[PATCH 0/3] thermal: Fix/cleanup error paths in __thermal_cooling_device_register()

[PATCH 0/3] thermal: Fix/cleanup error paths in __thermal_cooling_device_register()

[Viresh Kumar]

Hi,

The error paths in __thermal_cooling_device_register() are trying to free
un-acquired resources, fix that and clean it up a bit.

Caleb, can you please test this? I have just build tested it for now.

--
Viresh

Viresh Kumar (3):
  thermal: core: call put_device() only after device_register() fails
  thermal: core: Move cdev cleanup to thermal_release()
  thermal: core: Use device_unregister() instead of device_del/put()

 drivers/thermal/thermal_core.c | 29 +++++++++++++++--------------
 1 file changed, 15 insertions(+), 14 deletions(-)

-- 
2.31.1.272.g89b43f80a514

[PATCH V4 1/3] thermal: core: call put_device() only after device_register() fails

[Viresh Kumar]

put_device() shouldn't be called before a prior call to
device_register(). __thermal_cooling_device_register() doesn't follow
that properly and needs fixing. Also
thermal_cooling_device_destroy_sysfs() is getting called unnecessarily
on few error paths.

Fix all this by placing the calls at the right place.

Based on initial work done by Caleb Connolly.

Fixes: 4748f9687caa ("thermal: core: fix some possible name leaks in error paths")
Fixes: c408b3d1d9bb ("thermal: Validate new state in cur_state_store()")
Reported-by: Caleb Connolly <caleb.connolly@linaro.org>
Signed-off-by: Viresh Kumar <viresh.kumar@linaro.org>
---
For v6.2-rc.

V3->V4:
- The first three versions were sent by Caleb.
- The new version fixes the current bugs, without looking to optimize the
  code any further, which is done separately in the next two patches.

 drivers/thermal/thermal_core.c | 13 ++++++++++---
 1 file changed, 10 insertions(+), 3 deletions(-)

diff --git a/drivers/thermal/thermal_core.c b/drivers/thermal/thermal_core.c
index f17ab2316dbd..77bd47d976a2 100644
--- a/drivers/thermal/thermal_core.c
+++ b/drivers/thermal/thermal_core.c
@@ -909,15 +909,20 @@ __thermal_cooling_device_register(struct device_node *np,
 	cdev->devdata = devdata;
 
 	ret = cdev->ops->get_max_state(cdev, &cdev->max_state);
-	if (ret)
-		goto out_kfree_type;
+	if (ret) {
+		kfree(cdev->type);
+		goto out_ida_remove;
+	}
 
 	thermal_cooling_device_setup_sysfs(cdev);
+
 	ret = dev_set_name(&cdev->device, "cooling_device%d", cdev->id);
 	if (ret) {
+		kfree(cdev->type);
 		thermal_cooling_device_destroy_sysfs(cdev);
-		goto out_kfree_type;
+		goto out_ida_remove;
 	}
+
 	ret = device_register(&cdev->device);
 	if (ret)
 		goto out_kfree_type;
@@ -943,6 +948,8 @@ __thermal_cooling_device_register(struct device_node *np,
 	thermal_cooling_device_destroy_sysfs(cdev);
 	kfree(cdev->type);
 	put_device(&cdev->device);
+
+	/* thermal_release() takes care of the rest */
 	cdev = NULL;
 out_ida_remove:
 	ida_free(&thermal_cdev_ida, id);
-- 
2.31.1.272.g89b43f80a514

Re: [PATCH V4 1/3] thermal: core: call put_device() only after device_register() fails

[Frank Rowand]

On 1/18/23 02:38, Viresh Kumar wrote:
> put_device() shouldn't be called before a prior call to
> device_register(). __thermal_cooling_device_register() doesn't follow
> that properly and needs fixing. Also
> thermal_cooling_device_destroy_sysfs() is getting called unnecessarily
> on few error paths.
> 
> Fix all this by placing the calls at the right place.
> 
> Based on initial work done by Caleb Connolly.
> 
> Fixes: 4748f9687caa ("thermal: core: fix some possible name leaks in error paths")
> Fixes: c408b3d1d9bb ("thermal: Validate new state in cur_state_store()")
> Reported-by: Caleb Connolly <caleb.connolly@linaro.org>
> Signed-off-by: Viresh Kumar <viresh.kumar@linaro.org>
> ---
> For v6.2-rc.
> 
> V3->V4:
> - The first three versions were sent by Caleb.
> - The new version fixes the current bugs, without looking to optimize the
>   code any further, which is done separately in the next two patches.
> 
>  drivers/thermal/thermal_core.c | 13 ++++++++++---
>  1 file changed, 10 insertions(+), 3 deletions(-)
> 
> diff --git a/drivers/thermal/thermal_core.c b/drivers/thermal/thermal_core.c
> index f17ab2316dbd..77bd47d976a2 100644
> --- a/drivers/thermal/thermal_core.c
> +++ b/drivers/thermal/thermal_core.c
> @@ -909,15 +909,20 @@ __thermal_cooling_device_register(struct device_node *np,
>  	cdev->devdata = devdata;
>  
>  	ret = cdev->ops->get_max_state(cdev, &cdev->max_state);
> -	if (ret)
> -		goto out_kfree_type;
> +	if (ret) {
> +		kfree(cdev->type);
> +		goto out_ida_remove;
> +	}
>  
>  	thermal_cooling_device_setup_sysfs(cdev);
> +
>  	ret = dev_set_name(&cdev->device, "cooling_device%d", cdev->id);
>  	if (ret) {
> +		kfree(cdev->type);
>  		thermal_cooling_device_destroy_sysfs(cdev);
> -		goto out_kfree_type;
> +		goto out_ida_remove;
>  	}
> +
>  	ret = device_register(&cdev->device);
>  	if (ret)
>  		goto out_kfree_type;
> @@ -943,6 +948,8 @@ __thermal_cooling_device_register(struct device_node *np,
>  	thermal_cooling_device_destroy_sysfs(cdev);
>  	kfree(cdev->type);
>  	put_device(&cdev->device);
> +
> +	/* thermal_release() takes care of the rest */
>  	cdev = NULL;
>  out_ida_remove:
>  	ida_free(&thermal_cdev_ida, id);

My testing:

 Applied on top of v6.2-rc1
 The configuration is qcom_defconfig
 The system is a Qualcomm Dragon 8074

The two WARNING stack traces no longer occur after applying the patch.

Tested-by: Frank Rowand <frowand.list@gmail.com>

Re: [PATCH V4 1/3] thermal: core: call put_device() only after device_register() fails

[Rafael J. Wysocki]

On Wed, Jan 18, 2023 at 9:38 AM Viresh Kumar <viresh.kumar@linaro.org> wrote:
>
> put_device() shouldn't be called before a prior call to
> device_register(). __thermal_cooling_device_register() doesn't follow
> that properly and needs fixing. Also
> thermal_cooling_device_destroy_sysfs() is getting called unnecessarily
> on few error paths.
>
> Fix all this by placing the calls at the right place.
>
> Based on initial work done by Caleb Connolly.
>
> Fixes: 4748f9687caa ("thermal: core: fix some possible name leaks in error paths")
> Fixes: c408b3d1d9bb ("thermal: Validate new state in cur_state_store()")
> Reported-by: Caleb Connolly <caleb.connolly@linaro.org>
> Signed-off-by: Viresh Kumar <viresh.kumar@linaro.org>

OK, so I think that this patch is needed for 6.2 and the other two may
be queued up for later (they do depend on this one, though, of
course).  Is my understanding correct?

> ---
> For v6.2-rc.
>
> V3->V4:
> - The first three versions were sent by Caleb.
> - The new version fixes the current bugs, without looking to optimize the
>   code any further, which is done separately in the next two patches.
>
>  drivers/thermal/thermal_core.c | 13 ++++++++++---
>  1 file changed, 10 insertions(+), 3 deletions(-)
>
> diff --git a/drivers/thermal/thermal_core.c b/drivers/thermal/thermal_core.c
> index f17ab2316dbd..77bd47d976a2 100644
> --- a/drivers/thermal/thermal_core.c
> +++ b/drivers/thermal/thermal_core.c
> @@ -909,15 +909,20 @@ __thermal_cooling_device_register(struct device_node *np,
>         cdev->devdata = devdata;
>
>         ret = cdev->ops->get_max_state(cdev, &cdev->max_state);
> -       if (ret)
> -               goto out_kfree_type;
> +       if (ret) {
> +               kfree(cdev->type);
> +               goto out_ida_remove;
> +       }
>
>         thermal_cooling_device_setup_sysfs(cdev);
> +
>         ret = dev_set_name(&cdev->device, "cooling_device%d", cdev->id);
>         if (ret) {
> +               kfree(cdev->type);
>                 thermal_cooling_device_destroy_sysfs(cdev);
> -               goto out_kfree_type;
> +               goto out_ida_remove;
>         }
> +
>         ret = device_register(&cdev->device);
>         if (ret)
>                 goto out_kfree_type;
> @@ -943,6 +948,8 @@ __thermal_cooling_device_register(struct device_node *np,
>         thermal_cooling_device_destroy_sysfs(cdev);
>         kfree(cdev->type);
>         put_device(&cdev->device);
> +
> +       /* thermal_release() takes care of the rest */
>         cdev = NULL;
>  out_ida_remove:
>         ida_free(&thermal_cdev_ida, id);
> --
> 2.31.1.272.g89b43f80a514
>

Re: [PATCH V4 1/3] thermal: core: call put_device() only after device_register() fails

[Viresh Kumar]

On 18-01-23, 20:58, Rafael J. Wysocki wrote:
> On Wed, Jan 18, 2023 at 9:38 AM Viresh Kumar <viresh.kumar@linaro.org> wrote:
> >
> > put_device() shouldn't be called before a prior call to
> > device_register(). __thermal_cooling_device_register() doesn't follow
> > that properly and needs fixing. Also
> > thermal_cooling_device_destroy_sysfs() is getting called unnecessarily
> > on few error paths.
> >
> > Fix all this by placing the calls at the right place.
> >
> > Based on initial work done by Caleb Connolly.
> >
> > Fixes: 4748f9687caa ("thermal: core: fix some possible name leaks in error paths")
> > Fixes: c408b3d1d9bb ("thermal: Validate new state in cur_state_store()")
> > Reported-by: Caleb Connolly <caleb.connolly@linaro.org>
> > Signed-off-by: Viresh Kumar <viresh.kumar@linaro.org>
> 
> OK, so I think that this patch is needed for 6.2 and the other two may
> be queued up for later (they do depend on this one, though, of
> course).  Is my understanding correct?

Right.

-- 
viresh

Re: [PATCH V4 1/3] thermal: core: call put_device() only after device_register() fails

[Rafael J. Wysocki]

On Thu, Jan 19, 2023 at 6:16 AM Viresh Kumar <viresh.kumar@linaro.org> wrote:
>
> On 18-01-23, 20:58, Rafael J. Wysocki wrote:
> > On Wed, Jan 18, 2023 at 9:38 AM Viresh Kumar <viresh.kumar@linaro.org> wrote:
> > >
> > > put_device() shouldn't be called before a prior call to
> > > device_register(). __thermal_cooling_device_register() doesn't follow
> > > that properly and needs fixing. Also
> > > thermal_cooling_device_destroy_sysfs() is getting called unnecessarily
> > > on few error paths.
> > >
> > > Fix all this by placing the calls at the right place.
> > >
> > > Based on initial work done by Caleb Connolly.
> > >
> > > Fixes: 4748f9687caa ("thermal: core: fix some possible name leaks in error paths")
> > > Fixes: c408b3d1d9bb ("thermal: Validate new state in cur_state_store()")
> > > Reported-by: Caleb Connolly <caleb.connolly@linaro.org>
> > > Signed-off-by: Viresh Kumar <viresh.kumar@linaro.org>
> >
> > OK, so I think that this patch is needed for 6.2 and the other two may
> > be queued up for later (they do depend on this one, though, of
> > course).  Is my understanding correct?
>
> Right.

OK, applied as 6.2-rc material and I'll get to the other two when this goes in.

Re: [PATCH V4 1/3] thermal: core: call put_device() only after device_register() fails

[Rafael J. Wysocki]

On Thu, Jan 19, 2023 at 9:09 PM Rafael J. Wysocki <rafael@kernel.org> wrote:
>
> On Thu, Jan 19, 2023 at 6:16 AM Viresh Kumar <viresh.kumar@linaro.org> wrote:
> >
> > On 18-01-23, 20:58, Rafael J. Wysocki wrote:
> > > On Wed, Jan 18, 2023 at 9:38 AM Viresh Kumar <viresh.kumar@linaro.org> wrote:
> > > >
> > > > put_device() shouldn't be called before a prior call to
> > > > device_register(). __thermal_cooling_device_register() doesn't follow
> > > > that properly and needs fixing. Also
> > > > thermal_cooling_device_destroy_sysfs() is getting called unnecessarily
> > > > on few error paths.
> > > >
> > > > Fix all this by placing the calls at the right place.
> > > >
> > > > Based on initial work done by Caleb Connolly.
> > > >
> > > > Fixes: 4748f9687caa ("thermal: core: fix some possible name leaks in error paths")
> > > > Fixes: c408b3d1d9bb ("thermal: Validate new state in cur_state_store()")
> > > > Reported-by: Caleb Connolly <caleb.connolly@linaro.org>
> > > > Signed-off-by: Viresh Kumar <viresh.kumar@linaro.org>
> > >
> > > OK, so I think that this patch is needed for 6.2 and the other two may
> > > be queued up for later (they do depend on this one, though, of
> > > course).  Is my understanding correct?
> >
> > Right.
>
> OK, applied as 6.2-rc material and I'll get to the other two when this goes in.

Patches [2-3/3] from this series have been applied as 6.3 material now, thanks!

Re: [PATCH V4 1/3] thermal: core: call put_device() only after device_register() fails

[Yang Yingliang]

On 2023/1/18 16:38, Viresh Kumar wrote:
> put_device() shouldn't be called before a prior call to
> device_register(). __thermal_cooling_device_register() doesn't follow
> that properly and needs fixing. Also
> thermal_cooling_device_destroy_sysfs() is getting called unnecessarily
> on few error paths.
>
> Fix all this by placing the calls at the right place.
>
> Based on initial work done by Caleb Connolly.
>
> Fixes: 4748f9687caa ("thermal: core: fix some possible name leaks in error paths")
> Fixes: c408b3d1d9bb ("thermal: Validate new state in cur_state_store()")
> Reported-by: Caleb Connolly <caleb.connolly@linaro.org>
> Signed-off-by: Viresh Kumar <viresh.kumar@linaro.org>
> ---
Reviewed-by: Yang Yingliang <yangyingliang@huawei.com>
> For v6.2-rc.
>
> V3->V4:
> - The first three versions were sent by Caleb.
> - The new version fixes the current bugs, without looking to optimize the
>    code any further, which is done separately in the next two patches.
>
>   drivers/thermal/thermal_core.c | 13 ++++++++++---
>   1 file changed, 10 insertions(+), 3 deletions(-)
>
> diff --git a/drivers/thermal/thermal_core.c b/drivers/thermal/thermal_core.c
> index f17ab2316dbd..77bd47d976a2 100644
> --- a/drivers/thermal/thermal_core.c
> +++ b/drivers/thermal/thermal_core.c
> @@ -909,15 +909,20 @@ __thermal_cooling_device_register(struct device_node *np,
>   	cdev->devdata = devdata;
>   
>   	ret = cdev->ops->get_max_state(cdev, &cdev->max_state);
> -	if (ret)
> -		goto out_kfree_type;
> +	if (ret) {
> +		kfree(cdev->type);
> +		goto out_ida_remove;
> +	}
>   
>   	thermal_cooling_device_setup_sysfs(cdev);
> +
>   	ret = dev_set_name(&cdev->device, "cooling_device%d", cdev->id);
>   	if (ret) {
> +		kfree(cdev->type);
>   		thermal_cooling_device_destroy_sysfs(cdev);
> -		goto out_kfree_type;
> +		goto out_ida_remove;
>   	}
> +
>   	ret = device_register(&cdev->device);
>   	if (ret)
>   		goto out_kfree_type;
> @@ -943,6 +948,8 @@ __thermal_cooling_device_register(struct device_node *np,
>   	thermal_cooling_device_destroy_sysfs(cdev);
>   	kfree(cdev->type);
>   	put_device(&cdev->device);
> +
> +	/* thermal_release() takes care of the rest */
>   	cdev = NULL;
>   out_ida_remove:
>   	ida_free(&thermal_cdev_ida, id);

Re: [PATCH V4 1/3] thermal: core: call put_device() only after device_register() fails

[Caleb Connolly]

On 18/01/2023 08:38, Viresh Kumar wrote:
> put_device() shouldn't be called before a prior call to
> device_register(). __thermal_cooling_device_register() doesn't follow
> that properly and needs fixing. Also
> thermal_cooling_device_destroy_sysfs() is getting called unnecessarily
> on few error paths.
> 
> Fix all this by placing the calls at the right place.
> 
> Based on initial work done by Caleb Connolly.
> 
> Fixes: 4748f9687caa ("thermal: core: fix some possible name leaks in error paths")
> Fixes: c408b3d1d9bb ("thermal: Validate new state in cur_state_store()")
> Reported-by: Caleb Connolly <caleb.connolly@linaro.org>
> Signed-off-by: Viresh Kumar <viresh.kumar@linaro.org>

Tested-by: Caleb Connolly <caleb.connolly@linaro.org>

Thanks for sending this, with this I no longer hit the splats when
get_max_state() fails.
> ---
> For v6.2-rc.
> 
> V3->V4:
> - The first three versions were sent by Caleb.
> - The new version fixes the current bugs, without looking to optimize the
>   code any further, which is done separately in the next two patches.
> 
>  drivers/thermal/thermal_core.c | 13 ++++++++++---
>  1 file changed, 10 insertions(+), 3 deletions(-)
> 
> diff --git a/drivers/thermal/thermal_core.c b/drivers/thermal/thermal_core.c
> index f17ab2316dbd..77bd47d976a2 100644
> --- a/drivers/thermal/thermal_core.c
> +++ b/drivers/thermal/thermal_core.c
> @@ -909,15 +909,20 @@ __thermal_cooling_device_register(struct device_node *np,
>  	cdev->devdata = devdata;
>  
>  	ret = cdev->ops->get_max_state(cdev, &cdev->max_state);
> -	if (ret)
> -		goto out_kfree_type;
> +	if (ret) {
> +		kfree(cdev->type);
> +		goto out_ida_remove;
> +	}
>  
>  	thermal_cooling_device_setup_sysfs(cdev);
> +
>  	ret = dev_set_name(&cdev->device, "cooling_device%d", cdev->id);
>  	if (ret) {
> +		kfree(cdev->type);
>  		thermal_cooling_device_destroy_sysfs(cdev);
> -		goto out_kfree_type;
> +		goto out_ida_remove;
>  	}
> +
>  	ret = device_register(&cdev->device);
>  	if (ret)
>  		goto out_kfree_type;
> @@ -943,6 +948,8 @@ __thermal_cooling_device_register(struct device_node *np,
>  	thermal_cooling_device_destroy_sysfs(cdev);
>  	kfree(cdev->type);
>  	put_device(&cdev->device);
> +
> +	/* thermal_release() takes care of the rest */
>  	cdev = NULL;
>  out_ida_remove:
>  	ida_free(&thermal_cdev_ida, id);

-- 
Kind Regards,
Caleb (they/them)

[PATCH 2/3] thermal: core: Move cdev cleanup to thermal_release()

[Viresh Kumar]

thermal_release() already frees cdev, let it do rest of the cleanup as
well in order to simplify the error paths in
__thermal_cooling_device_register().

Signed-off-by: Viresh Kumar <viresh.kumar@linaro.org>
---
 drivers/thermal/thermal_core.c | 33 ++++++++++++++-------------------
 1 file changed, 14 insertions(+), 19 deletions(-)

diff --git a/drivers/thermal/thermal_core.c b/drivers/thermal/thermal_core.c
index 77bd47d976a2..1fb109a97ff6 100644
--- a/drivers/thermal/thermal_core.c
+++ b/drivers/thermal/thermal_core.c
@@ -774,6 +774,9 @@ static void thermal_release(struct device *dev)
 	} else if (!strncmp(dev_name(dev), "cooling_device",
 			    sizeof("cooling_device") - 1)) {
 		cdev = to_cooling_device(dev);
+		thermal_cooling_device_destroy_sysfs(cdev);
+		kfree(cdev->type);
+		ida_free(&thermal_cdev_ida, cdev->id);
 		kfree(cdev);
 	}
 }
@@ -909,23 +912,21 @@ __thermal_cooling_device_register(struct device_node *np,
 	cdev->devdata = devdata;
 
 	ret = cdev->ops->get_max_state(cdev, &cdev->max_state);
-	if (ret) {
-		kfree(cdev->type);
-		goto out_ida_remove;
-	}
+	if (ret)
+		goto out_cdev_type;
 
 	thermal_cooling_device_setup_sysfs(cdev);
 
 	ret = dev_set_name(&cdev->device, "cooling_device%d", cdev->id);
-	if (ret) {
-		kfree(cdev->type);
-		thermal_cooling_device_destroy_sysfs(cdev);
-		goto out_ida_remove;
-	}
+	if (ret)
+		goto out_cooling_dev;
 
 	ret = device_register(&cdev->device);
-	if (ret)
-		goto out_kfree_type;
+	if (ret) {
+		/* thermal_release() handles rest of the cleanup */
+		put_device(&cdev->device);
+		return ERR_PTR(ret);
+	}
 
 	/* Add 'this' new cdev to the global cdev list */
 	mutex_lock(&thermal_list_lock);
@@ -944,13 +945,10 @@ __thermal_cooling_device_register(struct device_node *np,
 
 	return cdev;
 
-out_kfree_type:
+out_cooling_dev:
 	thermal_cooling_device_destroy_sysfs(cdev);
+out_cdev_type:
 	kfree(cdev->type);
-	put_device(&cdev->device);
-
-	/* thermal_release() takes care of the rest */
-	cdev = NULL;
 out_ida_remove:
 	ida_free(&thermal_cdev_ida, id);
 out_kfree_cdev:
@@ -1111,10 +1109,7 @@ void thermal_cooling_device_unregister(struct thermal_cooling_device *cdev)
 
 	mutex_unlock(&thermal_list_lock);
 
-	ida_free(&thermal_cdev_ida, cdev->id);
 	device_del(&cdev->device);
-	thermal_cooling_device_destroy_sysfs(cdev);
-	kfree(cdev->type);
 	put_device(&cdev->device);
 }
 EXPORT_SYMBOL_GPL(thermal_cooling_device_unregister);
-- 
2.31.1.272.g89b43f80a514

Re: [PATCH 2/3] thermal: core: Move cdev cleanup to thermal_release()

[Yang Yingliang]

On 2023/1/18 16:38, Viresh Kumar wrote:
> thermal_release() already frees cdev, let it do rest of the cleanup as
> well in order to simplify the error paths in
> __thermal_cooling_device_register().
>
> Signed-off-by: Viresh Kumar <viresh.kumar@linaro.org>
> ---
Reviewed-by: Yang Yingliang <yangyingliang@huawei.com>
>   drivers/thermal/thermal_core.c | 33 ++++++++++++++-------------------
>   1 file changed, 14 insertions(+), 19 deletions(-)
>
> diff --git a/drivers/thermal/thermal_core.c b/drivers/thermal/thermal_core.c
> index 77bd47d976a2..1fb109a97ff6 100644
> --- a/drivers/thermal/thermal_core.c
> +++ b/drivers/thermal/thermal_core.c
> @@ -774,6 +774,9 @@ static void thermal_release(struct device *dev)
>   	} else if (!strncmp(dev_name(dev), "cooling_device",
>   			    sizeof("cooling_device") - 1)) {
>   		cdev = to_cooling_device(dev);
> +		thermal_cooling_device_destroy_sysfs(cdev);
> +		kfree(cdev->type);
> +		ida_free(&thermal_cdev_ida, cdev->id);
>   		kfree(cdev);
>   	}
>   }
> @@ -909,23 +912,21 @@ __thermal_cooling_device_register(struct device_node *np,
>   	cdev->devdata = devdata;
>   
>   	ret = cdev->ops->get_max_state(cdev, &cdev->max_state);
> -	if (ret) {
> -		kfree(cdev->type);
> -		goto out_ida_remove;
> -	}
> +	if (ret)
> +		goto out_cdev_type;
>   
>   	thermal_cooling_device_setup_sysfs(cdev);
>   
>   	ret = dev_set_name(&cdev->device, "cooling_device%d", cdev->id);
> -	if (ret) {
> -		kfree(cdev->type);
> -		thermal_cooling_device_destroy_sysfs(cdev);
> -		goto out_ida_remove;
> -	}
> +	if (ret)
> +		goto out_cooling_dev;
>   
>   	ret = device_register(&cdev->device);
> -	if (ret)
> -		goto out_kfree_type;
> +	if (ret) {
> +		/* thermal_release() handles rest of the cleanup */
> +		put_device(&cdev->device);
> +		return ERR_PTR(ret);
> +	}
>   
>   	/* Add 'this' new cdev to the global cdev list */
>   	mutex_lock(&thermal_list_lock);
> @@ -944,13 +945,10 @@ __thermal_cooling_device_register(struct device_node *np,
>   
>   	return cdev;
>   
> -out_kfree_type:
> +out_cooling_dev:
>   	thermal_cooling_device_destroy_sysfs(cdev);
> +out_cdev_type:
>   	kfree(cdev->type);
> -	put_device(&cdev->device);
> -
> -	/* thermal_release() takes care of the rest */
> -	cdev = NULL;
>   out_ida_remove:
>   	ida_free(&thermal_cdev_ida, id);
>   out_kfree_cdev:
> @@ -1111,10 +1109,7 @@ void thermal_cooling_device_unregister(struct thermal_cooling_device *cdev)
>   
>   	mutex_unlock(&thermal_list_lock);
>   
> -	ida_free(&thermal_cdev_ida, cdev->id);
>   	device_del(&cdev->device);
> -	thermal_cooling_device_destroy_sysfs(cdev);
> -	kfree(cdev->type);
>   	put_device(&cdev->device);
>   }
>   EXPORT_SYMBOL_GPL(thermal_cooling_device_unregister);

[PATCH 3/3] thermal: core: Use device_unregister() instead of device_del/put()

[Viresh Kumar]

Lets not open code device_unregister() unnecessarily.

Signed-off-by: Viresh Kumar <viresh.kumar@linaro.org>
---
 drivers/thermal/thermal_core.c | 3 +--
 1 file changed, 1 insertion(+), 2 deletions(-)

diff --git a/drivers/thermal/thermal_core.c b/drivers/thermal/thermal_core.c
index 1fb109a97ff6..9fb37c5d9c4f 100644
--- a/drivers/thermal/thermal_core.c
+++ b/drivers/thermal/thermal_core.c
@@ -1109,8 +1109,7 @@ void thermal_cooling_device_unregister(struct thermal_cooling_device *cdev)
 
 	mutex_unlock(&thermal_list_lock);
 
-	device_del(&cdev->device);
-	put_device(&cdev->device);
+	device_unregister(&cdev->device);
 }
 EXPORT_SYMBOL_GPL(thermal_cooling_device_unregister);
 
-- 
2.31.1.272.g89b43f80a514

Re: [PATCH 3/3] thermal: core: Use device_unregister() instead of device_del/put()

[Yang Yingliang]

On 2023/1/18 16:38, Viresh Kumar wrote:
> Lets not open code device_unregister() unnecessarily.
>
> Signed-off-by: Viresh Kumar <viresh.kumar@linaro.org>
> ---
Reviewed-by: Yang Yingliang <yangyingliang@huawei.com>
>   drivers/thermal/thermal_core.c | 3 +--
>   1 file changed, 1 insertion(+), 2 deletions(-)
>
> diff --git a/drivers/thermal/thermal_core.c b/drivers/thermal/thermal_core.c
> index 1fb109a97ff6..9fb37c5d9c4f 100644
> --- a/drivers/thermal/thermal_core.c
> +++ b/drivers/thermal/thermal_core.c
> @@ -1109,8 +1109,7 @@ void thermal_cooling_device_unregister(struct thermal_cooling_device *cdev)
>   
>   	mutex_unlock(&thermal_list_lock);
>   
> -	device_del(&cdev->device);
> -	put_device(&cdev->device);
> +	device_unregister(&cdev->device);
>   }
>   EXPORT_SYMBOL_GPL(thermal_cooling_device_unregister);
>