rtc: atcrtc100: Fix signedness bug in probe()

[PATCH next] rtc: atcrtc100: Fix signedness bug in probe()

Posted by Dan Carpenter 2 months, 3 weeks ago

The "atcrtc_dev->alarm_irq" variable is an unsigned int but it needs to
be signed for the error handling to work.  Use the "ret" variable
instead.

Fixes: 7adca706fe16 ("rtc: atcrtc100: Add ATCRTC100 RTC driver")
Signed-off-by: Dan Carpenter <dan.carpenter@linaro.org>
---
 drivers/rtc/rtc-atcrtc100.c | 8 +++++---
 1 file changed, 5 insertions(+), 3 deletions(-)

diff --git a/drivers/rtc/rtc-atcrtc100.c b/drivers/rtc/rtc-atcrtc100.c
index 51933ae1a2fa..9808fc2c5a49 100644
--- a/drivers/rtc/rtc-atcrtc100.c
+++ b/drivers/rtc/rtc-atcrtc100.c
@@ -296,10 +296,12 @@ static int atcrtc_probe(struct platform_device *pdev)
 				     "Failed to initialize RTC: unsupported hardware ID 0x%x\n",
 				     rtc_id);
 
-	atcrtc_dev->alarm_irq = platform_get_irq(pdev, 1);
-	if (atcrtc_dev->alarm_irq < 0)
-		return dev_err_probe(&pdev->dev, atcrtc_dev->alarm_irq,
+	ret = platform_get_irq(pdev, 1);
+	if (ret < 0)
+		return dev_err_probe(&pdev->dev, ret,
 				     "Failed to get IRQ for alarm\n");
+	atcrtc_dev->alarm_irq = ret;
+
 	ret = devm_request_irq(&pdev->dev,
 			       atcrtc_dev->alarm_irq,
 			       atcrtc_alarm_isr,
-- 
2.51.0

Re: [PATCH next] rtc: atcrtc100: Fix signedness bug in probe()

Posted by Alexandre Belloni 2 months ago

On Tue, 18 Nov 2025 13:48:56 +0300, Dan Carpenter wrote:
> The "atcrtc_dev->alarm_irq" variable is an unsigned int but it needs to
> be signed for the error handling to work.  Use the "ret" variable
> instead.
> 
> 

Applied, thanks!

[1/1] rtc: atcrtc100: Fix signedness bug in probe()
      https://git.kernel.org/abelloni/c/159a740c768e

Best regards,

-- 
Alexandre Belloni, co-owner and COO, Bootlin
Embedded Linux and Kernel engineering
https://bootlin.com

Re: [PATCH next] rtc: atcrtc100: Fix signedness bug in probe()

Posted by CL Wang 2 months, 2 weeks ago

On Tue, Nov 18, 2025 at 01:48:56PM +0300, Dan Carpenter wrote:

Hi Dan,

Thank you for pointing out the issue and for providing the fix.

You're absolutely correct that using an unsigned type for alarm_irq
prevents proper error handling when platform_get_irq() returns a
negative value. I will apply your patch and also review other return
value checks to ensure there are no similar issues elsewhere.

Thank you again for your detailed review and suggestions.

Best regards,
CL

> 
> The "atcrtc_dev->alarm_irq" variable is an unsigned int but it needs to
> be signed for the error handling to work.  Use the "ret" variable
> instead.
> 
> Fixes: 7adca706fe16 ("rtc: atcrtc100: Add ATCRTC100 RTC driver")
> Signed-off-by: Dan Carpenter <dan.carpenter@linaro.org>
> ---
>  drivers/rtc/rtc-atcrtc100.c | 8 +++++---
>  1 file changed, 5 insertions(+), 3 deletions(-)
> 
> diff --git a/drivers/rtc/rtc-atcrtc100.c b/drivers/rtc/rtc-atcrtc100.c
> index 51933ae1a2fa..9808fc2c5a49 100644
> --- a/drivers/rtc/rtc-atcrtc100.c
> +++ b/drivers/rtc/rtc-atcrtc100.c
> @@ -296,10 +296,12 @@ static int atcrtc_probe(struct platform_device *pdev)
>                                      "Failed to initialize RTC: unsupported hardware ID 0x%x\n",
>                                      rtc_id);
> 
> -       atcrtc_dev->alarm_irq = platform_get_irq(pdev, 1);
> -       if (atcrtc_dev->alarm_irq < 0)
> -               return dev_err_probe(&pdev->dev, atcrtc_dev->alarm_irq,
> +       ret = platform_get_irq(pdev, 1);
> +       if (ret < 0)
> +               return dev_err_probe(&pdev->dev, ret,
>                                      "Failed to get IRQ for alarm\n");
> +       atcrtc_dev->alarm_irq = ret;
> +
>         ret = devm_request_irq(&pdev->dev,
>                                atcrtc_dev->alarm_irq,
>                                atcrtc_alarm_isr,
> --
> 2.51.0
>

Re: [PATCH next] rtc: atcrtc100: Fix signedness bug in probe()

Posted by Dan Carpenter 2 months, 2 weeks ago

On Wed, Nov 19, 2025 at 10:35:23AM +0800, CL Wang wrote:
> On Tue, Nov 18, 2025 at 01:48:56PM +0300, Dan Carpenter wrote:
> 
> Hi Dan,
> 
> Thank you for pointing out the issue and for providing the fix.
> 
> You're absolutely correct that using an unsigned type for alarm_irq
> prevents proper error handling when platform_get_irq() returns a
> negative value. I will apply your patch and also review other return
> value checks to ensure there are no similar issues elsewhere.
> 
> Thank you again for your detailed review and suggestions.
> 

You're welcome.  These are just Smatch warnings.  Nothing super
major.

https://github.com/error27/smatch
https://github.com/error27/smatch/blob/master/Documentation/smatch.rst

regards,
dan carpenter