| Message ID | 1688545032-17748-1-git-send-email-quic_msarkar@quicinc.com (mailing list archive) |
|---|---|
| Headers | show |
| Series | arm64: qcom: sa8775p: add support for PCIe | expand |
On Wed, Jul 05, 2023 at 01:47:05PM +0530, Mrinmay Sarkar wrote: > Update the relavent DT bindings for PCIe, add new config to the phy > driver add pcie and phy nodes to the .dtsi file and enable then in > board .dts file for the sa8775p-ride platform. > > Mrinmay Sarkar (6): > dt-bindings: PCI: qcom: Add sa8775p compatible > dt-bindings: phy: qcom,qmp: Add sa8775p QMP PCIe PHY > PCI: qcom: Add support for sa8775p SoC > phy: qcom-qmp-pcie: add support for sa8775p > arm64: dts: qcom: sa8775p: Add pcie0 and pcie1 nodes > arm64: dts: qcom: sa8775p-ride: enable pcie nodes Please note that the dts patches should come before driver patches. - Mani > > .../devicetree/bindings/pci/qcom,pcie.yaml | 28 ++ > .../phy/qcom,sc8280xp-qmp-pcie-phy.yaml | 19 +- > arch/arm64/boot/dts/qcom/sa8775p-ride.dts | 68 ++++ > arch/arm64/boot/dts/qcom/sa8775p.dtsi | 201 ++++++++++- > drivers/pci/controller/dwc/pcie-qcom.c | 1 + > drivers/phy/qualcomm/phy-qcom-qmp-pcie.c | 328 ++++++++++++++++++ > .../qualcomm/phy-qcom-qmp-pcs-pcie-v5_20.h | 1 + > .../phy-qcom-qmp-qserdes-txrx-v5_20.h | 2 + > 8 files changed, 645 insertions(+), 3 deletions(-) > > -- > 2.39.2 >
On 05/07/2023 10:46, Manivannan Sadhasivam wrote: > On Wed, Jul 05, 2023 at 01:47:05PM +0530, Mrinmay Sarkar wrote: >> Update the relavent DT bindings for PCIe, add new config to the phy >> driver add pcie and phy nodes to the .dtsi file and enable then in >> board .dts file for the sa8775p-ride platform. >> >> Mrinmay Sarkar (6): >> dt-bindings: PCI: qcom: Add sa8775p compatible >> dt-bindings: phy: qcom,qmp: Add sa8775p QMP PCIe PHY >> PCI: qcom: Add support for sa8775p SoC >> phy: qcom-qmp-pcie: add support for sa8775p >> arm64: dts: qcom: sa8775p: Add pcie0 and pcie1 nodes >> arm64: dts: qcom: sa8775p-ride: enable pcie nodes > > Please note that the dts patches should come before driver patches. > Why? DTS is always independent thus usually put at the end of patchset or better separate. It is the first time I hear that DTS should be before driver. Best regards, Krzysztof
On Thu, Jul 06, 2023 at 08:39:54AM +0200, Krzysztof Kozlowski wrote: > On 05/07/2023 10:46, Manivannan Sadhasivam wrote: > > On Wed, Jul 05, 2023 at 01:47:05PM +0530, Mrinmay Sarkar wrote: > >> Update the relavent DT bindings for PCIe, add new config to the phy > >> driver add pcie and phy nodes to the .dtsi file and enable then in > >> board .dts file for the sa8775p-ride platform. > >> > >> Mrinmay Sarkar (6): > >> dt-bindings: PCI: qcom: Add sa8775p compatible > >> dt-bindings: phy: qcom,qmp: Add sa8775p QMP PCIe PHY > >> PCI: qcom: Add support for sa8775p SoC > >> phy: qcom-qmp-pcie: add support for sa8775p > >> arm64: dts: qcom: sa8775p: Add pcie0 and pcie1 nodes > >> arm64: dts: qcom: sa8775p-ride: enable pcie nodes > > > > Please note that the dts patches should come before driver patches. > > > > Why? DTS is always independent thus usually put at the end of patchset > or better separate. It is the first time I hear that DTS should be > before driver. > This is what I was suggested by Rob during my initial days and I've been following this pattern since then. If that's not the case, I have no issues. - Mani > Best regards, > Krzysztof >
On 06/07/2023 13:12, Manivannan Sadhasivam wrote: >>> Please note that the dts patches should come before driver patches. >>> >> >> Why? DTS is always independent thus usually put at the end of patchset >> or better separate. It is the first time I hear that DTS should be >> before driver. >> > > This is what I was suggested by Rob during my initial days and I've been > following this pattern since then. If that's not the case, I have no issues. I really doubt Rob ever suggested that. He probably suggested that bindings must come before user, but not the DTS. Best regards, Krzysztof
On 6.07.2023 14:09, Krzysztof Kozlowski wrote: > On 06/07/2023 13:12, Manivannan Sadhasivam wrote: >>>> Please note that the dts patches should come before driver patches. >>>> >>> >>> Why? DTS is always independent thus usually put at the end of patchset >>> or better separate. It is the first time I hear that DTS should be >>> before driver. >>> >> >> This is what I was suggested by Rob during my initial days and I've been >> following this pattern since then. If that's not the case, I have no issues. > > I really doubt Rob ever suggested that. He probably suggested that > bindings must come before user, but not the DTS. Besides, the flow in this patchset is very logical from the Linux POV 1. describe the hardware in bindings 2. implement the drivers, conforming to the description 3. describe the hardware in the device tree to ensure the (now existing) driver probes Konrad > > Best regards, > Krzysztof >
On Thu, Jul 06, 2023 at 02:09:09PM +0200, Krzysztof Kozlowski wrote: > On 06/07/2023 13:12, Manivannan Sadhasivam wrote: > >>> Please note that the dts patches should come before driver patches. > >>> > >> > >> Why? DTS is always independent thus usually put at the end of patchset > >> or better separate. It is the first time I hear that DTS should be > >> before driver. > >> > > > > This is what I was suggested by Rob during my initial days and I've been > > following this pattern since then. If that's not the case, I have no issues. > > I really doubt Rob ever suggested that. He probably suggested that > bindings must come before user, but not the DTS. > Maybe I misunderstood it to be both bindings and dts. Thanks for clarifying. - Mani > Best regards, > Krzysztof >