Re: [GIT PULL] Qualcomm clock driver updates for v6.3

[Kathiravan T <quic_kathirav@quicinc.com>]

On 2/21/2023 9:14 PM, Bjorn Andersson wrote:
> On Mon, Feb 20, 2023 at 10:10:07PM +0530, Kathiravan T wrote:
>> On 2/20/2023 9:05 PM, Bjorn Andersson wrote:
>>> On Mon, Feb 20, 2023 at 05:17:39PM +0530, Kathiravan T wrote:
>>>> On 2/19/2023 10:23 AM, Bjorn Andersson wrote:
>>>>> The following changes since commit 1b929c02afd37871d5afb9d498426f83432e71c2:
>>>>>
>>>>>      Linux 6.2-rc1 (2022-12-25 13:41:39 -0800)
>>>>>
>>>>> are available in the Git repository at:
>>>>>
>>>>>      https://git.kernel.org/pub/scm/linux/kernel/git/qcom/linux.git tags/qcom-clk-for-6.3
>>>>>
>>>>> for you to fetch changes up to 7935b534d32a1823a7d5db449d340f56c201f284:
>>>>>
>>>>>      dt-bindings: clock: Merge qcom,gpucc-sm8350 into qcom,gpucc.yaml (2023-02-08 19:29:25 -0800)
>>>>>
>>>>> ----------------------------------------------------------------
>>>>> Qualcomm clock driver updates for v6.3
>>>>>
>>>>> Support for requesting the next power_off operation for a genpd to be
>>>>> synchronous is introduced, and implemented in the GDSC driver. To allow
>>>>> the GPU driver to request power_off to wait for the GDSC to actually
>>>>> collapse.
>>>>>
>>>>> clk_disable_unused() is altered, to not consider clocks which comes from
>>>>> a provider with sync_state defined. This makes it possible for
>>>>> individual clock drivers to invoke this process once all clients has
>>>>> been probed, avoiding the need for booting many systems with
>>>>> clk_ignore_unused.
>>>>> This is then enabled for SDM845 and SC8280XP clock controllers.
>>>>>
>>>>> Support for QDU1000/QRU1000 Global clock controller, SA8775P Global
>>>>> clock controller, SM8550 TCSR and display clock controller, SM6350 clock
>>>>> controller, nd MSM8996 CBF and APCS clock controllers is introduced..
>>>>>
>>>>> Parent references are updated across a large number of clock drivers, to
>>>>> align with the design changes since those drivers where introduced.
>>>>> Similarly, test clocks has been dropped from a range of drivers.
>>>>>
>>>>> A range of fixes for the MSM8996 CPU clock controller is introduced.
>>>>>
>>>>> MSM8974 GCC is transitioned off the externally defined sleep_clk.
>>>>>
>>>>> GDSC in the global clock controller for QCS404 is added, and various
>>>>> parent definitions are cleaned up.
>>>>>
>>>>> The SDCC core clocks on SM6115 are moved for floor_ops.
>>>>>
>>>>> Programming of clk_dis_wait for GPU CX GDSC on SC7180 and SDM845 are
>>>>> moved to use the recently introduced properties in the GDSC struct.
>>>>>
>>>>> The RPMh clock driver gains SM8550 and SA8775P clocks, and the IPA clock
>>>>> is added on a variety of platforms.
>>>>>
>>>>> The SMD RPM driver receives a big cleanup, in particular a move away
>>>>> from duplicating declaration of identical clocks between multiple
>>>>> platforms.
>>>>> A few missing clocks across msm8998, msm8992, msm8916, qcs404 are added
>>>>> as well.
>>>>>
>>>>> Using devm_pm_runtime_enable() to clean up some duplication is done
>>>>> across SM8250 display and video clock controllers, SM8450 display clock
>>>>> controller and SC7280 LPASS clock controller.
>>>>>
>>>>> Devicetree binding changes for above mentioned additions and changes are
>>>>> introduced.
>>>>>
>>>>> Lastly, a change to pad a few registers in the SM8250 DTS to 8 digits
>>>>> was picked up in the wrong tree and kept here, to avoid rebasing.
>>>>>
>>>>> ----------------------------------------------------------------
>>>> <snip>
>>>>
>>>> Bjorn,
>>>>
>>> Hi Kathiravan,
>>>
>>>> I don't see the change got picked up v6.3 https://lore.kernel.org/linux-arm-msm/20230120082631.22053-1-quic_kathirav@quicinc.com/.
>>>> Please help to understand what is missing to get it picked up?
>>>>
>>> Afaict, this patch introduces support that doesn't have a user today,
>>> and as such should be merged together with:
>>> https://lore.kernel.org/linux-arm-msm/20230217083308.12017-4-quic_kathirav@quicinc.com/
>>
>> Thanks Bjorn. I got the point. This patch also needed for IPQ5332 Baseport
>> patches https://lore.kernel.org/linux-gpio/20230217075835.460-1-quic_kathirav@quicinc.com/
>>
>>
>>> Which hasn't yet been picked up...
>>>
>>> But please send such patches together in the future, to make it easier
>>> for people to review and me to merge them.
>>
>> If I re-spin the IPQ5332 base port patches, shall I add this patch to the
>> IPQ5332 series itself and send it or you are okay to pick it up as it is?
>>
> Doesn't:
> https://lore.kernel.org/linux-gpio/20230217075835.460-1-quic_kathirav@quicinc.com/
> depend on:
> https://lore.kernel.org/linux-arm-msm/20230217083308.12017-4-quic_kathirav@quicinc.com/
> being merged first?


Both the baseport[1] and APSS clock driver[2] series depends on the 
patch 
https://lore.kernel.org/linux-arm-msm/20230120082631.22053-1-quic_kathirav@quicinc.com/. 
GCC driver which is part of baseport uses the clk_alpha_pll_stromer_ops 
which is introduced in the mentioned patch.  However there is no 
dependency between baseport and APSS clock driver apart from DTS changes.


[1] 
https://lore.kernel.org/linux-gpio/20230217075835.460-1-quic_kathirav@quicinc.com/

[2] 
https://lore.kernel.org/linux-arm-msm/20230217083308.12017-1-quic_kathirav@quicinc.com/


>
> If you are going to resubmit the IPQ5332 baseport, please consider
> sending one series with all the clock patches together.
>
> You may send the dts separately, mentioning the dependency on the clock
> series - or any other independent driver series that you sent out. That
> way you make it easy for each maintainer to pick the patches destined
> for their tree.


Thanks got it. In the next re-spin let me include the mentioned patch 
along with the baseport series.


Thanks, Kathiravan T.


> Regards,
> Bjorn