On Fri, 30 May 2025 20:45:32 +0300
Andy Shevchenko <andy@xxxxxxxxxx> wrote:
> On Thu, May 29, 2025 at 07:50:29PM -0300, Jonathan Santos wrote:
> > The synchronization method using GPIO requires the generated pulse to be
> > truly synchronous with the base MCLK signal. When it is not possible to
> > do that in hardware, the datasheet recommends using synchronization over
> > SPI, where the generated pulse is already synchronous with MCLK. This
> > requires the SYNC_OUT pin to be connected to the SYNC_IN pin.
> >
> > Use trigger-sources property to enable device synchronization over SPI
> > and multi-device synchronization while replacing sync-in-gpios property.
>
> ...
>
> > struct ad7768_state {
>
> > struct iio_trigger *trig;
> > struct gpio_desc *gpio_sync_in;
> > struct gpio_desc *gpio_reset;
>
> > + bool en_spi_sync;
>
> I'm wondering if moving this...
>
> > const char *labels[ARRAY_SIZE(ad7768_channels)];
> > struct gpio_chip gpiochip;
>
> ...to here saves a few bytes in accordance to `pahole`.
>
> > };
>
> ...
>
> > +static int ad7768_trigger_sources_sync_setup(struct device *dev,
> > + struct fwnode_handle *dev_fwnode,
> > + struct ad7768_state *st)
> > +{
> > + struct fwnode_reference_args args;
> > +
> > + struct fwnode_handle *fwnode __free(fwnode_handle) =
> > + fwnode_find_reference_args(dev_fwnode, "trigger-sources",
> > + "#trigger-source-cells", 0,
> > + AD7768_TRIGGER_SOURCE_SYNC_IDX, &args);
>
> I don't see how args are being used. This puts in doubt the need of the first
> patch.
That did get discussed (more context needed in the commit message for patch 1).
I wasn't happy with ignoring #trigger-source-cells which is required in the
binding but here is known to be 0.
If it was larger than 0 but we didn't care about the arguments I believe
we'd still need to use this call to take the right stride through the
data array that this is coming from.
Ultimately I think that is this bit of code establishing the end of the phandle.
https://elixir.bootlin.com/linux/v6.15/source/drivers/of/base.c#L1300
I might have gotten it wrong how this all works though!
J
