Re: [PATCH 3/4] drivers: media: i2c: imx471: Add Device Tree matching support

[sashiko-bot@kernel.org]

Thank you for your contribution! Sashiko AI review found 1 potential issue(s) to consider:

Pre-existing issues:
- [High] The `dev_pm_ops` structure pointer `&imx471_pm_ops` is incorrectly wrapped with `pm_sleep_ptr()`, discarding runtime PM callbacks when `CONFIG_PM_SLEEP` is disabled.
--

--- Patch [3]: [PATCH 3/4] drivers: media: i2c: imx471: Add Device Tree matching support ---
commit 0a7fa78c5d7c772d5834bc87176b5fcc9a6246db
Author: Danila Tikhonov <danila@mainlining.org>

drivers: media: i2c: imx471: Add Device Tree matching support

The IMX471 driver currently only supports probing via ACPI. This commit adds
an OF device ID table with the sony,imx471 compatible string to allow
probing on Device Tree based platforms.

> diff --git a/drivers/media/i2c/imx471.c b/drivers/media/i2c/imx471.c
> --- a/drivers/media/i2c/imx471.c
> +++ b/drivers/media/i2c/imx471.c
[ ... ]
>  static struct i2c_driver imx471_i2c_driver = {
>  	.driver = {
>  		.name = "imx471",
>  		.acpi_match_table = ACPI_PTR(imx471_acpi_ids),
> +		.of_match_table = imx471_of_match,
>  		.pm = pm_sleep_ptr(&imx471_pm_ops),

[Severity: High]
This is a pre-existing issue, but is it intended to use pm_sleep_ptr() here 
instead of pm_ptr()? 

When CONFIG_PM is enabled but CONFIG_PM_SLEEP is disabled, pm_sleep_ptr() 
evaluates to NULL. This drops the entire dev_pm_ops structure and its runtime 
PM callbacks in imx471_i2c_driver. 

Will this cause the device to permanently leak power on embedded platforms 
because the regulators and clocks are never disabled after probe or between 
streams?

>  	},
>  	.probe = imx471_probe,

-- 
Sashiko AI review · https://sashiko.dev/#/patchset/20260805-imx471-v1-0-7be79f539ed8@mainlining.org?part=3