Re: [PATCH v6 2/2] iio: adc: ti-ads112c04: Add support for TI ADS112C04

[sashiko-bot@kernel.org]

Thank you for your contribution! Sashiko AI review found 1 potential issue(s) to consider:
- [High] Interrupt storm (hard lockup) if DRDY is configured as level-triggered.
--

--- Patch [2]: [PATCH v6 2/2] iio: adc: ti-ads112c04: Add support for TI ADS112C04 ---
commit 88f1a4a634251e87a5f5b33f385ea979bcaa7f17
Author: Kyle Hsieh <kylehsieh1995@gmail.com>

iio: adc: ti-ads112c04: Add support for TI ADS112C04

Add IIO driver support for the Texas Instruments ADS112C04 (16-bit)
delta-sigma ADCs. The driver implements single-shot conversions, dynamic
parsing of channels, hardware interrupts via the DRDY pin, and per-channel
reference source selection.

> diff --git a/drivers/iio/adc/ti-ads112c04.c b/drivers/iio/adc/ti-ads112c04.c
> new file mode 100644
> --- /dev/null
> +++ b/drivers/iio/adc/ti-ads112c04.c
[ ... ]
> +static irqreturn_t ads112c04_irq_handler(int irq, void *private)
> +{
> +	struct iio_dev *indio_dev = private;
> +	struct ads112c04_state *st = iio_priv(indio_dev);
> +
> +	complete(&st->completion);
> +
> +	return IRQ_HANDLED;
> +}
[ ... ]
> +static int ads112c04_probe(struct i2c_client *client)
> +{
[ ... ]
> +	if (client->irq > 0) {
> +		ret = devm_request_irq(dev, client->irq, ads112c04_irq_handler, 0,
> +				       indio_dev->name, indio_dev);

[Severity: High]
Since the hardware DRDY pin stays low until data is read, could an interrupt
storm occur here if the device tree configures the interrupt as
level-triggered?

By passing 0 for the irqflags, the driver accepts the device tree's default
trigger type. If it is set to a level trigger, ads112c04_irq_handler() will
fire, call complete(), and return IRQ_HANDLED without doing any I2C reads to
clear the device-side interrupt.

Because the interrupt source isn't cleared, the interrupt controller will
immediately re-trigger the hardirq. Won't this create an infinite loop that
starves the process context, preventing ads112c04_read_data() from ever
executing to clear the condition?

Should this request enforce an edge trigger, or use a threaded interrupt
handler with IRQF_ONESHOT so the interrupt is masked until the data is read?

> +		if (ret)
> +			return ret;
> +	}
> +
> +	return devm_iio_device_register(dev, indio_dev);
> +}

-- 
Sashiko AI review · https://sashiko.dev/#/patchset/20260820-ti-ads112c04-driver-v6-0-0e8f0aacce9b@gmail.com?part=2