Re: extending /sys/.../iio:deviceX/in_accelX_power_mode

[Martin Kepplinger]

Am 01.03.2016 10:53 schrieb Crt Mori:
> On 1 March 2016 at 10:47, Martin Kepplinger <martink@xxxxxxxxx> wrote:
> > Am 2016-03-01 um 10:38 schrieb Daniel Baluta:
> > > On Tue, Mar 1, 2016 at 10:59 AM, Martin Kepplinger 
> > > <martink@xxxxxxxxx> wrote:
> > > > Would it be ok, if adding in_accelX_power_mode to a driver, to 
> > > > extend it
> > > > so that in_accel_power_mode_available offers:
> > > >
> > > > low_noise low_power low_power_low_noise normal
> > > >
> > > > if there's a default "normal" mode, plus options to increase or 
> > > > decrease
> > > > oversampling / power consumption for my device?
> > > >
> > > > Specifically I'm unsure about "low_power_low_noise" being enough
> > > > user-friendly. The chip I work with just happens to offer these 4 
> > > > modes.
> > > > Would you leave out "low_power_low_noise" and go with
> > > >
> > > > low_noise low_power normal
> > > >
> > > > or is it not even desired to add "normal" to the list?
> > > >
> > > > Although strictly not necessary, I would add any new addition to the
> > > > Documentation as well.
> > >
> > > The problem with this is that is not uniform across sensors. What
> > > chip are you looking at?
> > >
> > > For example INV6500 has:
> > > * sleep mode
> > > * standby mode
> > > * etc.
> > >
> > > Daniel.
> >
> > I suspect these modes are something else. I'm looking at the mma8452
> > driver, and it also has "active" "standby" and "sleep" modes, but I'm
> > talking about different *power* (oversampling) configurations in
> > "active" mode, which is what said sysfs file is about.
> >
> > But yes, it should be potenially uniform across sensors, which is why 
> > I
> > would probably only add "normal" to the list. At least I can imagine
> > that many devices have an oversampling mode called "normal".
>
> If that is oversampling option then why don't you just use that as a
> setup? Power mode does not sound like oversampling to me... Maybe you
> should use a sampling_frequency parameter instead?

well, it doesn't affect the sampling frequency. Oversampling is a way 
chips get
more accurate values and use more power. But it's fine. It comes down to
trying patches and see what happens anyways :)

> > A simple user interface is important so right now I think the best is 
> > to
> > leave it as it is, and not to add complexity and every possible option
> > for the user.
> > --
> > To unsubscribe from this list: send the line "unsubscribe linux-iio" 
> > in
> > the body of a message to majordomo@xxxxxxxxxxxxxxx
> > More majordomo info at  http://vger.kernel.org/majordomo-info.html