Re: [PATCH] iio: inkern: Avoid risky abs() usage in iio_multiply_value()

[Andy Shevchenko <andriy.shevchenko@intel.com>]

On Tue, Mar 31, 2026 at 12:29:28PM +0300, Andy Shevchenko wrote:
> On Tue, Mar 31, 2026 at 10:49:59AM +0200, Romain Gantois wrote:
> > iio_multiply_value() passes integers val and val2 directly to abs(). This
> > is problematic because if a signed argument to abs is the lowest value for
> > its type, then the result is undefined due to overflow.
> > 
> > Cast val and val2 to s64 before passing them to abs() to avoid this issue.

...

> > -		*result = multiplier * abs(val);
> > -		*result += div_s64(multiplier * abs(val2), denominator);
> > +		*result = multiplier * abs((s64)val);
> > +		*result += div_s64(multiplier * abs((s64)val2), denominator);
> 
> Right, but here we get val and val2 from either static values from the driver
> (when it is SCALE channel), or when channel has PROCESSED support.
> In the latter one it might theoretically be possible to go till the INT_MIN,
> but practically I don't know how, except for the broken driver code in the
> first place. With that being said, I think it's better to validate somewhere
> the multipliers (when it's SCALE or PROCESSED channel). I also noted that
> for the _PROCESSED some drivers keep a garbage in val2. That probably needs
> to be addressed as well (exempli gratia: bmi270_read_raw() does that).

And start from the test cases actually.

-- 
With Best Regards,
Andy Shevchenko