From: Liam Beguin <liambeguin@gmail.com>
To: Peter Rosin <peda@axentia.se>
Cc: jic23@kernel.org, lars@metafoo.de, linux-kernel@vger.kernel.org,
linux-iio@vger.kernel.org, devicetree@vger.kernel.org,
robh+dt@kernel.org
Subject: Re: [PATCH v8 08/14] iio: afe: rescale: reduce risk of integer overflow
Date: Sun, 29 Aug 2021 00:01:07 -0400 [thread overview]
Message-ID: <YSsGgwlnO7Wd7BDS@shaak> (raw)
In-Reply-To: <1e18e5b3-9335-ca55-3506-4b4cb19d5661@axentia.se>
On Thu, Aug 26, 2021 at 11:13:38AM +0200, Peter Rosin wrote:
> On 2021-08-20 21:17, Liam Beguin wrote:
> > From: Liam Beguin <lvb@xiphos.com>
> >
> > Reduce the risk of integer overflow by doing the scale calculation on
> > a 64-bit integer. Since the rescaling is only performed on *val, reuse
> > the IIO_VAL_FRACTIONAL_LOG2 case.
>
> While this patch certainly helps with overflow problems, it also
> potentially kills precision in some cases where there currently are
> no overflow issues.
>
> E.g. this patch transforms 5/32768 scaled by 3/10000 from the exact
>
> 15 / 327680000 (0.0000000457763671875)
>
> to the heavily truncated plain old sorry "zero".
>
> Sure, 9/14 improves the situation, but patch 9/14 simply cannot
> make this example any better than returning 2 significant digits
> since the value is so small.
The 100 ppm check introduced in 09/14 is really objective and might not
be the best choice. Changing it to
- if (abs(rem) > 10000000 && abs(div64_s64(*val, tmp)) < 100) {
+ if (abs(rem)) {
Helps with the precision issues you brought up here, and in 09/14.
I was originally trying to keep the original scale as much as possible,
I'll continue the rest of the discussion on the 09/14 thread we already
have.
>
> Side note, there is also the same type of risk of overflow for
> IIO_VAL_INT. Why does that case not get the same treatment as
> IIO_VAL_FRACTIONAL?
>
Being totally honest, I noticed we have the same issue with IIO_VAL_INT,
but since I didn't run into the issue on my setup I left it out to focus
on getting the rest cleaned up.
I guess it couldn't hurt to fix that too while we're at it.
I'll work on it!
> But again, I see no elegant solution. The best I can think of is the
> inelegant solution to provide extra info on the input range, the
> exact desired scaling method, the desired output type, some mix of
> all of the above or something else that helps determining the
> appropriate scaling method w/o looking at the individual number.
I don't really like having to add a range parameter.
If changing the scale type dynamically isn't an issue, I think we can
get away with not adding a parameter.
If it is an issue, we might have to look into it...
Thanks,
Liam
>
> Cheers,
> Peter
>
> > Signed-off-by: Liam Beguin <lvb@xiphos.com>
> > ---
> > drivers/iio/afe/iio-rescale.c | 5 +----
> > 1 file changed, 1 insertion(+), 4 deletions(-)
> >
> > diff --git a/drivers/iio/afe/iio-rescale.c b/drivers/iio/afe/iio-rescale.c
> > index 809e966f7058..c408c4057c08 100644
> > --- a/drivers/iio/afe/iio-rescale.c
> > +++ b/drivers/iio/afe/iio-rescale.c
> > @@ -27,16 +27,13 @@ int rescale_process_scale(struct rescale *rescale, int scale_type,
> > u32 neg;
> >
> > switch (scale_type) {
> > - case IIO_VAL_FRACTIONAL:
> > - *val *= rescale->numerator;
> > - *val2 *= rescale->denominator;
> > - return scale_type;
> > case IIO_VAL_INT:
> > *val *= rescale->numerator;
> > if (rescale->denominator == 1)
> > return scale_type;
> > *val2 = rescale->denominator;
> > return IIO_VAL_FRACTIONAL;
> > + case IIO_VAL_FRACTIONAL:
> > case IIO_VAL_FRACTIONAL_LOG2:
> > tmp = (s64)*val * 1000000000LL;
> > tmp = div_s64(tmp, rescale->denominator);
> >
>
next prev parent reply other threads:[~2021-08-29 4:01 UTC|newest]
Thread overview: 39+ messages / expand[flat|nested] mbox.gz Atom feed top
2021-08-20 19:17 [PATCH v8 00/14] iio: afe: add temperature rescaling support Liam Beguin
2021-08-20 19:17 ` [PATCH v8 01/14] iio: inkern: apply consumer scale on IIO_VAL_INT cases Liam Beguin
2021-08-20 19:17 ` [PATCH v8 02/14] iio: inkern: apply consumer scale when no channel scale is available Liam Beguin
2021-08-20 19:17 ` [PATCH v8 03/14] iio: inkern: make a best effort on offset calculation Liam Beguin
2021-08-20 19:17 ` [PATCH v8 04/14] iio: afe: rescale: expose scale processing function Liam Beguin
2021-08-20 19:17 ` [PATCH v8 05/14] iio: afe: rescale: add INT_PLUS_{MICRO,NANO} support Liam Beguin
2021-08-26 8:11 ` Peter Rosin
2021-08-29 2:50 ` Liam Beguin
2021-08-20 19:17 ` [PATCH v8 06/14] iio: afe: rescale: add offset support Liam Beguin
2021-08-20 19:17 ` [PATCH v8 07/14] iio: afe: rescale: use s64 for temporary scale calculations Liam Beguin
2021-08-20 19:17 ` [PATCH v8 08/14] iio: afe: rescale: reduce risk of integer overflow Liam Beguin
2021-08-26 9:13 ` Peter Rosin
2021-08-29 4:01 ` Liam Beguin [this message]
2021-08-20 19:17 ` [PATCH v8 09/14] iio: afe: rescale: fix accuracy for small fractional scales Liam Beguin
2021-08-20 23:37 ` kernel test robot
2021-08-21 1:33 ` Liam Beguin
2021-08-23 6:53 ` Peter Rosin
2021-08-21 2:00 ` kernel test robot
2021-08-21 7:21 ` kernel test robot
2021-08-22 22:18 ` Peter Rosin
2021-08-24 20:28 ` [PATCH v8 09/14] iio: afe: rescale: fix accuracy for small Liam Beguin
2021-08-26 9:53 ` Peter Rosin
2021-08-29 4:41 ` Liam Beguin
2021-08-30 11:27 ` Jonathan Cameron
2021-09-11 23:31 ` Liam Beguin
2021-08-30 13:03 ` Peter Rosin
2021-09-11 23:20 ` Liam Beguin
2021-08-30 11:22 ` [PATCH v8 09/14] iio: afe: rescale: fix accuracy for small fractional scales Jonathan Cameron
2021-08-30 14:30 ` Peter Rosin
2021-08-30 17:03 ` Jonathan Cameron
2021-09-02 2:27 ` Liam Beguin
2021-09-02 21:52 ` Peter Rosin
2021-08-20 19:17 ` [PATCH v8 10/14] iio: test: add basic tests for the iio-rescale driver Liam Beguin
2021-08-20 19:17 ` [PATCH v8 11/14] iio: afe: rescale: add RTD temperature sensor support Liam Beguin
2021-08-20 19:17 ` [PATCH v8 12/14] iio: afe: rescale: add temperature transducers Liam Beguin
2021-08-26 8:56 ` Peter Rosin
2021-08-29 2:33 ` Liam Beguin
2021-08-20 19:17 ` [PATCH v8 13/14] dt-bindings: iio: afe: add bindings for temperature-sense-rtd Liam Beguin
2021-08-20 19:17 ` [PATCH v8 14/14] dt-bindings: iio: afe: add bindings for temperature transducers Liam Beguin
Reply instructions:
You may reply publicly to this message via plain-text email
using any one of the following methods:
* Save the following mbox file, import it into your mail client,
and reply-to-all from there: mbox
Avoid top-posting and favor interleaved quoting:
https://en.wikipedia.org/wiki/Posting_style#Interleaved_style
* Reply using the --to, --cc, and --in-reply-to
switches of git-send-email(1):
git send-email \
--in-reply-to=YSsGgwlnO7Wd7BDS@shaak \
--to=liambeguin@gmail.com \
--cc=devicetree@vger.kernel.org \
--cc=jic23@kernel.org \
--cc=lars@metafoo.de \
--cc=linux-iio@vger.kernel.org \
--cc=linux-kernel@vger.kernel.org \
--cc=peda@axentia.se \
--cc=robh+dt@kernel.org \
/path/to/YOUR_REPLY
https://kernel.org/pub/software/scm/git/docs/git-send-email.html
* If your mail client supports setting the In-Reply-To header
via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line
before the message body.
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox;
as well as URLs for NNTP newsgroup(s).