[PATCH] staging: gpib: Fix memory leak in ni_usb_init()

Zilin Guan posted 1 patch 1 month, 1 week ago
There is a newer version of this series
drivers/gpib/ni_usb/ni_usb_gpib.c | 6 +++++-
1 file changed, 5 insertions(+), 1 deletion(-)
[PATCH] staging: gpib: Fix memory leak in ni_usb_init()
Posted by Zilin Guan 1 month, 1 week ago
In ni_usb_init(), if ni_usb_setup_init() fails, the function returns
immediately without freeing the allocated memory for writes, leading
to a memory leak.

Fix this by jumping to the out label to ensure the memory is properly
freed.

Fixes: 4e127de14fa7 ("staging: gpib: Add National Instruments USB GPIB driver")
Co-developed-by: Jianhao Xu <jianhao.xu@seu.edu.cn>
Signed-off-by: Jianhao Xu <jianhao.xu@seu.edu.cn>
Signed-off-by: Zilin Guan <zilin@seu.edu.cn>
---
 drivers/gpib/ni_usb/ni_usb_gpib.c | 6 +++++-
 1 file changed, 5 insertions(+), 1 deletion(-)

diff --git a/drivers/gpib/ni_usb/ni_usb_gpib.c b/drivers/gpib/ni_usb/ni_usb_gpib.c
index 1f8412de9fa3..4cf45e94c750 100644
--- a/drivers/gpib/ni_usb/ni_usb_gpib.c
+++ b/drivers/gpib/ni_usb/ni_usb_gpib.c
@@ -1802,7 +1802,7 @@ static int ni_usb_init(struct gpib_board *board)
 	if (writes_len)
 		retval = ni_usb_write_registers(ni_priv, writes, writes_len, &ibsta);
 	else
-		return -EFAULT;
+		goto out;
 	kfree(writes);
 	if (retval) {
 		dev_err(&usb_dev->dev, "register write failed, retval=%i\n", retval);
@@ -1810,6 +1810,10 @@ static int ni_usb_init(struct gpib_board *board)
 	}
 	ni_usb_soft_update_status(board, ibsta, 0);
 	return 0;
+
+out:
+	kfree(writes);
+	return -EFAULT;
 }
 
 static void ni_usb_interrupt_complete(struct urb *urb)
-- 
2.34.1
Re: [PATCH] staging: gpib: Fix memory leak in ni_usb_init()
Posted by Dave Penkler 1 month, 1 week ago
On Sun, Dec 28, 2025 at 08:19:26AM +0000, Zilin Guan wrote:
> In ni_usb_init(), if ni_usb_setup_init() fails, the function returns
> immediately without freeing the allocated memory for writes, leading
> to a memory leak.
> 
> Fix this by jumping to the out label to ensure the memory is properly
> freed.
> 
> Fixes: 4e127de14fa7 ("staging: gpib: Add National Instruments USB GPIB driver")
> Co-developed-by: Jianhao Xu <jianhao.xu@seu.edu.cn>
> Signed-off-by: Jianhao Xu <jianhao.xu@seu.edu.cn>
> Signed-off-by: Zilin Guan <zilin@seu.edu.cn>
> ---
>  drivers/gpib/ni_usb/ni_usb_gpib.c | 6 +++++-
>  1 file changed, 5 insertions(+), 1 deletion(-)
> 
> diff --git a/drivers/gpib/ni_usb/ni_usb_gpib.c b/drivers/gpib/ni_usb/ni_usb_gpib.c
> index 1f8412de9fa3..4cf45e94c750 100644
> --- a/drivers/gpib/ni_usb/ni_usb_gpib.c
> +++ b/drivers/gpib/ni_usb/ni_usb_gpib.c
> @@ -1802,7 +1802,7 @@ static int ni_usb_init(struct gpib_board *board)
>  	if (writes_len)
>  		retval = ni_usb_write_registers(ni_priv, writes, writes_len, &ibsta);
>  	else
> -		return -EFAULT;
> +		goto out;
>  	kfree(writes);
>  	if (retval) {
>  		dev_err(&usb_dev->dev, "register write failed, retval=%i\n", retval);
> @@ -1810,6 +1810,10 @@ static int ni_usb_init(struct gpib_board *board)
>  	}
>  	ni_usb_soft_update_status(board, ibsta, 0);
>  	return 0;
> +
> +out:
> +	kfree(writes);
> +	return -EFAULT;
>  }
>  
>  static void ni_usb_interrupt_complete(struct urb *urb)
Good catch.
Prefer simpler variant with check for failure first:
  	if (!writes_len) {
		kfree(writes);
		return -EFAULT;
	}
  	retval = ni_usb_write_registers(ni_priv, writes, writes_len, &ibsta);
  	kfree(writes);
	if (retval) {
	...
cheers,
-Dave
> -- 
> 2.34.1
>
Re: [PATCH] staging: gpib: Fix memory leak in ni_usb_init()
Posted by Zilin Guan 1 month, 1 week ago
On Mon, Dec 29, 2025 at 01:19:22PM+0100, Dave Penkler wrote:
> Good catch.
> Prefer simpler variant with check for failure first:
>   	if (!writes_len) {
> 		kfree(writes);
> 		return -EFAULT;
> 	}
>   	retval = ni_usb_write_registers(ni_priv, writes, writes_len, &ibsta);
>   	kfree(writes);
> 	if (retval) {
> 	...
> cheers,
> -Dave

Thanks for the review and the suggestion. I will adopt this approach in v2.

Regards,
Zilin Guan