Re: [RESEND] devfreq: governor: Save void *data in the governor userspace

[Kant Fan <kant@allwinnertech.com>]

On 3/29/2022 5:14 PM, Kant Fan wrote:
> The member void *data in the structure devfreq can be overwrite
> by governor_userspace. For example:
> 1. The device driver assigned the devfreq governor to simple_ondemand
> by the function devfreq_add_device() and init the devfreq member
> void *data to a pointer of a static structure devfreq_simple_ondemand_data
> by the function devfreq_add_device().
> 2. The user changed the devfreq governor to userspace by the command
> "echo userspace > /sys/class/devfreq/.../governor".
> 3. The governor userspace alloced a dynamic memory for the struct
> userspace_data and assigend the member void *data of devfreq to
> this memory by the function userspace_init().
> 4. The user changed the devfreq governor back to simple_ondemand
> by the command "echo simple_ondemand > /sys/class/devfreq/.../governor".
> 5. The governor userspace exited and assigned the member void *data
> in the structure devfreq to NULL by the function userspace_exit().
> 6. The governor simple_ondemand fetched the static information of
> devfreq_simple_ondemand_data in the function
> devfreq_simple_ondemand_func() but the member void *data of devfreq was
> assigned to NULL by the function userspace_exit().
> 7. The information of upthreshold and downdifferential is lost
> and the governor simple_ondemand can't work correctly.
> 
> The member void *data in the structure devfreq is designed for
> a static pointer used in a governor and inited by the function
> devfreq_add_device(). So if a governor want to use void *data
> to do some other things, it must save void *data in the init()
> function and restore void *data in the exit() function.
> 
> Signed-off-by: Kant Fan <kant@allwinnertech.com>
> ---
>   drivers/devfreq/governor_userspace.c | 6 +++++-
>   1 file changed, 5 insertions(+), 1 deletion(-)
> 
> diff --git a/drivers/devfreq/governor_userspace.c b/drivers/devfreq/governor_userspace.c
> index ab9db7adb3ad..dbbb448dcbcf 100644
> --- a/drivers/devfreq/governor_userspace.c
> +++ b/drivers/devfreq/governor_userspace.c
> @@ -17,6 +17,7 @@
>   struct userspace_data {
>   	unsigned long user_frequency;
>   	bool valid;
> +	void *saved_data;
>   };
>   
>   static int devfreq_userspace_func(struct devfreq *df, unsigned long *freq)
> @@ -91,6 +92,7 @@ static int userspace_init(struct devfreq *devfreq)
>   		goto out;
>   	}
>   	data->valid = false;
> +	data->saved_data = devfreq->data;
>   	devfreq->data = data;
>   
>   	err = sysfs_create_group(&devfreq->dev.kobj, &dev_attr_group);
> @@ -100,6 +102,8 @@ static int userspace_init(struct devfreq *devfreq)
>   
>   static void userspace_exit(struct devfreq *devfreq)
>   {
> +	struct userspace_data *data = devfreq->data;
> +	void *saved_data = data->saved_data;
>   	/*
>   	 * Remove the sysfs entry, unless this is being called after
>   	 * device_del(), which should have done this already via kobject_del().
> @@ -108,7 +112,7 @@ static void userspace_exit(struct devfreq *devfreq)
>   		sysfs_remove_group(&devfreq->dev.kobj, &dev_attr_group);
>   
>   	kfree(devfreq->data);
> -	devfreq->data = NULL;
> +	devfreq->data = saved_data;
>   }
>   
>   static int devfreq_userspace_handler(struct devfreq *devfreq,

Dear MyungJoo, Kyungmin & Chanwoo,
Gently ping this issue... Does this patch has a chance to be accepted?
This seems to be a bug in devfreq userspace governor, which affects the 
switching between governors -- When switching from userspace to 
ondemand, the ondemand governor would be invalid.
If there's any question, please let me know.
Thank you.

-- 
Best Regards,
Kant Fan