[PATCH v2 0/5] PCI: qcom: Add PCIe support for Shikra SoC

[Sushrut Shree Trivedi]

On 9/2/2026 2:16 PM, Konrad Dybcio wrote:
> On 8/31/26 9:26 AM, Sushrut Shree Trivedi wrote:
>> Add PCIe support for Shikra target, by adding dt-bindings for phy,
>> controller and corresponding phy & controller drivers/device-tree
>> changes.
>>
>> Shikra RC is connected to TC9563 PCIe switch on all three
>> EVK variants: CQS, CQM and IQS. The individual downstream ports
>> of TC9563 connect like below:
>>
>> DSP1: M.2 B-Key for 5G Modem
>> DSP2: M.2 M-Key for NVMe
>> DSP3: Embedded ethernet device
>>
>> Power and reset to M.2 B and M.2 M slot are controlled via
>> TC9563 GPIO's. Since TC9563 GPIO driver from Alex is still
>> under active review, current series only contains basic PCIe
>> enablement for the controller / PHY. TC9563 enablement changes
>> would be posted as a separate series once the GPIO driver is
>> ready.
>>
>> [1] https://lore.kernel.org/all/e2inl7k5gsjj6oomv2k5ximuzpb3gfiz66ufet3b4hvov7zqt4@qz4pifbos7yf/
>>
>> Signed-off-by: Sushrut Shree Trivedi <sushrut.trivedi at oss.qualcomm.com>
>> ---
>> Changes in v2:
>> - Remove TC9563 specific changes [Mani]
>> - Fix commit descriptions/subjects [Bjorn Helgaas]
>> - Fix PCIe controller dt-binding errors [Krzysztof Kozlowski / Mani]
>> - Document PCIe Gen2 limitation in controller node & commit msg [Konrad Dybcio / Mani]
>> - Add GCC_PCIE_SLEEP_CLK (which is actually PHY AUX CLK) in PCIe PHY
>>    node.
> You also silently adjusted the OPP levels in the DT patch:
My bad;  I did ACK Mani's comment in v1 regarding higher voltage corner 
votes.
Will mention in cover letter next time onwards.
>
> < +                                     required-opps = <&rpmpd_opp_nom>;
> ---
>> +                                     required-opps = <&rpmpd_opp_low_svs>;
>
> is the lower one suitable?
Yes, Gen2 can be supported with low svs corner as per HW team.
>
> Konrad