IIO: necessity of manual reset attribute in drivers?

IIO: necessity of manual reset attribute in drivers?

[Jonathan Cameron]

Hi All,

I vaguely recall asking this in response to certain drivers over the
years, but what is the usecase for userspace 'resetting' a device via
a reset attribute?  It's reasonably common in the drivers but not
(as I've kept it out) in the abi docs.

It is common in some of Analog's inertial sensor drivers but I think
turns up in various other places.

Jonathan

RE: necessity of manual reset attribute in drivers?

[Hennerich, Michael]

Jonathan Cameron wrote on 2012-04-22:
> Hi All,
>
> I vaguely recall asking this in response to certain drivers over the
> years, but what is the usecase for userspace 'resetting' a device via
> a reset attribute?  It's reasonably common in the drivers but not
> (as I've kept it out) in the abi docs.
>
> It is common in some of Analog's inertial sensor drivers but I think
> turns up in various other places.

Ideally we shouldn't have such an attribute.
Error checking and handling should be all done in the drivers, if possible.
I don't remember exactly why it went into one of the original inertial sensor drivers.
But I think it went there to provide error correction, in case plausibility/error checking
of the delivered data failed.
I also suspect it then got copied over and over again.

Greetings,
Michael

--
Analog Devices GmbH      Wilhelm-Wagenfeld-Str. 6      80807 Muenchen
Sitz der Gesellschaft: Muenchen; Registergericht: Muenchen HRB 40368;
Geschaeftsfuehrer:Dr.Carsten Suckrow, Thomas Wessel, William A. Martin, Margaret Seif

Re: necessity of manual reset attribute in drivers?

[Jonathan Cameron]

On 4/24/2012 8:52 AM, Hennerich, Michael wrote:
> Jonathan Cameron wrote on 2012-04-22:
>> Hi All,
>>
>> I vaguely recall asking this in response to certain drivers over the
>> years, but what is the usecase for userspace 'resetting' a device via
>> a reset attribute?  It's reasonably common in the drivers but not
>> (as I've kept it out) in the abi docs.
>>
>> It is common in some of Analog's inertial sensor drivers but I think
>> turns up in various other places.
> Ideally we shouldn't have such an attribute.
> Error checking and handling should be all done in the drivers, if possible.
> I don't remember exactly why it went into one of the original inertial sensor drivers.
> But I think it went there to provide error correction, in case plausibility/error checking
> of the delivered data failed.
> I also suspect it then got copied over and over again.
That's what I suspected had happened.  Shall we aim to kill it off in 
these drivers
before they move out of staging? I'm always happy to delete code if no 
one else
gets to them first ;)