Re: [PATCH 4/5] Input: s6sy761 - move chip differences into per-chip data

[sashiko-bot@kernel.org]

Thank you for your contribution! Sashiko AI review found no new issues
in this patch, but noticed 1 related pre-existing issue it does not
introduce:
- [High] Input: s6sy761 - interrupt storm on I2C error in
  s6sy761_irq_handler() [1]
--

commit a2e70885f2393c600cd0947a2d064a0c055020a7
Author: Nizat Mikayilli <nacht@dxdmail.cc>

Input: s6sy761 - move chip differences into per-chip data

This commit moves controller-specific parameters and callbacks into a
per-chip data structure to prepare for adding support for the ST FTS1BA90A
controller.

> diff --git a/drivers/input/touchscreen/s6sy761.c b/drivers/input/touchscreen/s6sy761.c
> index 59166865d397c..f000ed0d7a9a4 100644
> --- a/drivers/input/touchscreen/s6sy761.c
> +++ b/drivers/input/touchscreen/s6sy761.c

> @@ -229,7 +244,7 @@ static irqreturn_t s6sy761_irq_handler(int irq, void *dev)
>  
>  	ret = i2c_smbus_read_i2c_block_data(sdata->client,
>  					    S6SY761_READ_ONE_EVENT,
> -					    S6SY761_EVENT_SIZE,
> +					    sdata->chip->event_size,
>  					    sdata->data);
>  	if (ret < 0) {
>  		dev_err(&sdata->client->dev, "failed to read events\n");

[Severity: High]
This problem wasn't introduced by this patch, but can returning IRQ_HANDLED
when i2c_smbus_read_i2c_block_data() or s6sy761_read_events() fails cause
an interrupt storm [1]?

The IRQ is registered with IRQF_TRIGGER_LOW.  When an I2C read fails, the
pending event is not drained from the controller, so the level-triggered
interrupt line stays asserted and immediately re-fires once unmasked, while
returning IRQ_HANDLED instead of IRQ_NONE prevents the spurious interrupt
detector from disabling the IRQ.

[1] https://sashiko.dev/#/bug/linux-d31827dd-6777-4c33-8908-c3b4fbd062c3

-- 
Sashiko AI review · https://sashiko.dev/#/patchset/20261108-b4-s6sy761-fts1ba90a-v1-0-999b3cfaf82d@dxdmail.cc?part=4