On 26/04/2024 07:24, Marek Szyprowski wrote:
> On 26.04.2024 02:24, William McVicker wrote:
>> On 04/25/2024, Marek Szyprowski wrote:
>>> Since commit 9484f2cb8332 ("clk: samsung: exynos-clkout: convert to
>>> module driver") this driver is instantiated as MFD-cell (matched by
>>> platform device name) not as a real platform device created by OF code.
>>> Remove then of_match_table and related MODULE_DEVICE_TABLE to avoid
>>> further confusion.
>>>
>>> Signed-off-by: Marek Szyprowski <m.szyprowski@xxxxxxxxxxx>
>>> ---
>>> drivers/clk/samsung/clk-exynos-clkout.c | 2 --
>>> 1 file changed, 2 deletions(-)
>>>
>>> diff --git a/drivers/clk/samsung/clk-exynos-clkout.c b/drivers/clk/samsung/clk-exynos-clkout.c
>>> index 503c6f5b20d5..0c7f4e2aa366 100644
>>> --- a/drivers/clk/samsung/clk-exynos-clkout.c
>>> +++ b/drivers/clk/samsung/clk-exynos-clkout.c
>>> @@ -75,7 +75,6 @@ static const struct of_device_id exynos_clkout_ids[] = {
>>> .data = &exynos_clkout_exynos5,
>>> }, { }
>>> };
>>> -MODULE_DEVICE_TABLE(of, exynos_clkout_ids);
>> I understand these are duplicates of the exynos-pmu driver, but was wondering
>> if this will impact the exynos-clkout module from getting auto-loaded? Without
>> the MODULE_DEVICE_TABLE() defined, the aliases won't be created that trigger
>> udev to load this module and the mfd driver is obviously not going to load it.
>
> This driver loaded and matched only against the platform device name
Matched yes, but "loaded"? As in module loaded? Are you sure? There is
no MODULE_ALIAS, no platform_id_table with MODULE_DEVICE_TABLE, so with
this patch all aliases are gone.
The old aliases indeed could have been used to load the clkout, even if
not used for matching.
> ("exynos-clkout") since the mentioned commit 9484f2cb8332. These OF IDs
> defined as MODULE_DEVICE_TABLE and of_match_table are leftovers from the
> old (pre-9484f2cb8332) instantiating method.and should be removed by
> that commit too.
Best regards,
Krzysztof
