Re: [PATCH v11 03/11] reset: realtek: Add RTD1625 reset controller driver

[sashiko-bot@kernel.org]

Thank you for your contribution! Sashiko AI review found 2 potential issue(s) to consider:
- [Low] The auxiliary driver's name is incorrectly initialized via `.driver = { .name = ... }`, causing the intended name to be silently overwritten.
- [Low] Missing `<linux/module.h>` include for module macros.
--

--- Patch [3]: [PATCH v11 03/11] reset: realtek: Add RTD1625 reset controller driver ---
commit 6ed7fb0b40e040995c4e1e41d96adfc82a758f96
Author: Cheng-Yu Lee <cylee12@realtek.com>

reset: realtek: Add RTD1625 reset controller driver

This adds support for the CRT and ISO domain reset controller on the Realtek
RTD1625 SoC. The reset driver is implemented as an auxiliary driver to handle
shared register space with the CRT and ISO clock controller.

> diff --git a/drivers/reset/realtek/reset-rtd1625.c b/drivers/reset/realtek/reset-rtd1625.c
> new file mode 100644
> index 0000000000000..fa87462f50e9d
> --- /dev/null
> +++ b/drivers/reset/realtek/reset-rtd1625.c
> @@ -0,0 +1,262 @@
> +// SPDX-License-Identifier: GPL-2.0-only
> +/*
> + * Copyright (C) 2026 Realtek Semiconductor Corporation
> + */
> +
> +#include <dt-bindings/reset/realtek,rtd1625.h>
> +#include <linux/auxiliary_bus.h>
> +#include <linux/array_size.h>
> +#include <linux/device.h>
> +#include <linux/errno.h>
> +#include <linux/slab.h>
> +#include "reset-rtk-common.h"

[Severity: Low]
This isn't a bug, but should <linux/module.h> be explicitly included here?
The code relies heavily on module macros like MODULE_DEVICE_TABLE,
module_auxiliary_driver, and MODULE_DESCRIPTION, which currently only build
because of an indirect include through <linux/device.h>.

[ ... ]

> +static struct auxiliary_driver rtd1625_reset_driver = {
> +	.probe    = rtd1625_reset_probe,
> +	.id_table = rtd1625_reset_ids,
> +	.driver = {
> +		.name = "rtd1625-reset",
> +	},
> +};

[Severity: Low]
Is this driver name initialization correct for the auxiliary bus? Setting the
name inside the embedded device_driver struct (.driver = { .name = ... })
causes the intended name to be ignored, as the auxiliary bus framework will
dynamically allocate and overwrite it with the module name. 

Should this be initialized using the .name field of the parent
auxiliary_driver struct instead?

> +module_auxiliary_driver(rtd1625_reset_driver);
> +
> +MODULE_DESCRIPTION("Realtek RTD1625 Reset Controller Driver");
> +MODULE_LICENSE("GPL");
> +MODULE_IMPORT_NS("REALTEK_RESET");

-- 
Sashiko AI review · https://sashiko.dev/#/patchset/20260728142806.1954638-1-eleanor.lin@realtek.com?part=3