diff mbox series

[v3,7/9] iio: cros_ec: use devm_iio_triggered_buffer_setup_ext()

Message ID 20200929125949.69934-8-alexandru.ardelean@analog.com (mailing list archive)
State New
Headers show
Series iio: remove iio_buffer_set_attrs() and assign buffer attrs during alloc | expand

Commit Message

Alexandru Ardelean Sept. 29, 2020, 12:59 p.m. UTC
This change switches to the new devm_iio_triggered_buffer_setup_ext()
function and removes the iio_buffer_set_attrs() call, for assigning the
HW FIFO attributes to the buffer.

Signed-off-by: Alexandru Ardelean <alexandru.ardelean@analog.com>
---
 .../common/cros_ec_sensors/cros_ec_sensors_core.c | 15 +++++++++------
 1 file changed, 9 insertions(+), 6 deletions(-)

Comments

Andy Shevchenko Sept. 29, 2020, 1:08 p.m. UTC | #1
On Tue, Sep 29, 2020 at 3:55 PM Alexandru Ardelean
<alexandru.ardelean@analog.com> wrote:

> This change switches to the new devm_iio_triggered_buffer_setup_ext()
> function and removes the iio_buffer_set_attrs() call, for assigning the
> HW FIFO attributes to the buffer.

Sorry, you were too fast with the version, below one nit.

> Signed-off-by: Alexandru Ardelean <alexandru.ardelean@analog.com>
> ---
>  .../common/cros_ec_sensors/cros_ec_sensors_core.c | 15 +++++++++------
>  1 file changed, 9 insertions(+), 6 deletions(-)
>
> diff --git a/drivers/iio/common/cros_ec_sensors/cros_ec_sensors_core.c b/drivers/iio/common/cros_ec_sensors/cros_ec_sensors_core.c
> index c62cacc04672..1eafcf04ad69 100644
> --- a/drivers/iio/common/cros_ec_sensors/cros_ec_sensors_core.c
> +++ b/drivers/iio/common/cros_ec_sensors/cros_ec_sensors_core.c
> @@ -353,19 +353,22 @@ int cros_ec_sensors_core_init(struct platform_device *pdev,
>                         if (ret)
>                                 return ret;
>                 } else {
> +                       const struct attribute **fifo_attrs;
> +
> +                       if (has_hw_fifo)
> +                               fifo_attrs = cros_ec_sensor_fifo_attributes;
> +                       else
> +                               fifo_attrs = NULL;
> +
>                         /*
>                          * The only way to get samples in buffer is to set a
>                          * software trigger (systrig, hrtimer).
>                          */
> -                       ret = devm_iio_triggered_buffer_setup(

> +                       ret = devm_iio_triggered_buffer_setup_ext(
>                                         dev, indio_dev, NULL, trigger_capture,
> -                                       NULL);
> +                                       NULL, fifo_attrs);

Perhaps it's time to reformat a bit, i.e. move dev to the first line
and do the rest accordingly?

>                         if (ret)
>                                 return ret;
> -
> -                       if (has_hw_fifo)
> -                               iio_buffer_set_attrs(indio_dev->buffer,
> -                                                    cros_ec_sensor_fifo_attributes);
>                 }
>         }
>
> --
> 2.17.1
>
Alexandru Ardelean Sept. 29, 2020, 2:31 p.m. UTC | #2
On Tue, Sep 29, 2020 at 4:09 PM Andy Shevchenko
<andy.shevchenko@gmail.com> wrote:
>
> On Tue, Sep 29, 2020 at 3:55 PM Alexandru Ardelean
> <alexandru.ardelean@analog.com> wrote:
>
> > This change switches to the new devm_iio_triggered_buffer_setup_ext()
> > function and removes the iio_buffer_set_attrs() call, for assigning the
> > HW FIFO attributes to the buffer.
>
> Sorry, you were too fast with the version, below one nit.
>
> > Signed-off-by: Alexandru Ardelean <alexandru.ardelean@analog.com>
> > ---
> >  .../common/cros_ec_sensors/cros_ec_sensors_core.c | 15 +++++++++------
> >  1 file changed, 9 insertions(+), 6 deletions(-)
> >
> > diff --git a/drivers/iio/common/cros_ec_sensors/cros_ec_sensors_core.c b/drivers/iio/common/cros_ec_sensors/cros_ec_sensors_core.c
> > index c62cacc04672..1eafcf04ad69 100644
> > --- a/drivers/iio/common/cros_ec_sensors/cros_ec_sensors_core.c
> > +++ b/drivers/iio/common/cros_ec_sensors/cros_ec_sensors_core.c
> > @@ -353,19 +353,22 @@ int cros_ec_sensors_core_init(struct platform_device *pdev,
> >                         if (ret)
> >                                 return ret;
> >                 } else {
> > +                       const struct attribute **fifo_attrs;
> > +
> > +                       if (has_hw_fifo)
> > +                               fifo_attrs = cros_ec_sensor_fifo_attributes;
> > +                       else
> > +                               fifo_attrs = NULL;
> > +
> >                         /*
> >                          * The only way to get samples in buffer is to set a
> >                          * software trigger (systrig, hrtimer).
> >                          */
> > -                       ret = devm_iio_triggered_buffer_setup(
>
> > +                       ret = devm_iio_triggered_buffer_setup_ext(
> >                                         dev, indio_dev, NULL, trigger_capture,
> > -                                       NULL);
> > +                                       NULL, fifo_attrs);
>
> Perhaps it's time to reformat a bit, i.e. move dev to the first line
> and do the rest accordingly?

this feels like a mix of preferences here;
for once, the patch here [as-is], is the minimal form for this change
[in terms of patch-noise];
so, some people would choose the least noisiest patch;

also, this indentation was chosen [as-is here] from the start [for
this code block];
not sure if it was preferred; i'd suspect it was due to the old 80-col limit;

i'd leave it as-is [for now], or defer the decision to a maintainer to
decide [either IIO or chromium];

>
> >                         if (ret)
> >                                 return ret;
> > -
> > -                       if (has_hw_fifo)
> > -                               iio_buffer_set_attrs(indio_dev->buffer,
> > -                                                    cros_ec_sensor_fifo_attributes);
> >                 }
> >         }
> >
> > --
> > 2.17.1
> >
>
>
> --
> With Best Regards,
> Andy Shevchenko
Jonathan Cameron Sept. 29, 2020, 3:40 p.m. UTC | #3
On Tue, 29 Sep 2020 17:31:55 +0300
Alexandru Ardelean <ardeleanalex@gmail.com> wrote:

> On Tue, Sep 29, 2020 at 4:09 PM Andy Shevchenko
> <andy.shevchenko@gmail.com> wrote:
> >
> > On Tue, Sep 29, 2020 at 3:55 PM Alexandru Ardelean
> > <alexandru.ardelean@analog.com> wrote:
> >  
> > > This change switches to the new devm_iio_triggered_buffer_setup_ext()
> > > function and removes the iio_buffer_set_attrs() call, for assigning the
> > > HW FIFO attributes to the buffer.  
> >
> > Sorry, you were too fast with the version, below one nit.
> >  
> > > Signed-off-by: Alexandru Ardelean <alexandru.ardelean@analog.com>
> > > ---
> > >  .../common/cros_ec_sensors/cros_ec_sensors_core.c | 15 +++++++++------
> > >  1 file changed, 9 insertions(+), 6 deletions(-)
> > >
> > > diff --git a/drivers/iio/common/cros_ec_sensors/cros_ec_sensors_core.c b/drivers/iio/common/cros_ec_sensors/cros_ec_sensors_core.c
> > > index c62cacc04672..1eafcf04ad69 100644
> > > --- a/drivers/iio/common/cros_ec_sensors/cros_ec_sensors_core.c
> > > +++ b/drivers/iio/common/cros_ec_sensors/cros_ec_sensors_core.c
> > > @@ -353,19 +353,22 @@ int cros_ec_sensors_core_init(struct platform_device *pdev,
> > >                         if (ret)
> > >                                 return ret;
> > >                 } else {
> > > +                       const struct attribute **fifo_attrs;
> > > +
> > > +                       if (has_hw_fifo)
> > > +                               fifo_attrs = cros_ec_sensor_fifo_attributes;
> > > +                       else
> > > +                               fifo_attrs = NULL;
> > > +
> > >                         /*
> > >                          * The only way to get samples in buffer is to set a
> > >                          * software trigger (systrig, hrtimer).
> > >                          */
> > > -                       ret = devm_iio_triggered_buffer_setup(  
> >  
> > > +                       ret = devm_iio_triggered_buffer_setup_ext(
> > >                                         dev, indio_dev, NULL, trigger_capture,
> > > -                                       NULL);
> > > +                                       NULL, fifo_attrs);  
> >
> > Perhaps it's time to reformat a bit, i.e. move dev to the first line
> > and do the rest accordingly?  
> 
> this feels like a mix of preferences here;
> for once, the patch here [as-is], is the minimal form for this change
> [in terms of patch-noise];
> so, some people would choose the least noisiest patch;
> 
> also, this indentation was chosen [as-is here] from the start [for
> this code block];
> not sure if it was preferred; i'd suspect it was due to the old 80-col limit;
> 
> i'd leave it as-is [for now], or defer the decision to a maintainer to
> decide [either IIO or chromium];

The indenting of this whole code block is a bit too deep.

Looks to me like we should flip the sense of the outer if statement

if (!physical_device)
	return 0;

That would lead to a whole bunch of reformatting around here including
picking up this.

For now I can just shuffle it a bit whilst applying.

This set isn't likely to make the merge window anyway now as I'd like
it to sit on list a little longer just because it touches several
drivers with active maintainers and I'd like time for them to sanity
check.

Jonathan


> 
> >  
> > >                         if (ret)
> > >                                 return ret;
> > > -
> > > -                       if (has_hw_fifo)
> > > -                               iio_buffer_set_attrs(indio_dev->buffer,
> > > -                                                    cros_ec_sensor_fifo_attributes);
> > >                 }
> > >         }
> > >
> > > --
> > > 2.17.1
> > >  
> >
> >
> > --
> > With Best Regards,
> > Andy Shevchenko
Alexandru Ardelean Nov. 18, 2020, 10:35 a.m. UTC | #4
On Tue, Sep 29, 2020 at 6:40 PM Jonathan Cameron <jic23@kernel.org> wrote:
>
> On Tue, 29 Sep 2020 17:31:55 +0300
> Alexandru Ardelean <ardeleanalex@gmail.com> wrote:
>
> > On Tue, Sep 29, 2020 at 4:09 PM Andy Shevchenko
> > <andy.shevchenko@gmail.com> wrote:
> > >
> > > On Tue, Sep 29, 2020 at 3:55 PM Alexandru Ardelean
> > > <alexandru.ardelean@analog.com> wrote:
> > >
> > > > This change switches to the new devm_iio_triggered_buffer_setup_ext()
> > > > function and removes the iio_buffer_set_attrs() call, for assigning the
> > > > HW FIFO attributes to the buffer.
> > >
> > > Sorry, you were too fast with the version, below one nit.
> > >
> > > > Signed-off-by: Alexandru Ardelean <alexandru.ardelean@analog.com>
> > > > ---
> > > >  .../common/cros_ec_sensors/cros_ec_sensors_core.c | 15 +++++++++------
> > > >  1 file changed, 9 insertions(+), 6 deletions(-)
> > > >
> > > > diff --git a/drivers/iio/common/cros_ec_sensors/cros_ec_sensors_core.c b/drivers/iio/common/cros_ec_sensors/cros_ec_sensors_core.c
> > > > index c62cacc04672..1eafcf04ad69 100644
> > > > --- a/drivers/iio/common/cros_ec_sensors/cros_ec_sensors_core.c
> > > > +++ b/drivers/iio/common/cros_ec_sensors/cros_ec_sensors_core.c
> > > > @@ -353,19 +353,22 @@ int cros_ec_sensors_core_init(struct platform_device *pdev,
> > > >                         if (ret)
> > > >                                 return ret;
> > > >                 } else {
> > > > +                       const struct attribute **fifo_attrs;
> > > > +
> > > > +                       if (has_hw_fifo)
> > > > +                               fifo_attrs = cros_ec_sensor_fifo_attributes;
> > > > +                       else
> > > > +                               fifo_attrs = NULL;
> > > > +
> > > >                         /*
> > > >                          * The only way to get samples in buffer is to set a
> > > >                          * software trigger (systrig, hrtimer).
> > > >                          */
> > > > -                       ret = devm_iio_triggered_buffer_setup(
> > >
> > > > +                       ret = devm_iio_triggered_buffer_setup_ext(
> > > >                                         dev, indio_dev, NULL, trigger_capture,
> > > > -                                       NULL);
> > > > +                                       NULL, fifo_attrs);
> > >
> > > Perhaps it's time to reformat a bit, i.e. move dev to the first line
> > > and do the rest accordingly?
> >
> > this feels like a mix of preferences here;
> > for once, the patch here [as-is], is the minimal form for this change
> > [in terms of patch-noise];
> > so, some people would choose the least noisiest patch;
> >
> > also, this indentation was chosen [as-is here] from the start [for
> > this code block];
> > not sure if it was preferred; i'd suspect it was due to the old 80-col limit;
> >
> > i'd leave it as-is [for now], or defer the decision to a maintainer to
> > decide [either IIO or chromium];
>
> The indenting of this whole code block is a bit too deep.
>
> Looks to me like we should flip the sense of the outer if statement
>
> if (!physical_device)
>         return 0;
>
> That would lead to a whole bunch of reformatting around here including
> picking up this.
>
> For now I can just shuffle it a bit whilst applying.
>
> This set isn't likely to make the merge window anyway now as I'd like
> it to sit on list a little longer just because it touches several
> drivers with active maintainers and I'd like time for them to sanity
> check.
>

ping on this;
should i do a V4 for this?

this is related to the multiple IIO buffer support:
https://lore.kernel.org/linux-iio/20201117162340.43924-1-alexandru.ardelean@analog.com/T/#t

it's one of the patchsets i could split away on it's own;

> Jonathan
>
>
> >
> > >
> > > >                         if (ret)
> > > >                                 return ret;
> > > > -
> > > > -                       if (has_hw_fifo)
> > > > -                               iio_buffer_set_attrs(indio_dev->buffer,
> > > > -                                                    cros_ec_sensor_fifo_attributes);
> > > >                 }
> > > >         }
> > > >
> > > > --
> > > > 2.17.1
> > > >
> > >
> > >
> > > --
> > > With Best Regards,
> > > Andy Shevchenko
>
Jonathan Cameron Nov. 21, 2020, 2:47 p.m. UTC | #5
On Wed, 18 Nov 2020 12:35:16 +0200
Alexandru Ardelean <ardeleanalex@gmail.com> wrote:

> On Tue, Sep 29, 2020 at 6:40 PM Jonathan Cameron <jic23@kernel.org> wrote:
> >
> > On Tue, 29 Sep 2020 17:31:55 +0300
> > Alexandru Ardelean <ardeleanalex@gmail.com> wrote:
> >  
> > > On Tue, Sep 29, 2020 at 4:09 PM Andy Shevchenko
> > > <andy.shevchenko@gmail.com> wrote:  
> > > >
> > > > On Tue, Sep 29, 2020 at 3:55 PM Alexandru Ardelean
> > > > <alexandru.ardelean@analog.com> wrote:
> > > >  
> > > > > This change switches to the new devm_iio_triggered_buffer_setup_ext()
> > > > > function and removes the iio_buffer_set_attrs() call, for assigning the
> > > > > HW FIFO attributes to the buffer.  
> > > >
> > > > Sorry, you were too fast with the version, below one nit.
> > > >  
> > > > > Signed-off-by: Alexandru Ardelean <alexandru.ardelean@analog.com>
> > > > > ---
> > > > >  .../common/cros_ec_sensors/cros_ec_sensors_core.c | 15 +++++++++------
> > > > >  1 file changed, 9 insertions(+), 6 deletions(-)
> > > > >
> > > > > diff --git a/drivers/iio/common/cros_ec_sensors/cros_ec_sensors_core.c b/drivers/iio/common/cros_ec_sensors/cros_ec_sensors_core.c
> > > > > index c62cacc04672..1eafcf04ad69 100644
> > > > > --- a/drivers/iio/common/cros_ec_sensors/cros_ec_sensors_core.c
> > > > > +++ b/drivers/iio/common/cros_ec_sensors/cros_ec_sensors_core.c
> > > > > @@ -353,19 +353,22 @@ int cros_ec_sensors_core_init(struct platform_device *pdev,
> > > > >                         if (ret)
> > > > >                                 return ret;
> > > > >                 } else {
> > > > > +                       const struct attribute **fifo_attrs;
> > > > > +
> > > > > +                       if (has_hw_fifo)
> > > > > +                               fifo_attrs = cros_ec_sensor_fifo_attributes;
> > > > > +                       else
> > > > > +                               fifo_attrs = NULL;
> > > > > +
> > > > >                         /*
> > > > >                          * The only way to get samples in buffer is to set a
> > > > >                          * software trigger (systrig, hrtimer).
> > > > >                          */
> > > > > -                       ret = devm_iio_triggered_buffer_setup(  
> > > >  
> > > > > +                       ret = devm_iio_triggered_buffer_setup_ext(
> > > > >                                         dev, indio_dev, NULL, trigger_capture,
> > > > > -                                       NULL);
> > > > > +                                       NULL, fifo_attrs);  
> > > >
> > > > Perhaps it's time to reformat a bit, i.e. move dev to the first line
> > > > and do the rest accordingly?  
> > >
> > > this feels like a mix of preferences here;
> > > for once, the patch here [as-is], is the minimal form for this change
> > > [in terms of patch-noise];
> > > so, some people would choose the least noisiest patch;
> > >
> > > also, this indentation was chosen [as-is here] from the start [for
> > > this code block];
> > > not sure if it was preferred; i'd suspect it was due to the old 80-col limit;
> > >
> > > i'd leave it as-is [for now], or defer the decision to a maintainer to
> > > decide [either IIO or chromium];  
> >
> > The indenting of this whole code block is a bit too deep.
> >
> > Looks to me like we should flip the sense of the outer if statement
> >
> > if (!physical_device)
> >         return 0;
> >
> > That would lead to a whole bunch of reformatting around here including
> > picking up this.
> >
> > For now I can just shuffle it a bit whilst applying.
> >
> > This set isn't likely to make the merge window anyway now as I'd like
> > it to sit on list a little longer just because it touches several
> > drivers with active maintainers and I'd like time for them to sanity
> > check.
> >  
> 
> ping on this;
> should i do a V4 for this?
Yes, probably worth sending out again. I'd like to see a few more acks
on the individual drivers ideally and a v4 will get this to the
top of peoples' inboxes.

If we don't get them I won't let it block this series, but it's nice
to try at least!

Thanks,

Jonathan

> 
> this is related to the multiple IIO buffer support:
> https://lore.kernel.org/linux-iio/20201117162340.43924-1-alexandru.ardelean@analog.com/T/#t
> 
> it's one of the patchsets i could split away on it's own;
> 
> > Jonathan
> >
> >  
> > >  
> > > >  
> > > > >                         if (ret)
> > > > >                                 return ret;
> > > > > -
> > > > > -                       if (has_hw_fifo)
> > > > > -                               iio_buffer_set_attrs(indio_dev->buffer,
> > > > > -                                                    cros_ec_sensor_fifo_attributes);
> > > > >                 }
> > > > >         }
> > > > >
> > > > > --
> > > > > 2.17.1
> > > > >  
> > > >
> > > >
> > > > --
> > > > With Best Regards,
> > > > Andy Shevchenko  
> >
Jonathan Cameron Nov. 21, 2020, 2:55 p.m. UTC | #6
On Sat, 21 Nov 2020 14:47:39 +0000
Jonathan Cameron <jic23@kernel.org> wrote:

> On Wed, 18 Nov 2020 12:35:16 +0200
> Alexandru Ardelean <ardeleanalex@gmail.com> wrote:
> 
> > On Tue, Sep 29, 2020 at 6:40 PM Jonathan Cameron <jic23@kernel.org> wrote:  
> > >
> > > On Tue, 29 Sep 2020 17:31:55 +0300
> > > Alexandru Ardelean <ardeleanalex@gmail.com> wrote:
> > >    
> > > > On Tue, Sep 29, 2020 at 4:09 PM Andy Shevchenko
> > > > <andy.shevchenko@gmail.com> wrote:    
> > > > >
> > > > > On Tue, Sep 29, 2020 at 3:55 PM Alexandru Ardelean
> > > > > <alexandru.ardelean@analog.com> wrote:
> > > > >    
> > > > > > This change switches to the new devm_iio_triggered_buffer_setup_ext()
> > > > > > function and removes the iio_buffer_set_attrs() call, for assigning the
> > > > > > HW FIFO attributes to the buffer.    
> > > > >
> > > > > Sorry, you were too fast with the version, below one nit.
> > > > >    
> > > > > > Signed-off-by: Alexandru Ardelean <alexandru.ardelean@analog.com>
> > > > > > ---
> > > > > >  .../common/cros_ec_sensors/cros_ec_sensors_core.c | 15 +++++++++------
> > > > > >  1 file changed, 9 insertions(+), 6 deletions(-)
> > > > > >
> > > > > > diff --git a/drivers/iio/common/cros_ec_sensors/cros_ec_sensors_core.c b/drivers/iio/common/cros_ec_sensors/cros_ec_sensors_core.c
> > > > > > index c62cacc04672..1eafcf04ad69 100644
> > > > > > --- a/drivers/iio/common/cros_ec_sensors/cros_ec_sensors_core.c
> > > > > > +++ b/drivers/iio/common/cros_ec_sensors/cros_ec_sensors_core.c
> > > > > > @@ -353,19 +353,22 @@ int cros_ec_sensors_core_init(struct platform_device *pdev,
> > > > > >                         if (ret)
> > > > > >                                 return ret;
> > > > > >                 } else {
> > > > > > +                       const struct attribute **fifo_attrs;
> > > > > > +
> > > > > > +                       if (has_hw_fifo)
> > > > > > +                               fifo_attrs = cros_ec_sensor_fifo_attributes;
> > > > > > +                       else
> > > > > > +                               fifo_attrs = NULL;
> > > > > > +
> > > > > >                         /*
> > > > > >                          * The only way to get samples in buffer is to set a
> > > > > >                          * software trigger (systrig, hrtimer).
> > > > > >                          */
> > > > > > -                       ret = devm_iio_triggered_buffer_setup(    
> > > > >    
> > > > > > +                       ret = devm_iio_triggered_buffer_setup_ext(
> > > > > >                                         dev, indio_dev, NULL, trigger_capture,
> > > > > > -                                       NULL);
> > > > > > +                                       NULL, fifo_attrs);    
> > > > >
> > > > > Perhaps it's time to reformat a bit, i.e. move dev to the first line
> > > > > and do the rest accordingly?    
> > > >
> > > > this feels like a mix of preferences here;
> > > > for once, the patch here [as-is], is the minimal form for this change
> > > > [in terms of patch-noise];
> > > > so, some people would choose the least noisiest patch;
> > > >
> > > > also, this indentation was chosen [as-is here] from the start [for
> > > > this code block];
> > > > not sure if it was preferred; i'd suspect it was due to the old 80-col limit;
> > > >
> > > > i'd leave it as-is [for now], or defer the decision to a maintainer to
> > > > decide [either IIO or chromium];    
> > >
> > > The indenting of this whole code block is a bit too deep.
> > >
> > > Looks to me like we should flip the sense of the outer if statement
> > >
> > > if (!physical_device)
> > >         return 0;
> > >
> > > That would lead to a whole bunch of reformatting around here including
> > > picking up this.
> > >
> > > For now I can just shuffle it a bit whilst applying.
> > >
> > > This set isn't likely to make the merge window anyway now as I'd like
> > > it to sit on list a little longer just because it touches several
> > > drivers with active maintainers and I'd like time for them to sanity
> > > check.
> > >    
> > 
> > ping on this;
> > should i do a V4 for this?  
> Yes, probably worth sending out again. I'd like to see a few more acks
> on the individual drivers ideally and a v4 will get this to the
> top of peoples' inboxes.
> 
> If we don't get them I won't let it block this series, but it's nice
> to try at least!
Actually scratch that this is old now and should be fine to check by
inspection.

I'll have a go at applying it and if nothing odd happens we should
be good.

I was too lazy to fix the alignment Andy pointed out but would like
to see the more major refactoring as discussed for that patch if one
of us gets round to it sometime :)

Hmm. b4 doesn't deal with accidental half typed email addresses so
just went through and fixed all those.

Thanks,

Jonathan


> 
> Thanks,
> 
> Jonathan
> 
> > 
> > this is related to the multiple IIO buffer support:
> > https://lore.kernel.org/linux-iio/20201117162340.43924-1-alexandru.ardelean@analog.com/T/#t
> > 
> > it's one of the patchsets i could split away on it's own;
> >   
> > > Jonathan
> > >
> > >    
> > > >    
> > > > >    
> > > > > >                         if (ret)
> > > > > >                                 return ret;
> > > > > > -
> > > > > > -                       if (has_hw_fifo)
> > > > > > -                               iio_buffer_set_attrs(indio_dev->buffer,
> > > > > > -                                                    cros_ec_sensor_fifo_attributes);
> > > > > >                 }
> > > > > >         }
> > > > > >
> > > > > > --
> > > > > > 2.17.1
> > > > > >    
> > > > >
> > > > >
> > > > > --
> > > > > With Best Regards,
> > > > > Andy Shevchenko    
> > >    
>
diff mbox series

Patch

diff --git a/drivers/iio/common/cros_ec_sensors/cros_ec_sensors_core.c b/drivers/iio/common/cros_ec_sensors/cros_ec_sensors_core.c
index c62cacc04672..1eafcf04ad69 100644
--- a/drivers/iio/common/cros_ec_sensors/cros_ec_sensors_core.c
+++ b/drivers/iio/common/cros_ec_sensors/cros_ec_sensors_core.c
@@ -353,19 +353,22 @@  int cros_ec_sensors_core_init(struct platform_device *pdev,
 			if (ret)
 				return ret;
 		} else {
+			const struct attribute **fifo_attrs;
+
+			if (has_hw_fifo)
+				fifo_attrs = cros_ec_sensor_fifo_attributes;
+			else
+				fifo_attrs = NULL;
+
 			/*
 			 * The only way to get samples in buffer is to set a
 			 * software trigger (systrig, hrtimer).
 			 */
-			ret = devm_iio_triggered_buffer_setup(
+			ret = devm_iio_triggered_buffer_setup_ext(
 					dev, indio_dev, NULL, trigger_capture,
-					NULL);
+					NULL, fifo_attrs);
 			if (ret)
 				return ret;
-
-			if (has_hw_fifo)
-				iio_buffer_set_attrs(indio_dev->buffer,
-						     cros_ec_sensor_fifo_attributes);
 		}
 	}