linux-kernel - Re: [PATCH 01/10] hwmon: (amc6821) Stop accepting invalid pwm values

lists.openwall.net		lists / announce owl-users owl-dev john-users john-dev passwdqc-users yescrypt popa3d-users / oss-security kernel-hardening musl sabotage tlsify passwords / crypt-dev xvendor / Bugtraq Full-Disclosure linux-kernel linux-netdev linux-ext4 linux-hardening linux-cve-announce PHC
Open Source and information security mailing list archives

Hash Suite: Windows password security audit tool. GUI, reports in PDF.

[<prev] [next>] [<thread-prev] [thread-next>] [day] [month] [year] [list]

Date: Mon, 1 Jul 2024 12:19:30 +0200
From: Quentin Schulz <quentin.schulz@...rry.de>
To: Guenter Roeck <linux@...ck-us.net>, linux-hwmon@...r.kernel.org
Cc: linux-kernel@...r.kernel.org, Farouk Bouabid <farouk.bouabid@...rry.de>
Subject: Re: [PATCH 01/10] hwmon: (amc6821) Stop accepting invalid pwm values

Hi Guenter,

On 6/28/24 5:13 PM, Guenter Roeck wrote:
> The pwm value range is well defined from 0..255. Don't accept
> any values outside this range.
> 
> Signed-off-by: Guenter Roeck <linux@...ck-us.net>
> ---
>   drivers/hwmon/amc6821.c | 14 ++++++++++----
>   1 file changed, 10 insertions(+), 4 deletions(-)
> 
> diff --git a/drivers/hwmon/amc6821.c b/drivers/hwmon/amc6821.c
> index 9b02b304c2f5..3c614a0bd192 100644
> --- a/drivers/hwmon/amc6821.c
> +++ b/drivers/hwmon/amc6821.c
> @@ -360,8 +360,11 @@ static ssize_t pwm1_store(struct device *dev,
>   	if (ret)
>   		return ret;
>   
> +	if (val < 0 || val > 255)
> +		return -EINVAL;
> +

Why not use kstrtoul to avoid having to check for negative val? The same 
way that is done just below in this patch?

Additionally, why not using kstrtou8 so we don't have to do this check 
ourselves in the driver?

>   	mutex_lock(&data->update_lock);
> -	data->pwm1 = clamp_val(val , 0, 255);
> +	data->pwm1 = val;
>   	i2c_smbus_write_byte_data(client, AMC6821_REG_DCY, data->pwm1);
>   	mutex_unlock(&data->update_lock);
>   	return count;
> @@ -558,13 +561,16 @@ static ssize_t pwm1_auto_point_pwm_store(struct device *dev,
>   	struct amc6821_data *data = dev_get_drvdata(dev);
>   	struct i2c_client *client = data->client;
>   	int dpwm;
> -	long val;
> -	int ret = kstrtol(buf, 10, &val);
> +	unsigned long val;
> +	int ret = kstrtoul(buf, 10, &val);

Same remark concerning kstrtou8 use.

>   	if (ret)
>   		return ret;
>   
> +	if (val > 255)
> +		return -EINVAL;
> +
>   	mutex_lock(&data->update_lock);
> -	data->pwm1_auto_point_pwm[1] = clamp_val(val, 0, 254);

We're suddenly allowing 255 as a valid value.

I don't see 255 triggering an obvious divide-by-0 issue in the code, nor 
any limitation from a quick look at the datasheet. 254 was introduced in 
the introducing commit, as well as the other 255... so probably an 
oversight by the original author? In any case, I would make this a 
separate commit or at the very least make this explicit in the commit 
log to show this isn't an oversight **right now** and that this change 
was desired.

Cheers,
Quentin