Re:Re: [PATCH] clk/meson: Hold reference returned by of_get_parent()

["Liang He" <windhl@126.com>]

At 2022-06-27 06:02:45, "Martin Blumenstingl" <martin.blumenstingl@googlemail.com> wrote:
>Hello,
>
>thank you for submitting this patch!
>
>On Fri, Jun 24, 2022 at 12:28 PM Liang He <windhl@126.com> wrote:
>[...]
>>  these bugs are compiled tested in 5.19rc2 with at91_dt_defconfig
>Please note that you're changing some drivers which are only in the
>arch/arm/configs/multi_v7_defconfig while others are only in
>arch/arm64/configs/defconfig
>I think at91_dt_defconfig will not compile any of the drivers below
>

Hi, Martin.

To simplify the building test, I use the at91_dt_defconfig to build .config
for these similar bugs in clk-at91, clk-sprd，etc.
Then I use following command to complete different compile tests for clk-meson:

make CONFIG_ARCH_MESON=y CONFIG_COMMON_CLK_MESON8B=y CROSS_COMPILE=arm-linux-gnueabi- ARCH=arm ./drivers/clk/meson/meson8b.o
make CONFIG_ARCH_MESON=y CONFIG_COMMON_CLK_MESON_EE_CLKC=y CROSS_COMPILE=arm-linux-gnueabi- ARCH=arm ./drivers/clk/meson/meson-eeclk.o
make CONFIG_ARCH_MESON=y CONFIG_COMMON_CLK_MESON_AO_CLKC=y CROSS_COMPILE=arm-linux-gnueabi- ARCH=arm ./drivers/clk/meson/meson-aoclk.o

I will add these detailed compile commands in my version-2 patch and also in future patch commit.

>[...]
>> diff --git a/drivers/clk/meson/meson8b.c b/drivers/clk/meson/meson8b.c
>> index 8f3b7a94a667..54188319f349 100644
>> --- a/drivers/clk/meson/meson8b.c
>> +++ b/drivers/clk/meson/meson8b.c
>> @@ -3796,8 +3796,11 @@ static void __init meson8b_clkc_init_common(struct device_node *np,
>>         struct clk *notifier_clk;
>>         struct regmap *map;
>>         int i, ret;
>> +       struct device_node *tp;
>This function uses reverse christmas tree sorting, so the longest line
>goes on top while the shortest should stay on the bottom.
>Can you please explain what "tp" stands for? Personally I would call
>this variable parent_np, but maybe "tp" has a similar meaning.
>
>
>Best regards,
>Martin

Thanks very much, Martin.

I will send version-2 patch to keep the  reverse christmas tree.
And I will follow this rule in my future patch code.

I just want to ues 'tp' as 'tmp_pointer', but your advice is
more reasonable and I will change that in version-2 patch.

Liang