  * statement and not group variable definitions at the top of the
>  * function when __free() is used.

This is really just a guideline.

>  *
>  * Lastly, given that the benefit of cleanup helpers is removal of
>  * "goto", and that the "goto" statement can jump between scopes, the
>  * expectation is that usage of "goto" and cleanup helpers is never
>  * mixed in the same function. I.e. for a given routine, convert all
>  * resources that need a "goto" cleanup to scope-based cleanup, or
>  * convert none of them.

The above has more to do with guard()s and not so much with a __free() at
the top of the function. The reason is because things get broken if a goto
jumps over a guard().

In other words, if you don't know what you are doing, don't mix them.

-- Steve

Re: [PATCH] tracing: Cleanup event_enable_trigger_parse() by using __free()

[Markus Elfring]

>>  * Given that the "__free(...) = NULL" pattern for variables defined at
>>  * the top of the function poses this potential interdependency problem
>>  * the recommendation is to always define and assign variables in one
>>  * statement and not group variable definitions at the top of the
>>  * function when __free() is used.
> 
> This is really just a guideline.
Do any contributors care a bit more for involved technical details?
https://lore.kernel.org/arm-scmi/20251208020907.5476-4-krzysztof.kozlowski@oss.qualcomm.com/

Regards,
Markus

Re: [PATCH] tracing: Cleanup event_enable_trigger_parse() by using __free()

[Steven Rostedt]

On Fri, 3 Jul 2026 20:37:09 +0200
Markus Elfring <Markus.Elfring@web.de> wrote:
> Do any contributors care a bit more for involved technical details?
> https://lore.kernel.org/arm-scmi/20251208020907.5476-4-krzysztof.kozlowski@oss.qualcomm.com/

That example uses the variable in a specific block whereas this variable is
used in most of the function, which makes it appropriate where it lies.

-- Steve

Re: tracing: Cleanup event_enable_trigger_parse() by using __free()

[Markus Elfring]

> That example uses the variable in a specific block whereas this variable is
> used in most of the function, which makes it appropriate where it lies.

What does hinder you to apply scope reductions for selected local variables
more often?

Regards,
Markus

Re: tracing: Cleanup event_enable_trigger_parse() by using __free()

[Steven Rostedt]

On Fri, 3 Jul 2026 21:20:46 +0200
Markus Elfring <Markus.Elfring@web.de> wrote:

> > That example uses the variable in a specific block whereas this variable is
> > used in most of the function, which makes it appropriate where it lies.  
> 
> What does hinder you to apply scope reductions for selected local variables
> more often?
> 

OK, now you are just talking like Eliza.

-- Steve

Re: [PATCH] tracing: Cleanup event_enable_trigger_parse() by using __free()

[Markus Elfring]

> The enable_data variable gets freed on most error paths in
> event_enable_trigger_parse(). Use free() to free it and just before
> returning normally, call no_free_ptr(enable_data) to keep it from being
> freed.

The conversion is incomplete for the application of scope-based resource management so far.

Regards,
Markus