Re: [PATCH 1/2] clk: meson-g12a: add cpu clock bindings

From: Neil Armstrong
Date: Fri Mar 01 2019 - 11:43:24 EST


Hi Martin,

On 01/03/2019 16:26, Martin Blumenstingl wrote:
> Hi Neil,
>
> On Fri, Mar 1, 2019 at 11:22 AM Neil Armstrong <narmstrong@xxxxxxxxxxxx> wrote:
>>
>> Add Amlogic G12A Family CPU clocks bindings, only export CPU_CLK since
>> it should be the only ID used.
> is this also true for the CPU post-dividers (APB, ATB, AXI, CPU CLK TRACE)?

Do we need these to be exported ?

>
>> Signed-off-by: Neil Armstrong <narmstrong@xxxxxxxxxxxx>
>> ---
>> drivers/clk/meson/g12a.h | 24 +++++++++++++++++++++++-
>> include/dt-bindings/clock/g12a-clkc.h | 1 +
>> 2 files changed, 24 insertions(+), 1 deletion(-)
>>
>> diff --git a/drivers/clk/meson/g12a.h b/drivers/clk/meson/g12a.h
>> index f399dfe1401c..4854750df902 100644
>> --- a/drivers/clk/meson/g12a.h
>> +++ b/drivers/clk/meson/g12a.h
>> @@ -166,8 +166,30 @@
>> #define CLKID_MALI_0_DIV 170
>> #define CLKID_MALI_1_DIV 173
>> #define CLKID_MPLL_5OM_DIV 176
>> +#define CLKID_PCIE_PLL_DCO 178
>> +#define CLKID_PCIE_PLL_DCO_DIV2 179
>> +#define CLKID_PCIE_PLL_OD 180
> how are these PCIe clock related to the CPU clocks?

Oops, bad merge...

>
>> +#define CLKID_SYS_PLL_DIV16_EN 181
>> +#define CLKID_SYS_PLL_DIV16 182
>> +#define CLKID_CPU_CLK_DYN0_SEL 183
>> +#define CLKID_CPU_CLK_DYN0_DIV 184
>> +#define CLKID_CPU_CLK_DYN0 185
>> +#define CLKID_CPU_CLK_DYN1_SEL 186
>> +#define CLKID_CPU_CLK_DYN1_DIV 187
>> +#define CLKID_CPU_CLK_DYN1 188
>> +#define CLKID_CPU_CLK_DYN 189
>> +#define CLKID_CPU_CLK_DIV16_EN 191
>> +#define CLKID_CPU_CLK_DIV16 192
>> +#define CLKID_CPU_CLK_APB_DIV 193
>> +#define CLKID_CPU_CLK_APB 194
>> +#define CLKID_CPU_CLK_ATB_DIV 195
>> +#define CLKID_CPU_CLK_ATB 196
>> +#define CLKID_CPU_CLK_AXI_DIV 197
>> +#define CLKID_CPU_CLK_AXI 198
>> +#define CLKID_CPU_CLK_TRACE_DIV 299
>> +#define CLKID_CPU_CLK_TRACE 200
>>
>> -#define NR_CLKS 178
>> +#define NR_CLKS 201
> shouldn't all changes to this file (drivers/clk/meson/g12a.h) be part
> of the patch which adds the actual clocks so the dt-bindings patch is
> independent of the clock driver patches?
> in this case the subject should also be updated to "dt-bindings:
> clock: g12a: ..."

I don't have any opinion, Jerome ?

>
>
> Regards
> Martin
>

Thanks !

Neil