diff mbox series

[v2,1/3] iio: adc: ad7192: Add IRQ flag

Message ID 20210906065630.16325-2-alexandru.tachici@analog.com (mailing list archive)
State Accepted
Headers show
Series iio: adc: Fix flags in sigma-delta drivers | expand

Commit Message

Alexandru Tachici Sept. 6, 2021, 6:56 a.m. UTC
From: Alexandru Tachici <alexandru.tachici@analog.com>

IRQ type in ad_sigma_delta_info struct was missing.

In Sigma-Delta devices the SDO line is also used as an interrupt.
Leaving IRQ on level instead of falling might trigger a sample read
when the IRQ is enabled, as the SDO line is already low. Not sure
if SDO line will always imediately go high in ad_sd_buffer_postenable
before the IRQ is enabled.

Also the datasheet seem to explicitly say the falling edge of the SDO
should be used as an interrupt:
From the AD7192 datasheet: "The DOUT/RDY falling edge can be used
as an interrupt to a processor,"

Fixes: da4d3d6bb9f6 ("iio: adc: ad-sigma-delta: Allow custom IRQ flags")
Signed-off-by: Alexandru Tachici <alexandru.tachici@analog.com>
---
 drivers/iio/adc/ad7192.c | 1 +
 1 file changed, 1 insertion(+)

Comments

Jonathan Cameron Sept. 11, 2021, 5:39 p.m. UTC | #1
On Mon, 6 Sep 2021 09:56:28 +0300
<alexandru.tachici@analog.com> wrote:

> From: Alexandru Tachici <alexandru.tachici@analog.com>
> 
> IRQ type in ad_sigma_delta_info struct was missing.
> 
> In Sigma-Delta devices the SDO line is also used as an interrupt.
> Leaving IRQ on level instead of falling might trigger a sample read
> when the IRQ is enabled, as the SDO line is already low. Not sure
> if SDO line will always imediately go high in ad_sd_buffer_postenable
> before the IRQ is enabled.
> 
> Also the datasheet seem to explicitly say the falling edge of the SDO
> should be used as an interrupt:
> From the AD7192 datasheet: "The DOUT/RDY falling edge can be used
> as an interrupt to a processor,"
> 
> Fixes: da4d3d6bb9f6 ("iio: adc: ad-sigma-delta: Allow custom IRQ flags")
> Signed-off-by: Alexandru Tachici <alexandru.tachici@analog.com>

As in thread discussing previous version.  This one should probably be left
to firmware to configure.  It's not 'wrong' before this point as firmware
should be specifying it correctly.  That allows for the interrupt line
to the processor to be inverted and other silliness that happens on real
boards.

Jonathan

> ---
>  drivers/iio/adc/ad7192.c | 1 +
>  1 file changed, 1 insertion(+)
> 
> diff --git a/drivers/iio/adc/ad7192.c b/drivers/iio/adc/ad7192.c
> index ee8ed9481025..2121a812b0c3 100644
> --- a/drivers/iio/adc/ad7192.c
> +++ b/drivers/iio/adc/ad7192.c
> @@ -293,6 +293,7 @@ static const struct ad_sigma_delta_info ad7192_sigma_delta_info = {
>  	.has_registers = true,
>  	.addr_shift = 3,
>  	.read_mask = BIT(6),
> +	.irq_flags = IRQF_TRIGGER_FALLING,
>  };
>  
>  static const struct ad_sd_calib_data ad7192_calib_arr[8] = {
Jonathan Cameron Sept. 11, 2021, 5:44 p.m. UTC | #2
On Sat, 11 Sep 2021 18:39:21 +0100
Jonathan Cameron <jic23@kernel.org> wrote:

> On Mon, 6 Sep 2021 09:56:28 +0300
> <alexandru.tachici@analog.com> wrote:
> 
> > From: Alexandru Tachici <alexandru.tachici@analog.com>
> > 
> > IRQ type in ad_sigma_delta_info struct was missing.
> > 
> > In Sigma-Delta devices the SDO line is also used as an interrupt.
> > Leaving IRQ on level instead of falling might trigger a sample read
> > when the IRQ is enabled, as the SDO line is already low. Not sure
> > if SDO line will always imediately go high in ad_sd_buffer_postenable
> > before the IRQ is enabled.
> > 
> > Also the datasheet seem to explicitly say the falling edge of the SDO
> > should be used as an interrupt:
> > From the AD7192 datasheet: "The DOUT/RDY falling edge can be used
> > as an interrupt to a processor,"
> > 
> > Fixes: da4d3d6bb9f6 ("iio: adc: ad-sigma-delta: Allow custom IRQ flags")
> > Signed-off-by: Alexandru Tachici <alexandru.tachici@analog.com>
> 
> As in thread discussing previous version.  This one should probably be left
> to firmware to configure.  It's not 'wrong' before this point as firmware
> should be specifying it correctly.  That allows for the interrupt line
> to the processor to be inverted and other silliness that happens on real
> boards.

Ignore that - I'd forgotten this is a joint DOUT / Data ready pin so it's
polarity needs to track that of the SPI pin and inverting that isn't
something we cope with.

So applied to the fixes-togreg branch of iio.git and marked for stable.

Thanks,

Jonathan

> 
> Jonathan
> 
> > ---
> >  drivers/iio/adc/ad7192.c | 1 +
> >  1 file changed, 1 insertion(+)
> > 
> > diff --git a/drivers/iio/adc/ad7192.c b/drivers/iio/adc/ad7192.c
> > index ee8ed9481025..2121a812b0c3 100644
> > --- a/drivers/iio/adc/ad7192.c
> > +++ b/drivers/iio/adc/ad7192.c
> > @@ -293,6 +293,7 @@ static const struct ad_sigma_delta_info ad7192_sigma_delta_info = {
> >  	.has_registers = true,
> >  	.addr_shift = 3,
> >  	.read_mask = BIT(6),
> > +	.irq_flags = IRQF_TRIGGER_FALLING,
> >  };
> >  
> >  static const struct ad_sd_calib_data ad7192_calib_arr[8] = {
>
diff mbox series

Patch

diff --git a/drivers/iio/adc/ad7192.c b/drivers/iio/adc/ad7192.c
index ee8ed9481025..2121a812b0c3 100644
--- a/drivers/iio/adc/ad7192.c
+++ b/drivers/iio/adc/ad7192.c
@@ -293,6 +293,7 @@  static const struct ad_sigma_delta_info ad7192_sigma_delta_info = {
 	.has_registers = true,
 	.addr_shift = 3,
 	.read_mask = BIT(6),
+	.irq_flags = IRQF_TRIGGER_FALLING,
 };
 
 static const struct ad_sd_calib_data ad7192_calib_arr[8] = {