linux-kernel - Re: [PATCH v10 6/7] iio: accel: adxl345: extend inactivity time for less than 1s

lists.openwall.net		lists / announce owl-users owl-dev john-users john-dev passwdqc-users yescrypt popa3d-users / oss-security kernel-hardening musl sabotage tlsify passwords / crypt-dev xvendor / Bugtraq Full-Disclosure linux-kernel linux-netdev linux-ext4 linux-hardening linux-cve-announce PHC
Open Source and information security mailing list archives

Hash Suite: Windows password security audit tool. GUI, reports in PDF.

[<prev] [next>] [<thread-prev] [thread-next>] [day] [month] [year] [list]

Message-ID: <CAFXKEHYhXekE29Ljfv=c7oRuzo0irWtJNM7fjW516xQ-ydsm=Q@mail.gmail.com>
Date: Tue, 24 Jun 2025 10:18:35 +0200
From: Lothar Rubusch <l.rubusch@...il.com>
To: Andy Shevchenko <andriy.shevchenko@...el.com>
Cc: lars@...afoo.de, Michael.Hennerich@...log.com, jic23@...nel.org, 
	dlechner@...libre.com, nuno.sa@...log.com, andy@...nel.org, corbet@....net, 
	linux-iio@...r.kernel.org, linux-kernel@...r.kernel.org, 
	linux-doc@...r.kernel.org, eraretuya@...il.com
Subject: Re: [PATCH v10 6/7] iio: accel: adxl345: extend inactivity time for
 less than 1s

On Tue, Jun 24, 2025 at 9:38 AM Andy Shevchenko
<andriy.shevchenko@...el.com> wrote:
>
> On Mon, Jun 23, 2025 at 11:21:01PM +0200, Lothar Rubusch wrote:
> > On Mon, Jun 23, 2025 at 12:17 PM Andy Shevchenko
> > <andriy.shevchenko@...el.com> wrote:
> > > On Sun, Jun 22, 2025 at 03:50:09PM +0000, Lothar Rubusch wrote:
>
> ...
>
> > > > -static int adxl345_set_inact_time(struct adxl345_state *st, u32 val_s)
> > > > +static int adxl345_set_inact_time(struct adxl345_state *st, u32 val_int,
> > > > +                               u32 val_fract)
> > > >  {
> > > >       int max_boundary = U8_MAX;
> > > >       int min_boundary = 10;
> > > > -     unsigned int val = min(val_s, U8_MAX);
> > > > +     unsigned int val;
> > >
> > > You see, I even suggested splitting this assignment to begin with.
> > > The change will be clearer with that done.
> > >
> > > >       enum adxl345_odr odr;
> > > >       unsigned int regval;
> > > >       int ret;
> > > >
> > > > -     if (val == 0) {
> > > > +     if (val_int == 0 && val_fract == 0) {
> >
> > The case for 0sec, 0.0 or setting "0" and fract will consequently be
> > "0". 0 is an invalid input for this period and sensor, so it will
> > default to an optimized period based on given ODR.
> >
> > > > +             /* Generated inactivity time based on ODR */
> > > >               ret = regmap_read(st->regmap, ADXL345_REG_BW_RATE, &regval);
> > > >               if (ret)
> > > >                       return ret;
> > >
> > > >               odr = FIELD_GET(ADXL345_BW_RATE_MSK, regval);
> > > >               val = clamp(max_boundary - adxl345_odr_tbl[odr][0],
> > > >                           min_boundary, max_boundary);
> > > > +             st->inact_time_ms = MILLI * val;
> > > > +
> > > > +             /* Inactivity time in s */
> > > > +             return regmap_write(st->regmap, ADXL345_REG_TIME_INACT, val);
> > > > +     } else if (val_int == 0 && val_fract > 0) {
> > >
> > > val_fract check is not needed here.
> >
> > Case for e.g. 0.123, numbers under 1s. This goes into the free-fall register.
>
> 0.0 is already checked above, and since the val_fract is unsigned this is check
> is redundant.
>
> > > > +             /* time < 1s, free-fall */
> > > > +
> > > > +             /*
> > > > +              * Datasheet max. value is 255 * 5000 us = 1.275000 seconds.
> > > > +              *
> > > > +              * Recommended values between 100ms and 350ms (0x14 to 0x46)
> > > > +              */
> > > > +             st->inact_time_ms = DIV_ROUND_UP(val_fract, MILLI);
> > > > +
> > > > +             return regmap_write(st->regmap, ADXL345_REG_TIME_FF,
> > > > +                                 DIV_ROUND_CLOSEST(val_fract, 5));
> > > > +     } else if (val_int > 0) {
> > >
> > > if now is redundant here, right?
> >
> > So, this will be 1s through 255s. Periods above 1sec. This goes into
> > the inactivity register.
>
> See above,
>

I agree, that checking for val_fract is actually done as a sub case of
val_int, and only if val_int was 0. So, would the following make it
clearer?

if (val_int  == 0) {
    if (val_fract == 0) {
        // 0 provided, default values
    } else {
        // >0s, e.g. 0.123s, use free-fall register
} else {
    // 1s - 255s, use inactivity register
}

Actually - I did not touch that - I saw some places where I'm already
using nested if/else in the third level. I guess, by the style advice
according to switch/case, this also applies to if/else, right?

If yes, when the according parts go into another round, I might give
it a try to separate as well using helper functions.

Best,
L

> > > > +             /* Time >= 1s, inactivity */
> > > > +             st->inact_time_ms = MILLI * val_int;
> > > > +
> > > > +             return regmap_write(st->regmap, ADXL345_REG_TIME_INACT, val_int);
> > > >       }
> > > >
> > > > -     return regmap_write(st->regmap, ADXL345_REG_TIME_INACT, val);
> > > > +     /* Do not support negative or wrong input. */
> > > > +     return -EINVAL;
> > > >  }
>
> --
> With Best Regards,
> Andy Shevchenko
>
>