[lm-sensors] [PATCH v2 2/2] hwmon: (lm92) Prevent overflow problem when writing large limits

[lm-sensors] [PATCH v2 2/2] hwmon: (lm92) Prevent overflow problem when writing large limits

[Axel Lin]

On platforms with sizeof(int) < sizeof(long), writing a temperature
limit larger than MAXINT will result in unpredictable limit values
written to the chip. Avoid auto-conversion from long to int to fix
the problem.
For set_temp_hyst, clamp the input values to the supported limits first
to fix the problem.
Also uses clamp_val in TEMP_TO_REG to simplify the code a bit.

Signed-off-by: Axel Lin <axel.lin@ingics.com>
---
 drivers/hwmon/lm92.c | 8 +++-----
 1 file changed, 3 insertions(+), 5 deletions(-)

diff --git a/drivers/hwmon/lm92.c b/drivers/hwmon/lm92.c
index d2060e2..481abe2 100644
--- a/drivers/hwmon/lm92.c
+++ b/drivers/hwmon/lm92.c
@@ -74,12 +74,9 @@ static inline int TEMP_FROM_REG(s16 reg)
 	return reg / 8 * 625 / 10;
 }
 
-static inline s16 TEMP_TO_REG(int val)
+static inline s16 TEMP_TO_REG(long val)
 {
-	if (val <= -60000)
-		return -60000 * 10 / 625 * 8;
-	if (val >= 160000)
-		return 160000 * 10 / 625 * 8;
+	val = clamp_val(val, -60000, 160000);
 	return val * 10 / 625 * 8;
 }
 
@@ -206,6 +203,7 @@ static ssize_t set_temp_hyst(struct device *dev,
 	if (err)
 		return err;
 
+	val = clamp_val(val, -60000, 160000);
 	mutex_lock(&data->update_lock);
 	data->temp[t_hyst] = TEMP_FROM_REG(data->temp[attr->index]) - val;
 	i2c_smbus_write_word_swapped(client, LM92_REG_TEMP_HYST,
-- 
1.9.1




_______________________________________________
lm-sensors mailing list
lm-sensors@lm-sensors.org
http://lists.lm-sensors.org/mailman/listinfo/lm-sensors

Re: [lm-sensors] [PATCH v2 2/2] hwmon: (lm92) Prevent overflow problem when writing large limits

[Guenter Roeck]

On Thu, Jul 31, 2014 at 03:31:54PM +0800, Axel Lin wrote:
> On platforms with sizeof(int) < sizeof(long), writing a temperature
> limit larger than MAXINT will result in unpredictable limit values
> written to the chip. Avoid auto-conversion from long to int to fix
> the problem.
> For set_temp_hyst, clamp the input values to the supported limits first
> to fix the problem.
> Also uses clamp_val in TEMP_TO_REG to simplify the code a bit.
> 
> Signed-off-by: Axel Lin <axel.lin@ingics.com>
> ---
>  drivers/hwmon/lm92.c | 8 +++-----
>  1 file changed, 3 insertions(+), 5 deletions(-)
> 
> diff --git a/drivers/hwmon/lm92.c b/drivers/hwmon/lm92.c
> index d2060e2..481abe2 100644
> --- a/drivers/hwmon/lm92.c
> +++ b/drivers/hwmon/lm92.c
> @@ -74,12 +74,9 @@ static inline int TEMP_FROM_REG(s16 reg)
>  	return reg / 8 * 625 / 10;
>  }
>  
> -static inline s16 TEMP_TO_REG(int val)
> +static inline s16 TEMP_TO_REG(long val)
>  {
> -	if (val <= -60000)
> -		return -60000 * 10 / 625 * 8;
> -	if (val >= 160000)
> -		return 160000 * 10 / 625 * 8;
> +	val = clamp_val(val, -60000, 160000);
>  	return val * 10 / 625 * 8;
>  }
>  
> @@ -206,6 +203,7 @@ static ssize_t set_temp_hyst(struct device *dev,
>  	if (err)
>  		return err;
>  
> +	val = clamp_val(val, -60000, 160000);
>  	mutex_lock(&data->update_lock);
>  	data->temp[t_hyst] = TEMP_FROM_REG(data->temp[attr->index]) - val;

Hi Axel,

unfortunately this function is all wrong to start with.
Problem is that data->temp[t_hyst] is supposed to store the
register value, not the converted value, meaning the above clamp
does not really do any good. We also need something like

  	data->temp[t_hyst] = TEMP_TO_REG(TEMP_FROM_REG(data->temp[attr->index]) - val);

Also, the hysteresis temperature range isn't really [-60000, 160000],
but depends on the value of data->temp[attr->index], since val is
subtracted from it.  One solution for that problem might be to use
a wider clamp. [-120000, 220000] should do to cover the possible range.
TEMP_TO_REG has its own clamp, so it will fix any range error introduced
by the subtraction.

Can you resubmit with those changes ?

Thanks,
Guenter

_______________________________________________
lm-sensors mailing list
lm-sensors@lm-sensors.org
http://lists.lm-sensors.org/mailman/listinfo/lm-sensors