Re: [PATCH 7/8] arm64: dts: qcom: msm8996: Add MSM8996 Pro support

[Yassine Oudjana <y.oudjana@protonmail.com>]

On Sat, Oct 16 2021 at 19:09:43 +0400, Dmitry Baryshkov 
<dmitry.baryshkov@linaro.org> wrote:
> On Sat, 16 Oct 2021 at 17:51, Yassine Oudjana 
> <y.oudjana@protonmail.com> wrote:
>> 
>> 
>>  On Fri, Oct 15 2021 at 23:01:54 +0400, Konrad Dybcio
>>  <konrad.dybcio@somainline.org> wrote:
>>  >
>>  > On 14.10.2021 10:32, Yassine Oudjana wrote:
>>  >>  Add a new DTSI for MSM8996 Pro (MSM8996SG) with msm-id and 
>> CPU/GPU
>>  >> OPPs.
>>  >>  CBF OPPs and CPR parameters will be added to it as well once
>>  >> support for
>>  >>  CBF scaling and CPR is introduced.
>>  >>
>>  >>  Signed-off-by: Yassine Oudjana <y.oudjana@protonmail.com>
>>  >>  ---
>>  >>   arch/arm64/boot/dts/qcom/msm8996.dtsi    |  82 +++----
>>  >>   arch/arm64/boot/dts/qcom/msm8996pro.dtsi | 281
>>  >> +++++++++++++++++++++++
>>  >>   2 files changed, 322 insertions(+), 41 deletions(-)
>>  >>   create mode 100644 arch/arm64/boot/dts/qcom/msm8996pro.dtsi
>>  >>
>>  >>  diff --git a/arch/arm64/boot/dts/qcom/msm8996.dtsi
>>  >> b/arch/arm64/boot/dts/qcom/msm8996.dtsi
>>  >>  index 94a846c3f1ee..5b2600a4fb2a 100644
>>  >>  --- a/arch/arm64/boot/dts/qcom/msm8996.dtsi
>>  >>  +++ b/arch/arm64/boot/dts/qcom/msm8996.dtsi
>>  >>  @@ -142,82 +142,82 @@ cluster0_opp: opp_table0 {
>>  >>              /* Nominal fmax for now */
>>  >>              opp-307200000 {
>>  >>                      opp-hz = /bits/ 64 <307200000>;
>>  >>  -                   opp-supported-hw = <0x77>;
>>  >>  +                   opp-supported-hw = <0x7>;
>>  >
>>  > You didn't describe what's the reason for changing this 
>> everywhere.
>>  >
>>  > If it's been always broken, perhaps make it a separate commit
>>  > describing
>>  >
>>  > the issue.
>>  >
>>  >
>>  > Konrad
>>  >
>> 
>>  Before removing reading msm-id in qcom_cpufreq_nvmem, bits 0-2 
>> (0x07)
>>  were MSM8996 speed bins, while bits 4-6 (0x70) were MSM8996 Pro 
>> speed
>>  bins. Now, only bits 0-2 are used for either one, so basically I 
>> moved
>>  bits 4-6 into msm8996pro.dtsi after shifting them right to become 
>> bits
>>  0-2.
>> 
>>  I'll put this in a separate patch and describe the change.
> 
> Could you please describe in the commit message why is it changed?
> IOW, what prompted you to split 8996SG support from main msm8996.dtsi?
> 

I will add that as well. For now it is described in the cover of this 
series[1].

	Yassine

> --
> With best wishes
> Dmitry
> 

[1] 
https://lore.kernel.org/linux-arm-msm/20211014083016.137441-1-y.oudjana@protonmail.com/T/#m01711ca6f962760cc9283965c5a82f8455fbe06c