Re: [PATCH v2 17/35] clk: mediatek: mt2712-apmixedsys: Switch to common probe/remove helpers

[sashiko-bot@kernel.org]

Thank you for your contribution! Sashiko AI review found 1 potential issue(s) to consider:
- [Medium] clk: mediatek: missing shared_io flag causes redundant exclusive syscon mapping
--

commit c93284125c00a17383c7e9f49de6dd50d3cdc4e9
Author: Akari Tsuyukusa <akkun11.open@gmail.com>

clk: mediatek: mt2712-apmixedsys: Switch to common probe/remove helpers

This patch updates the MT2712 apmixedsys driver to use the common
mtk_clk_simple_probe() and mtk_clk_simple_remove() helpers. It removes the
custom probe and remove callbacks that previously handled PLL registration
and unregistration.

> diff --git a/drivers/clk/mediatek/clk-mt2712-apmixedsys.c b/drivers/clk/mediatek/clk-mt2712-apmixedsys.c
> --- a/drivers/clk/mediatek/clk-mt2712-apmixedsys.c
> +++ b/drivers/clk/mediatek/clk-mt2712-apmixedsys.c

[ ... ]

> +static const struct mtk_clk_desc apmixed_desc = {
> +	.plls = plls,
> +	.num_plls = ARRAY_SIZE(plls),
> +};

[Severity: Medium]
Does this descriptor need the .shared_io = true flag?

By omitting .shared_io = true here, it looks like mtk_clk_simple_probe()
will automatically call devm_platform_ioremap_resource(). This requests
the memory region exclusively, which seems to conflict with the shared
nature of a syscon hardware region.

Additionally, the resulting base pointer from mtk_clk_simple_probe()
doesn't appear to be passed to mtk_clk_register_plls(), which internally
maps the exact same region again using of_iomap().

Could this lead to an unused, redundant ioremap mapping and an improper
request_mem_region() on the syscon?

[ ... ]

-- 
Sashiko AI review · https://sashiko.dev/#/patchset/20261013151734.3177690-1-akkun11.open@gmail.com?part=17