Re: [PATCH] hwmon: (ads1015) Handle negative conversion values correctly

[Guenter Roeck <linux@roeck-us.net>]

On Fri, Feb 19, 2016 at 12:13:43AM +0100, Peter Rosin wrote:
> Hi!
> 
> Sorry for the noise, but...
> 
> On 2016-02-18 14:07, Peter Rosin wrote:
> > From: Peter Rosin <peda@axentia.se>
> > 
> > Make the divisor signed as DIV_ROUND_CLOSEST is undefined for negative
> > dividends when the divisor is unsigned.
> > 
> > Signed-off-by: Peter Rosin <peda@axentia.se>
> 
> ...I forgot to add this to the commit message
> Cc: stable@vger.kernel.org
> 
No problem, I prefer to do that myself anyway.

Patch applied.

Thanks,
Guenter

> I also didn't explicitly mention any example platform where this
> bug causes real problems; it was triggered on ARM (atmel sama5d).
> 
> Cheers,
> Peter
> 
> > ---
> >  drivers/hwmon/ads1015.c | 2 +-
> >  1 file changed, 1 insertion(+), 1 deletion(-)
> > 
> > diff --git a/drivers/hwmon/ads1015.c b/drivers/hwmon/ads1015.c
> > index f155b8380481..2b3105c8aed3 100644
> > --- a/drivers/hwmon/ads1015.c
> > +++ b/drivers/hwmon/ads1015.c
> > @@ -126,7 +126,7 @@ static int ads1015_reg_to_mv(struct i2c_client *client, unsigned int channel,
> >  	struct ads1015_data *data = i2c_get_clientdata(client);
> >  	unsigned int pga = data->channel_data[channel].pga;
> >  	int fullscale = fullscale_table[pga];
> > -	const unsigned mask = data->id == ads1115 ? 0x7fff : 0x7ff0;
> > +	const int mask = data->id == ads1115 ? 0x7fff : 0x7ff0;
> >  
> >  	return DIV_ROUND_CLOSEST(reg * fullscale, mask);
> >  }
> >