Groups | Search | Server Info | Keyboard shortcuts | Login | Register [http] [https] [nntp] [nntps]

Groups > linux.kernel > #1662786 > unrolled thread

Re: [PATCH] thermal: int340x: check for sensor when PTYP is missing

Back to article view | Back to linux.kernel

This discussion starts older than the indexed window; earlier articles aren't shown. The article labeled Started by below is the oldest one visible, not the original post.

Contents

  Re: [PATCH] thermal: int340x: check for sensor when PTYP is missing Srinivas Pandruvada <srinivas.pandruvada@linux.intel.com> - 2017-06-10 00:30 +0200
    Re: [PATCH] thermal: int340x: check for sensor when PTYP is missing Andy Shevchenko <andy.shevchenko@gmail.com> - 2017-06-10 00:40 +0200

#1662786 — Re: [PATCH] thermal: int340x: check for sensor when PTYP is missing

On Wed, 2017-06-07 at 13:55 +0300, Andy Shevchenko wrote:
> On Wed, Jun 7, 2017 at 2:00 AM, Srinivas Pandruvada
> <srinivas.pandruvada@linux.intel.com> wrote:
> > 
> > For INT3403 sensor PTYP field is mandatory. But some platforms
> > didn't
> > have this field for sensors. This cause load failure for int3403
> > driver.
> > 
> > This change checks for the presence of _TMP method and if present,
> > then
> > treats this device as a sensor.
> > 
> >         status = acpi_evaluate_integer(priv->adev->handle, "PTYP",
> >                                        NULL, &priv->type);
> >         if (ACPI_FAILURE(status)) {
> > -               result = -EINVAL;
> > -               goto err;
> > 
> > +               unsigned long long tmp;
> You may use &priv->type as temporary variable, though I would go
> other
> way around:
> declare tmp for function, then
> 
>     unsigned long long tmp;
> ...
>     status = acpi_evaluate_integer(priv->adev->handle, "PTYP",  NULL,
> &tmp);
>     if (ACPI_FAILURE(status)) {
>         status = acpi_evaluate_integer(priv->adev->handle, "_TMP",
> NULL, &tmp);
>         if (ACPI_FAILURE(status)) {
>             result = -EINVAL;
>             goto err;
>         }
>         tmp = INT3403_TYPE_SENSOR;
>     }
>     priv->type = tmp;
> 
So what are we saving by doing this way?

Thanks,
Srinivas

> > 
> > +               } else {
> This is redundant.
> 
> > 
> > +                       priv->type = INT3403_TYPE_SENSOR;
> > +               }
> >         }

[toc] | [next] | [standalone]

#1662794

On Sat, Jun 10, 2017 at 1:23 AM, Srinivas Pandruvada
<srinivas.pandruvada@linux.intel.com> wrote:
> On Wed, 2017-06-07 at 13:55 +0300, Andy Shevchenko wrote:
>> On Wed, Jun 7, 2017 at 2:00 AM, Srinivas Pandruvada
>> <srinivas.pandruvada@linux.intel.com> wrote:

>> > This change checks for the presence of _TMP method and if present,
>> > then
>> > treats this device as a sensor.
>> >
>> >         status = acpi_evaluate_integer(priv->adev->handle, "PTYP",
>> >                                        NULL, &priv->type);
>> >         if (ACPI_FAILURE(status)) {
>> > -               result = -EINVAL;
>> > -               goto err;
>> >
>> > +               unsigned long long tmp;
>> You may use &priv->type as temporary variable, though I would go
>> other
>> way around:
>> declare tmp for function, then
>>
>>     unsigned long long tmp;
>> ...
>>     status = acpi_evaluate_integer(priv->adev->handle, "PTYP",  NULL,
>> &tmp);
>>     if (ACPI_FAILURE(status)) {
>>         status = acpi_evaluate_integer(priv->adev->handle, "_TMP",
>> NULL, &tmp);
>>         if (ACPI_FAILURE(status)) {
>>             result = -EINVAL;
>>             goto err;
>>         }
>>         tmp = INT3403_TYPE_SENSOR;
>>     }
>>     priv->type = tmp;
>>
> So what are we saving by doing this way?

We make priv->type assignment in one place.

That's the difference.

So, you may choose your way of course. It's not a big deal.

-- 
With Best Regards,
Andy Shevchenko

[toc] | [prev] | [standalone]

Back to top | Article view | linux.kernel

csiph-web