Re: [PATCH 2/5] clk: qcom: gcc-ipq5018: fix GE PHY reset

[Konrad Dybcio <konrad.dybcio@oss.qualcomm.com>]

On 5/27/25 1:14 PM, George Moussalem wrote:
> Hi Konrad,
> 
> On 5/27/25 15:00, Konrad Dybcio wrote:
>> On 5/25/25 7:56 PM, George Moussalem via B4 Relay wrote:
>>> From: George Moussalem <george.moussalem@outlook.com>
>>>
>>> The MISC reset is supposed to trigger a resets across the MDC, DSP, and
>>> RX & TX clocks of the IPQ5018 internal GE PHY. So let's set the bitmask
>>> of the reset definition accordingly in the GCC as per the downstream
>>> driver.
>>>
>>> Link: https://git.codelinaro.org/clo/qsdk/oss/kernel/linux-ipq-5.4/-/commit/00743c3e82fa87cba4460e7a2ba32f473a9ce932
>>>
>>> Signed-off-by: George Moussalem <george.moussalem@outlook.com>
>>> ---
>>>   drivers/clk/qcom/gcc-ipq5018.c | 2 +-
>>>   1 file changed, 1 insertion(+), 1 deletion(-)
>>>
>>> diff --git a/drivers/clk/qcom/gcc-ipq5018.c b/drivers/clk/qcom/gcc-ipq5018.c
>>> index 70f5dcb96700f55da1fb19fc893d22350a7e63bf..02d6f08f389f24eccc961b9a4271288c6b635bbc 100644
>>> --- a/drivers/clk/qcom/gcc-ipq5018.c
>>> +++ b/drivers/clk/qcom/gcc-ipq5018.c
>>> @@ -3660,7 +3660,7 @@ static const struct qcom_reset_map gcc_ipq5018_resets[] = {
>>>       [GCC_WCSS_AXI_S_ARES] = { 0x59008, 6 },
>>>       [GCC_WCSS_Q6_BCR] = { 0x18004, 0 },
>>>       [GCC_WCSSAON_RESET] = { 0x59010, 0},
>>> -    [GCC_GEPHY_MISC_ARES] = { 0x56004, 0 },
>>> +    [GCC_GEPHY_MISC_ARES] = { 0x56004, .bitmask = 0xf },
>>
>> The computer tells me there aren't any bits beyond this mask..
>>
>> Does this actually fix anything?
> 
> The mask is documented in the referenced downstream driver and allows for consolidating:
> 
> resets = <&gcc GCC_GEPHY_MDC_SW_ARES>,
>      <&gcc GCC_GEPHY_DSP_HW_ARES>,
>      <&gcc GCC_GEPHY_RX_ARES>,
>      <&gcc GCC_GEPHY_TX_ARES>;
> to:
> 
> resets = <&gcc GCC_MISC_ARES>;
> 
> to conform to this bindings restriction in ethernet-phy.yaml
> 
>   resets:
>     maxItems: 1
> 
> Effectively, there's no functional change. So we can also list all the resets in the device tree, whatever is preferred.

+ Kathiravan

are there any recommendations from the hw team on which one to use?
As far as I can tell, the _MISC one simply pulls all the aforementioned
resets, like George described.. it seems weird that it would be designed
like this

Konrad