Received-SPF: pass (google.com: domain of linux-kernel-owner@vger.kernel.org designates 23.128.96.18 as permitted sender) client-ip=23.128.96.18;
Mime-Version: 1.0
Content-Transfer-Encoding: quoted-printable
Content-Type: text/plain; charset=UTF-8
Date:   Tue, 27 Jul 2021 20:07:30 -0400
Message-Id: <CD4C37PEMPOM.91UZ60Q6534Q@shaak>
Cc:     <linux-kernel@vger.kernel.org>, <linux-iio@vger.kernel.org>,
        <devicetree@vger.kernel.org>, <robh+dt@kernel.org>
Subject: Re: [PATCH v6 08/13] iio: afe: rescale: reduce risk of integer
 overflow
From:   "Liam Beguin" <liambeguin@gmail.com>
To:     "Peter Rosin" <peda@axentia.se>, <jic23@kernel.org>,
        <lars@metafoo.de>, <pmeerw@pmeerw.net>
References: <20210721030613.3105327-1-liambeguin@gmail.com>
 <20210721030613.3105327-9-liambeguin@gmail.com>
 <9e0e4398-873e-b5c0-0f0c-50a186ed2228@axentia.se>
In-Reply-To: <9e0e4398-873e-b5c0-0f0c-50a186ed2228@axentia.se>
Precedence: bulk

On Fri Jul 23, 2021 at 5:17 PM EDT, Peter Rosin wrote:
> On 2021-07-21 05:06, Liam Beguin wrote:
> > From: Liam Beguin <lvb@xiphos.com>
> >=20
> > Reduce the risk of integer overflow by doing the scale calculation with
> > 64bit integers and looking for a Greatest Common Divider for both parts
> > of the fractional value when required.
> >=20
> > Signed-off-by: Liam Beguin <lvb@xiphos.com>
> > ---
> >  drivers/iio/afe/iio-rescale.c | 15 ++++++++++++---
> >  1 file changed, 12 insertions(+), 3 deletions(-)
> >=20
> > diff --git a/drivers/iio/afe/iio-rescale.c b/drivers/iio/afe/iio-rescal=
e.c
> > index 6f6a711ae3ae..35fa3b4e53e0 100644
> > --- a/drivers/iio/afe/iio-rescale.c
> > +++ b/drivers/iio/afe/iio-rescale.c
> > @@ -21,12 +21,21 @@
> >  int rescale_process_scale(struct rescale *rescale, int scale_type,
> >  			  int *val, int *val2)
> >  {
> > -	unsigned long long tmp;
> > +	s64 tmp, tmp2;
> > +	u32 factor;
> > =20
> >  	switch (scale_type) {
> >  	case IIO_VAL_FRACTIONAL:
> > -		*val *=3D rescale->numerator;
> > -		*val2 *=3D rescale->denominator;
> > +		if (check_mul_overflow(*val, rescale->numerator, (s32 *)&tmp) ||
> > +		    check_mul_overflow(*val2, rescale->denominator, (s32 *)&tmp2)) {
> > +			tmp =3D (s64)*val * rescale->numerator;
> > +			tmp2 =3D (s64)*val2 * rescale->denominator;
> > +			factor =3D gcd(tmp, tmp2);

Hi Peter,

>
> Hi!
>
> Reiterating that gcd() only works for unsigned operands, so this is
> broken for
> negative values.

Apologies, I didn't mean to make it seem like I ignored your comments. I
should've added a note. After you pointed out that gcd() only works for
unsigned elements, I added test cases for negative values, and all tests
passed. I'll look into it more.

rescale_voltage_divider_props() seems to also use gcd() with signed
integers.

Thanks,
Liam

>
> Cheers,
> Peter
>
> > +			tmp =3D div_s64(tmp, factor);
> > +			tmp2 =3D div_s64(tmp2, factor);
> > +		}
> > +		*val =3D tmp;
> > +		*val2 =3D tmp2;
> >  		return scale_type;
> >  	case IIO_VAL_INT:
> >  		*val *=3D rescale->numerator;
> >=20