Re: [PATCH v3] hwmon: (ina2xx) Fix overflow issues

[sashiko-bot@kernel.org]

Thank you for your contribution! Sashiko AI review found 1 potential issue(s) to consider:
- [High] Clamping to LONG_MAX/LONG_MIN before DIV_ROUND_CLOSEST guarantees signed integer overflow/underflow on 32-bit platforms.
--

commit 6fb047c808b5329542311277defe6fdce7b455f0
Author: Guenter Roeck <linux@roeck-us.net>

hwmon: (ina2xx) Fix overflow issues

This commit fixes several integer overflow issues in the ina2xx driver
by upgrading intermediate calculations to 64-bit and adjusting the
clamping limits. It also updates the return type of ina2xx_get_value 
to long to better support 64-bit systems.

> diff --git a/drivers/hwmon/ina2xx.c b/drivers/hwmon/ina2xx.c
> --- a/drivers/hwmon/ina2xx.c
> +++ b/drivers/hwmon/ina2xx.c

[ ... ]

> @@ -283,12 +285,13 @@ static int ina2xx_get_value(struct ina2xx_data *data, u8 reg,
>  		val = DIV_ROUND_CLOSEST(val, 1000);
>  		break;
>  	case INA2XX_POWER:
> -		val = regval * data->power_lsb_uW;
> +		val = clamp_val((u64)regval * data->power_lsb_uW, 0, LONG_MAX);
>  		break;
>  	case INA2XX_CURRENT:
>  		/* signed register, result in mA */
> -		val = ((s16)regval >> data->config->current_shift) *
> +		val64 = (s64)((s16)regval >> data->config->current_shift) *
>  		  data->current_lsb_uA;
> +		val = clamp_val(val64, LONG_MIN, LONG_MAX);
>  		val = DIV_ROUND_CLOSEST(val, 1000);

[Severity: High]
Can clamping to LONG_MAX and LONG_MIN here cause a signed integer overflow
on 32-bit systems?

Since val is clamped to the absolute limits of a signed long, when it is
subsequently passed to DIV_ROUND_CLOSEST(val, 1000), the rounding macro
will add 500 to positive values or subtract 500 from negative values.

Could this operation overflow LONG_MAX and underflow LONG_MIN, causing the
value to wrap around and return the opposite sign instead of the intended
clamped maximum limit?

-- 
Sashiko AI review · https://sashiko.dev/#/patchset/20260611135706.826074-1-linux@roeck-us.net?part=1