2022-08-17 06:21:40

by Farber, Eliav

Subject: [PATCH v2 11/16] hwmon: (mr75203) add protection for negative voltage value

This change makes sure the returned voltage vlaue is 0 or positive.

Signed-off-by: Eliav Farber <[email protected]>
---
drivers/hwmon/mr75203.c | 7 +++++++
1 file changed, 7 insertions(+)

diff --git a/drivers/hwmon/mr75203.c b/drivers/hwmon/mr75203.c
index 24a00339cfd8..e3191f590167 100644
--- a/drivers/hwmon/mr75203.c
+++ b/drivers/hwmon/mr75203.c
@@ -218,6 +218,13 @@ static int pvt_read_in(struct device *dev, u32 attr, int channel, long *val)
return ret;

n &= SAMPLE_DATA_MSK;
+
+ /* Voltage can't be negative */
+ if (PVT_N_CONST * n < PVT_R_CONST) {
+ *val = 0;
+ return 0;
+ }
+
/* Convert the N bitstream count into voltage */
*val = pvt->vd[channel].pre_scaler;
*val *= (PVT_N_CONST * n - PVT_R_CONST);
--
2.37.1

2022-08-18 21:24:52

by Guenter Roeck

Subject: Re: [PATCH v2 11/16] hwmon: (mr75203) add protection for negative voltage value

On Wed, Aug 17, 2022 at 05:43:16AM +0000, Eliav Farber wrote:
> This change makes sure the returned voltage vlaue is 0 or positive.
>
> Signed-off-by: Eliav Farber <[email protected]>
> ---
> drivers/hwmon/mr75203.c | 7 +++++++
> 1 file changed, 7 insertions(+)
>
> diff --git a/drivers/hwmon/mr75203.c b/drivers/hwmon/mr75203.c
> index 24a00339cfd8..e3191f590167 100644
> --- a/drivers/hwmon/mr75203.c
> +++ b/drivers/hwmon/mr75203.c
> @@ -218,6 +218,13 @@ static int pvt_read_in(struct device *dev, u32 attr, int channel, long *val)
> return ret;
>
> n &= SAMPLE_DATA_MSK;
> +
> + /* Voltage can't be negative */

Who says, and what does that mean ? Under which conditions would
the value be negative, and why would that be a problem / bug ?
After all, negative voltages do exist.

Guenter

> + if (PVT_N_CONST * n < PVT_R_CONST) {
> + *val = 0;
> + return 0;
> + }
> +
> /* Convert the N bitstream count into voltage */
> *val = pvt->vd[channel].pre_scaler;
> *val *= (PVT_N_CONST * n - PVT_R_CONST);