Fix possible NULL Pointer Dereference in 'extcon_rtk_type_c_probe'

[PATCH] Fix possible NULL Pointer Dereference in 'extcon_rtk_type_c_probe'

Posted by Gax-c 1 year, 4 months ago

A 'devm_kzalloc' in 'extcon_rtk_type_c_probe' could possibly return null pointer.
Add a null check for the returned pointer.

Fixes: 8a590d7371f0 ("extcon: add Realtek DHC RTD SoC Type-C driver")
Signed-off-by: Zichen Xie <zichenxie0106@gmail.com>
Reported-by: Zichen Xie <zichenxie0106@gmail.com>
Reported-by: Zijie Zhao <zzjas98@gmail.com>
Reported-by: Chenyuan Yang <chenyuan0y@gmail.com>
---
 drivers/extcon/extcon-rtk-type-c.c | 6 ++++++
 1 file changed, 6 insertions(+)

diff --git a/drivers/extcon/extcon-rtk-type-c.c b/drivers/extcon/extcon-rtk-type-c.c
index a592bab77538..9a0824fc7cbe 100644
--- a/drivers/extcon/extcon-rtk-type-c.c
+++ b/drivers/extcon/extcon-rtk-type-c.c
@@ -1371,6 +1371,12 @@ static int extcon_rtk_type_c_probe(struct platform_device *pdev)
 
 	type_c->type_c_cfg = devm_kzalloc(dev, sizeof(*type_c_cfg), GFP_KERNEL);
 
+	if (!type_c->type_c_cfg) {
+		dev_err(dev, "Failed to allocate memory for type_c_cfg!\n");
+		ret = -ENOMEM;
+		goto err;
+	}
+
 	memcpy(type_c->type_c_cfg, type_c_cfg, sizeof(*type_c_cfg));
 
 	if (setup_type_c_parameter(type_c)) {
-- 
2.25.1

Re: [PATCH] Fix possible NULL Pointer Dereference in 'extcon_rtk_type_c_probe'

Posted by Krzysztof Kozlowski 1 year, 4 months ago

On 30/09/2024 00:38, Gax-c wrote:
> A 'devm_kzalloc' in 'extcon_rtk_type_c_probe' could possibly return null pointer.
> Add a null check for the returned pointer.
> 
> Fixes: 8a590d7371f0 ("extcon: add Realtek DHC RTD SoC Type-C driver")
> Signed-off-by: Zichen Xie <zichenxie0106@gmail.com>
> Reported-by: Zichen Xie <zichenxie0106@gmail.com>
> Reported-by: Zijie Zhao <zzjas98@gmail.com>
> Reported-by: Chenyuan Yang <chenyuan0y@gmail.com>

All previous comments apply.

> ---
>  drivers/extcon/extcon-rtk-type-c.c | 6 ++++++
>  1 file changed, 6 insertions(+)
> 
> diff --git a/drivers/extcon/extcon-rtk-type-c.c b/drivers/extcon/extcon-rtk-type-c.c
> index a592bab77538..9a0824fc7cbe 100644
> --- a/drivers/extcon/extcon-rtk-type-c.c
> +++ b/drivers/extcon/extcon-rtk-type-c.c
> @@ -1371,6 +1371,12 @@ static int extcon_rtk_type_c_probe(struct platform_device *pdev)
>  
>  	type_c->type_c_cfg = devm_kzalloc(dev, sizeof(*type_c_cfg), GFP_KERNEL);
>  

Same as in other patches - drop blank line.

> +	if (!type_c->type_c_cfg) {
> +		dev_err(dev, "Failed to allocate memory for type_c_cfg!\n");

No, drop.


> +		ret = -ENOMEM;
> +		goto err;

Best regards,
Krzysztof

Re: [PATCH] Fix possible NULL Pointer Dereference in 'extcon_rtk_type_c_probe'

Posted by Markus Elfring 1 year, 4 months ago

> A 'devm_kzalloc' in 'extcon_rtk_type_c_probe' could possibly return null pointer.
> Add a null check for the returned pointer.

How do you think about a wording variant like the following?

  A devm_kzalloc() call can fail in this function implementation.
  Thus add a null pointer check and corresponding exception handling
  for the data structure member “type_c_cfg”.


…
> Signed-off-by: Zichen Xie <zichenxie0106@gmail.com>
> Reported-by: Zichen Xie <zichenxie0106@gmail.com>
…

How good does such a tag combination fit together for the same person?
https://git.kernel.org/pub/scm/linux/kernel/git/torvalds/linux.git/tree/Documentation/process/submitting-patches.rst?h=v6.11#n525


Can a subject like “[PATCH] extcon: realtek: Prevent null pointer dereference in extcon_rtk_type_c_probe()”
be more appropriate?
https://git.kernel.org/pub/scm/linux/kernel/git/torvalds/linux.git/tree/Documentation/process/submitting-patches.rst?h=v6.11#n613


…
> +++ b/drivers/extcon/extcon-rtk-type-c.c
> @@ -1371,6 +1371,12 @@ static int extcon_rtk_type_c_probe(struct platform_device *pdev)
>
>  	type_c->type_c_cfg = devm_kzalloc(dev, sizeof(*type_c_cfg), GFP_KERNEL);
>
> +	if (!type_c->type_c_cfg) {
> +		dev_err(dev, "Failed to allocate memory for type_c_cfg!\n");
> +		ret = -ENOMEM;
> +		goto err;
> +	}
> +
>  	memcpy(type_c->type_c_cfg, type_c_cfg, sizeof(*type_c_cfg));
…


I suggest to omit an extra error message here.


By the way:
Would you become interested to add a jump target for the statement “ret = -EINVAL;”
(in two other if branches)?

Regards,
Markus