Re: [PATCH v3 5/5] iio: adc: ltc2378: Enable triggered buffer data capture

[David Lechner <dlechner@baylibre.com>]

On 6/16/26 9:04 PM, Marcelo Schmitt wrote:
> Enable users to run triggered data captures with LTC2378 and similar ADCs.
> 
> Signed-off-by: Marcelo Schmitt <marcelo.schmitt@analog.com>
> ---
> Change log v2 -> v3:
> - Extracted from main driver file into a separate buffer-specific containment.
> 
>  drivers/iio/adc/ltc2378-lib-core.c         | 17 +++++++-
>  drivers/iio/adc/ltc2378-triggered-buffer.c | 49 ++++++++++++++++++++++
>  drivers/iio/adc/ltc2378.h                  | 15 +++++++
>  3 files changed, 80 insertions(+), 1 deletion(-)
>  create mode 100644 drivers/iio/adc/ltc2378-triggered-buffer.c
> 
> diff --git a/drivers/iio/adc/ltc2378-lib-core.c b/drivers/iio/adc/ltc2378-lib-core.c
> index 1160f4324d01..ec83e9f2ae81 100644
> --- a/drivers/iio/adc/ltc2378-lib-core.c
> +++ b/drivers/iio/adc/ltc2378-lib-core.c
> @@ -18,7 +18,7 @@ int ltc2378_lib_buffer_setup(struct iio_dev *indio_dev, struct ltc2378_state *st
>  
>  	ret = __ltc2378_set_offload_ops(st);
>  	if (ret == -EOPNOTSUPP)
> -		return 0; /* Let device setup complete without buffer support */
> +		goto trigger_buf_setup;
>  
>  	if (!ret)
>  		ret = st->ops->buffer_setup(indio_dev, st);
> @@ -27,6 +27,21 @@ int ltc2378_lib_buffer_setup(struct iio_dev *indio_dev, struct ltc2378_state *st
>  		return dev_err_probe(dev, ret, "error on SPI offload setup\n");
>  
>  	return 0;
> +
> +trigger_buf_setup:
> +	ret = __ltc2378_set_triggered_buf_ops(st);
> +	if (ret == -EOPNOTSUPP)
> +		return 0; /* Let device setup complete without buffer support */
> +
> +	if (!ret)
> +		ret = st->ops->buffer_setup(indio_dev, st);
> +
> +	if (ret)
> +		return dev_err_probe(dev, ret, "error on buffer setup\n");

This is repeating the code above. Seems like it would be better without
the goto.

> +
> +	st->chans[st->num_iio_chans++] = IIO_CHAN_SOFT_TIMESTAMP(1);

Adding a channel here seems messy. I still think static const channel
data would be better. But at least would be better if this was moved
to the same function as the other channel setup.

> +
> +	return 0;
>  }
>  EXPORT_SYMBOL_NS_GPL(ltc2378_lib_buffer_setup, "IIO_LTC2378");
>  
> diff --git a/drivers/iio/adc/ltc2378-triggered-buffer.c b/drivers/iio/adc/ltc2378-triggered-buffer.c
> new file mode 100644
> index 000000000000..d1d788fb5cb4
> --- /dev/null
> +++ b/drivers/iio/adc/ltc2378-triggered-buffer.c
> @@ -0,0 +1,49 @@
> +// SPDX-License-Identifier: GPL-2.0+
> +/*
> + * Copyright (C) 2026 Analog Devices, Inc.
> + * Author: Marcelo Schmitt <marcelo.schmitt@analog.com>
> + */
> +
> +#include <linux/iio/buffer.h>
> +#include <linux/iio/triggered_buffer.h>
> +#include <linux/iio/trigger_consumer.h>
> +
> +#include <ltc2378.h>
> +
> +static irqreturn_t ltc2378_trigger_handler(int irq, void *p)
> +{
> +	struct iio_poll_func *pf = p;
> +	struct iio_dev *indio_dev = pf->indio_dev;
> +	struct ltc2378_state *st = iio_priv(indio_dev);
> +	int ret;
> +
> +	ret = ltc2378_convert_and_acquire(st);
> +	if (ret < 0)
> +		goto err_out;

As mentioned elsewhere, SPI xfer size may be 2 or 4 bytes, but
we are always pushing 4 bytes, so this only works on little-endian
architecture.

I think best would be to have storagesize = 16 when appropriate to
avoid having to manually handle this.

> +
> +	iio_push_to_buffers_with_ts(indio_dev, &st->scan, sizeof(st->scan),
> +				    pf->timestamp);
> +
> +err_out:
> +	iio_trigger_notify_done(indio_dev->trig);
> +	return IRQ_HANDLED;
> +}
> +