[PATCH 0/4] Drop unrelated clocks from SM8150/SC8180X PCIe RCs

[PATCH 0/4] Drop unrelated clocks from SM8150/SC8180X PCIe RCs

[Konrad Dybcio]

Smoke tested on both, but more is always welcome.

Signed-off-by: Konrad Dybcio <konrad.dybcio@oss.qualcomm.com>
---
Konrad Dybcio (4):
      dt-bindings: PCI: qcom,pcie-sc8180x: Drop unrelated clocks from PCIe hosts
      dt-bindings: PCI: qcom,pcie-sm8150: Drop unrelated clocks from PCIe hosts
      arm64: dts: qcom: sc8180x: Drop unrelated clocks from PCIe hosts
      arm64: dts: qcom: sm8150: Drop unrelated clocks from PCIe hosts

 .../devicetree/bindings/pci/qcom,pcie-sc8180x.yaml | 14 +++-------
 .../devicetree/bindings/pci/qcom,pcie-sm8150.yaml  | 14 +++-------
 arch/arm64/boot/dts/qcom/sc8180x.dtsi              | 32 ++++++----------------
 arch/arm64/boot/dts/qcom/sm8150.dtsi               | 16 +++--------
 4 files changed, 20 insertions(+), 56 deletions(-)
---
base-commit: edef457004774e598fc4c1b7d1d4f0bcd9d0bb30
change-id: 20250521-topic-8150_pcie_drop_clocks-f57a903ec8d3

Best regards,
-- 
Konrad Dybcio <konrad.dybcio@oss.qualcomm.com>

[PATCH 1/4] dt-bindings: PCI: qcom,pcie-sc8180x: Drop unrelated clocks from PCIe hosts

[Konrad Dybcio]

From: Konrad Dybcio <konrad.dybcio@oss.qualcomm.com>

The TBU clock belongs to the Translation Buffer Unit, part of the SMMU.
The ref clock is already being driven upstream through some of the
branches.

Signed-off-by: Konrad Dybcio <konrad.dybcio@oss.qualcomm.com>
---
 .../devicetree/bindings/pci/qcom,pcie-sc8180x.yaml         | 14 ++++----------
 1 file changed, 4 insertions(+), 10 deletions(-)

diff --git a/Documentation/devicetree/bindings/pci/qcom,pcie-sc8180x.yaml b/Documentation/devicetree/bindings/pci/qcom,pcie-sc8180x.yaml
index 331fc25d7a17d657d4db3863f0c538d0e44dc840..34a4d7b2c8459aeb615736f54c1971014adb205f 100644
--- a/Documentation/devicetree/bindings/pci/qcom,pcie-sc8180x.yaml
+++ b/Documentation/devicetree/bindings/pci/qcom,pcie-sc8180x.yaml
@@ -33,8 +33,8 @@ properties:
       - const: mhi # MHI registers
 
   clocks:
-    minItems: 8
-    maxItems: 8
+    minItems: 6
+    maxItems: 6
 
   clock-names:
     items:
@@ -44,8 +44,6 @@ properties:
       - const: bus_master # Master AXI clock
       - const: bus_slave # Slave AXI clock
       - const: slave_q2a # Slave Q2A clock
-      - const: ref # REFERENCE clock
-      - const: tbu # PCIe TBU clock
 
   interrupts:
     minItems: 8
@@ -117,17 +115,13 @@ examples:
                      <&gcc GCC_PCIE_0_CFG_AHB_CLK>,
                      <&gcc GCC_PCIE_0_MSTR_AXI_CLK>,
                      <&gcc GCC_PCIE_0_SLV_AXI_CLK>,
-                     <&gcc GCC_PCIE_0_SLV_Q2A_AXI_CLK>,
-                     <&gcc GCC_PCIE_0_CLKREF_CLK>,
-                     <&gcc GCC_AGGRE_NOC_PCIE_TBU_CLK>;
+                     <&gcc GCC_PCIE_0_SLV_Q2A_AXI_CLK>;
             clock-names = "pipe",
                           "aux",
                           "cfg",
                           "bus_master",
                           "bus_slave",
-                          "slave_q2a",
-                          "ref",
-                          "tbu";
+                          "slave_q2a";
 
             dma-coherent;
 

-- 
2.49.0

Re: [PATCH 1/4] dt-bindings: PCI: qcom,pcie-sc8180x: Drop unrelated clocks from PCIe hosts

[Rob Herring (Arm)]

On Wed, 21 May 2025 15:38:10 +0200, Konrad Dybcio wrote:
> From: Konrad Dybcio <konrad.dybcio@oss.qualcomm.com>
> 
> The TBU clock belongs to the Translation Buffer Unit, part of the SMMU.
> The ref clock is already being driven upstream through some of the
> branches.
> 
> Signed-off-by: Konrad Dybcio <konrad.dybcio@oss.qualcomm.com>
> ---
>  .../devicetree/bindings/pci/qcom,pcie-sc8180x.yaml         | 14 ++++----------
>  1 file changed, 4 insertions(+), 10 deletions(-)
> 

Acked-by: Rob Herring (Arm) <robh@kernel.org>

Re: [PATCH 1/4] dt-bindings: PCI: qcom,pcie-sc8180x: Drop unrelated clocks from PCIe hosts

[Manivannan Sadhasivam]

+ Krishna

On Wed, May 21, 2025 at 03:38:10PM +0200, Konrad Dybcio wrote:
> From: Konrad Dybcio <konrad.dybcio@oss.qualcomm.com>
> 
> The TBU clock belongs to the Translation Buffer Unit, part of the SMMU.
> The ref clock is already being driven upstream through some of the
> branches.
> 

Can you please cross check with the hardware programming guide (I don't have
access to atm) that the 'ref' clock is no longer voted by the driver?

- Mani

> Signed-off-by: Konrad Dybcio <konrad.dybcio@oss.qualcomm.com>
> ---
>  .../devicetree/bindings/pci/qcom,pcie-sc8180x.yaml         | 14 ++++----------
>  1 file changed, 4 insertions(+), 10 deletions(-)
> 
> diff --git a/Documentation/devicetree/bindings/pci/qcom,pcie-sc8180x.yaml b/Documentation/devicetree/bindings/pci/qcom,pcie-sc8180x.yaml
> index 331fc25d7a17d657d4db3863f0c538d0e44dc840..34a4d7b2c8459aeb615736f54c1971014adb205f 100644
> --- a/Documentation/devicetree/bindings/pci/qcom,pcie-sc8180x.yaml
> +++ b/Documentation/devicetree/bindings/pci/qcom,pcie-sc8180x.yaml
> @@ -33,8 +33,8 @@ properties:
>        - const: mhi # MHI registers
>  
>    clocks:
> -    minItems: 8
> -    maxItems: 8
> +    minItems: 6
> +    maxItems: 6
>  
>    clock-names:
>      items:
> @@ -44,8 +44,6 @@ properties:
>        - const: bus_master # Master AXI clock
>        - const: bus_slave # Slave AXI clock
>        - const: slave_q2a # Slave Q2A clock
> -      - const: ref # REFERENCE clock
> -      - const: tbu # PCIe TBU clock
>  
>    interrupts:
>      minItems: 8
> @@ -117,17 +115,13 @@ examples:
>                       <&gcc GCC_PCIE_0_CFG_AHB_CLK>,
>                       <&gcc GCC_PCIE_0_MSTR_AXI_CLK>,
>                       <&gcc GCC_PCIE_0_SLV_AXI_CLK>,
> -                     <&gcc GCC_PCIE_0_SLV_Q2A_AXI_CLK>,
> -                     <&gcc GCC_PCIE_0_CLKREF_CLK>,
> -                     <&gcc GCC_AGGRE_NOC_PCIE_TBU_CLK>;
> +                     <&gcc GCC_PCIE_0_SLV_Q2A_AXI_CLK>;
>              clock-names = "pipe",
>                            "aux",
>                            "cfg",
>                            "bus_master",
>                            "bus_slave",
> -                          "slave_q2a",
> -                          "ref",
> -                          "tbu";
> +                          "slave_q2a";
>  
>              dma-coherent;
>  
> 
> -- 
> 2.49.0
> 

-- 
மணிவண்ணன் சதாசிவம்

Re: [PATCH 1/4] dt-bindings: PCI: qcom,pcie-sc8180x: Drop unrelated clocks from PCIe hosts

[Qiang Yu]

On Fri, Jun 13, 2025 at 02:43:38PM +0530, Manivannan Sadhasivam wrote:
> + Krishna
> 
> On Wed, May 21, 2025 at 03:38:10PM +0200, Konrad Dybcio wrote:
> > From: Konrad Dybcio <konrad.dybcio@oss.qualcomm.com>
> > 
> > The TBU clock belongs to the Translation Buffer Unit, part of the SMMU.
> > The ref clock is already being driven upstream through some of the
> > branches.
> > 
> 
> Can you please cross check with the hardware programming guide (I don't have
> access to atm) that the 'ref' clock is no longer voted by the driver?
>

CLKREF is required for PHY. Since it has been voted in PCIe PHY driver,
omitting it here is reasonable.

- Qiang Yu

> - Mani
> 
> > Signed-off-by: Konrad Dybcio <konrad.dybcio@oss.qualcomm.com>
> > ---
> >  .../devicetree/bindings/pci/qcom,pcie-sc8180x.yaml         | 14 ++++----------
> >  1 file changed, 4 insertions(+), 10 deletions(-)
> > 
> > diff --git a/Documentation/devicetree/bindings/pci/qcom,pcie-sc8180x.yaml b/Documentation/devicetree/bindings/pci/qcom,pcie-sc8180x.yaml
> > index 331fc25d7a17d657d4db3863f0c538d0e44dc840..34a4d7b2c8459aeb615736f54c1971014adb205f 100644
> > --- a/Documentation/devicetree/bindings/pci/qcom,pcie-sc8180x.yaml
> > +++ b/Documentation/devicetree/bindings/pci/qcom,pcie-sc8180x.yaml
> > @@ -33,8 +33,8 @@ properties:
> >        - const: mhi # MHI registers
> >  
> >    clocks:
> > -    minItems: 8
> > -    maxItems: 8
> > +    minItems: 6
> > +    maxItems: 6
> >  
> >    clock-names:
> >      items:
> > @@ -44,8 +44,6 @@ properties:
> >        - const: bus_master # Master AXI clock
> >        - const: bus_slave # Slave AXI clock
> >        - const: slave_q2a # Slave Q2A clock
> > -      - const: ref # REFERENCE clock
> > -      - const: tbu # PCIe TBU clock
> >  
> >    interrupts:
> >      minItems: 8
> > @@ -117,17 +115,13 @@ examples:
> >                       <&gcc GCC_PCIE_0_CFG_AHB_CLK>,
> >                       <&gcc GCC_PCIE_0_MSTR_AXI_CLK>,
> >                       <&gcc GCC_PCIE_0_SLV_AXI_CLK>,
> > -                     <&gcc GCC_PCIE_0_SLV_Q2A_AXI_CLK>,
> > -                     <&gcc GCC_PCIE_0_CLKREF_CLK>,
> > -                     <&gcc GCC_AGGRE_NOC_PCIE_TBU_CLK>;
> > +                     <&gcc GCC_PCIE_0_SLV_Q2A_AXI_CLK>;
> >              clock-names = "pipe",
> >                            "aux",
> >                            "cfg",
> >                            "bus_master",
> >                            "bus_slave",
> > -                          "slave_q2a",
> > -                          "ref",
> > -                          "tbu";
> > +                          "slave_q2a";
> >  
> >              dma-coherent;
> >  
> > 
> > -- 
> > 2.49.0
> > 
> 
> -- 
> மணிவண்ணன் சதாசிவம்
> 

[PATCH 2/4] dt-bindings: PCI: qcom,pcie-sm8150: Drop unrelated clocks from PCIe hosts

[Konrad Dybcio]

From: Konrad Dybcio <konrad.dybcio@oss.qualcomm.com>

The TBU clock belongs to the Translation Buffer Unit, part of the SMMU.
The ref clock is already being driven upstream through some of the
branches.

Signed-off-by: Konrad Dybcio <konrad.dybcio@oss.qualcomm.com>
---
 .../devicetree/bindings/pci/qcom,pcie-sm8150.yaml          | 14 ++++----------
 1 file changed, 4 insertions(+), 10 deletions(-)

diff --git a/Documentation/devicetree/bindings/pci/qcom,pcie-sm8150.yaml b/Documentation/devicetree/bindings/pci/qcom,pcie-sm8150.yaml
index a604f2a79de3b28863a0b8933e6679df4953402c..434448cd816ad4241ade402bbc621efc49646bdd 100644
--- a/Documentation/devicetree/bindings/pci/qcom,pcie-sm8150.yaml
+++ b/Documentation/devicetree/bindings/pci/qcom,pcie-sm8150.yaml
@@ -33,8 +33,8 @@ properties:
       - const: mhi # MHI registers
 
   clocks:
-    minItems: 8
-    maxItems: 8
+    minItems: 6
+    maxItems: 6
 
   clock-names:
     items:
@@ -44,8 +44,6 @@ properties:
       - const: bus_master # Master AXI clock
       - const: bus_slave # Slave AXI clock
       - const: slave_q2a # Slave Q2A clock
-      - const: tbu # PCIe TBU clock
-      - const: ref # REFERENCE clock
 
   interrupts:
     minItems: 8
@@ -111,17 +109,13 @@ examples:
                      <&gcc GCC_PCIE_0_CFG_AHB_CLK>,
                      <&gcc GCC_PCIE_0_MSTR_AXI_CLK>,
                      <&gcc GCC_PCIE_0_SLV_AXI_CLK>,
-                     <&gcc GCC_PCIE_0_SLV_Q2A_AXI_CLK>,
-                     <&gcc GCC_AGGRE_NOC_PCIE_TBU_CLK>,
-                     <&rpmhcc RPMH_CXO_CLK>;
+                     <&gcc GCC_PCIE_0_SLV_Q2A_AXI_CLK>;
             clock-names = "pipe",
                           "aux",
                           "cfg",
                           "bus_master",
                           "bus_slave",
-                          "slave_q2a",
-                          "tbu",
-                          "ref";
+                          "slave_q2a";
 
             interrupts = <GIC_SPI 141 IRQ_TYPE_LEVEL_HIGH>,
                          <GIC_SPI 142 IRQ_TYPE_LEVEL_HIGH>,

-- 
2.49.0

Re: [PATCH 2/4] dt-bindings: PCI: qcom,pcie-sm8150: Drop unrelated clocks from PCIe hosts

[Rob Herring (Arm)]

On Wed, 21 May 2025 15:38:11 +0200, Konrad Dybcio wrote:
> From: Konrad Dybcio <konrad.dybcio@oss.qualcomm.com>
> 
> The TBU clock belongs to the Translation Buffer Unit, part of the SMMU.
> The ref clock is already being driven upstream through some of the
> branches.
> 
> Signed-off-by: Konrad Dybcio <konrad.dybcio@oss.qualcomm.com>
> ---
>  .../devicetree/bindings/pci/qcom,pcie-sm8150.yaml          | 14 ++++----------
>  1 file changed, 4 insertions(+), 10 deletions(-)
> 

Acked-by: Rob Herring (Arm) <robh@kernel.org>

[PATCH 3/4] arm64: dts: qcom: sc8180x: Drop unrelated clocks from PCIe hosts

[Konrad Dybcio]

From: Konrad Dybcio <konrad.dybcio@oss.qualcomm.com>

The TBU clock belongs to the Translation Buffer Unit, part of the SMMU.
The ref clock is already being driven upstream through some of the
branches.

Signed-off-by: Konrad Dybcio <konrad.dybcio@oss.qualcomm.com>
---
 arch/arm64/boot/dts/qcom/sc8180x.dtsi | 32 ++++++++------------------------
 1 file changed, 8 insertions(+), 24 deletions(-)

diff --git a/arch/arm64/boot/dts/qcom/sc8180x.dtsi b/arch/arm64/boot/dts/qcom/sc8180x.dtsi
index b84e47a461a014871ef11e08d18af70bec8e2d63..fa8bd1ddfb39c3d46095f94a6c97fedb71db00ba 100644
--- a/arch/arm64/boot/dts/qcom/sc8180x.dtsi
+++ b/arch/arm64/boot/dts/qcom/sc8180x.dtsi
@@ -1747,17 +1747,13 @@ pcie0: pcie@1c00000 {
 				 <&gcc GCC_PCIE_0_CFG_AHB_CLK>,
 				 <&gcc GCC_PCIE_0_MSTR_AXI_CLK>,
 				 <&gcc GCC_PCIE_0_SLV_AXI_CLK>,
-				 <&gcc GCC_PCIE_0_SLV_Q2A_AXI_CLK>,
-				 <&gcc GCC_PCIE_0_CLKREF_CLK>,
-				 <&gcc GCC_AGGRE_NOC_PCIE_TBU_CLK>;
+				 <&gcc GCC_PCIE_0_SLV_Q2A_AXI_CLK>;
 			clock-names = "pipe",
 				      "aux",
 				      "cfg",
 				      "bus_master",
 				      "bus_slave",
-				      "slave_q2a",
-				      "ref",
-				      "tbu";
+				      "slave_q2a";
 
 			assigned-clocks = <&gcc GCC_PCIE_0_AUX_CLK>;
 			assigned-clock-rates = <19200000>;
@@ -1868,17 +1864,13 @@ pcie3: pcie@1c08000 {
 				 <&gcc GCC_PCIE_3_CFG_AHB_CLK>,
 				 <&gcc GCC_PCIE_3_MSTR_AXI_CLK>,
 				 <&gcc GCC_PCIE_3_SLV_AXI_CLK>,
-				 <&gcc GCC_PCIE_3_SLV_Q2A_AXI_CLK>,
-				 <&gcc GCC_PCIE_3_CLKREF_CLK>,
-				 <&gcc GCC_AGGRE_NOC_PCIE_TBU_CLK>;
+				 <&gcc GCC_PCIE_3_SLV_Q2A_AXI_CLK>;
 			clock-names = "pipe",
 				      "aux",
 				      "cfg",
 				      "bus_master",
 				      "bus_slave",
-				      "slave_q2a",
-				      "ref",
-				      "tbu";
+				      "slave_q2a";
 
 			assigned-clocks = <&gcc GCC_PCIE_3_AUX_CLK>;
 			assigned-clock-rates = <19200000>;
@@ -1990,17 +1982,13 @@ pcie1: pcie@1c10000 {
 				 <&gcc GCC_PCIE_1_CFG_AHB_CLK>,
 				 <&gcc GCC_PCIE_1_MSTR_AXI_CLK>,
 				 <&gcc GCC_PCIE_1_SLV_AXI_CLK>,
-				 <&gcc GCC_PCIE_1_SLV_Q2A_AXI_CLK>,
-				 <&gcc GCC_PCIE_1_CLKREF_CLK>,
-				 <&gcc GCC_AGGRE_NOC_PCIE_TBU_CLK>;
+				 <&gcc GCC_PCIE_1_SLV_Q2A_AXI_CLK>;
 			clock-names = "pipe",
 				      "aux",
 				      "cfg",
 				      "bus_master",
 				      "bus_slave",
-				      "slave_q2a",
-				      "ref",
-				      "tbu";
+				      "slave_q2a";
 
 			assigned-clocks = <&gcc GCC_PCIE_1_AUX_CLK>;
 			assigned-clock-rates = <19200000>;
@@ -2112,17 +2100,13 @@ pcie2: pcie@1c18000 {
 				 <&gcc GCC_PCIE_2_CFG_AHB_CLK>,
 				 <&gcc GCC_PCIE_2_MSTR_AXI_CLK>,
 				 <&gcc GCC_PCIE_2_SLV_AXI_CLK>,
-				 <&gcc GCC_PCIE_2_SLV_Q2A_AXI_CLK>,
-				 <&gcc GCC_PCIE_2_CLKREF_CLK>,
-				 <&gcc GCC_AGGRE_NOC_PCIE_TBU_CLK>;
+				 <&gcc GCC_PCIE_2_SLV_Q2A_AXI_CLK>;
 			clock-names = "pipe",
 				      "aux",
 				      "cfg",
 				      "bus_master",
 				      "bus_slave",
-				      "slave_q2a",
-				      "ref",
-				      "tbu";
+				      "slave_q2a";
 
 			assigned-clocks = <&gcc GCC_PCIE_2_AUX_CLK>;
 			assigned-clock-rates = <19200000>;

-- 
2.49.0

[PATCH 4/4] arm64: dts: qcom: sm8150: Drop unrelated clocks from PCIe hosts

[Konrad Dybcio]

From: Konrad Dybcio <konrad.dybcio@oss.qualcomm.com>

The TBU clock belongs to the Translation Buffer Unit, part of the SMMU.
The ref clock is already being driven upstream through some of the
branches.

Signed-off-by: Konrad Dybcio <konrad.dybcio@oss.qualcomm.com>
---
 arch/arm64/boot/dts/qcom/sm8150.dtsi | 16 ++++------------
 1 file changed, 4 insertions(+), 12 deletions(-)

diff --git a/arch/arm64/boot/dts/qcom/sm8150.dtsi b/arch/arm64/boot/dts/qcom/sm8150.dtsi
index cdb47359c4c88af5c73956ba0ba1710ca312a9af..da3d86a17a4f61dc0d1318efeaaf43ac213835ab 100644
--- a/arch/arm64/boot/dts/qcom/sm8150.dtsi
+++ b/arch/arm64/boot/dts/qcom/sm8150.dtsi
@@ -1874,17 +1874,13 @@ pcie0: pcie@1c00000 {
 				 <&gcc GCC_PCIE_0_CFG_AHB_CLK>,
 				 <&gcc GCC_PCIE_0_MSTR_AXI_CLK>,
 				 <&gcc GCC_PCIE_0_SLV_AXI_CLK>,
-				 <&gcc GCC_PCIE_0_SLV_Q2A_AXI_CLK>,
-				 <&gcc GCC_AGGRE_NOC_PCIE_TBU_CLK>,
-				 <&rpmhcc RPMH_CXO_CLK>;
+				 <&gcc GCC_PCIE_0_SLV_Q2A_AXI_CLK>;
 			clock-names = "pipe",
 				      "aux",
 				      "cfg",
 				      "bus_master",
 				      "bus_slave",
-				      "slave_q2a",
-				      "tbu",
-				      "ref";
+				      "slave_q2a";
 
 			iommu-map = <0x0   &apps_smmu 0x1d80 0x1>,
 				    <0x100 &apps_smmu 0x1d81 0x1>;
@@ -1991,17 +1987,13 @@ pcie1: pcie@1c08000 {
 				 <&gcc GCC_PCIE_1_CFG_AHB_CLK>,
 				 <&gcc GCC_PCIE_1_MSTR_AXI_CLK>,
 				 <&gcc GCC_PCIE_1_SLV_AXI_CLK>,
-				 <&gcc GCC_PCIE_1_SLV_Q2A_AXI_CLK>,
-				 <&gcc GCC_AGGRE_NOC_PCIE_TBU_CLK>,
-				 <&rpmhcc RPMH_CXO_CLK>;
+				 <&gcc GCC_PCIE_1_SLV_Q2A_AXI_CLK>;
 			clock-names = "pipe",
 				      "aux",
 				      "cfg",
 				      "bus_master",
 				      "bus_slave",
-				      "slave_q2a",
-				      "tbu",
-				      "ref";
+				      "slave_q2a";
 
 			assigned-clocks = <&gcc GCC_PCIE_1_AUX_CLK>;
 			assigned-clock-rates = <19200000>;

-- 
2.49.0

Re: (subset) [PATCH 0/4] Drop unrelated clocks from SM8150/SC8180X PCIe RCs

[Manivannan Sadhasivam]

On Wed, 21 May 2025 15:38:09 +0200, Konrad Dybcio wrote:
> Smoke tested on both, but more is always welcome.
> 
> 

Applied, thanks!

[1/4] dt-bindings: PCI: qcom,pcie-sc8180x: Drop unrelated clocks from PCIe hosts
      commit: 26daa18e35ebc4e192ff55d021f1cd7e69d55487
[2/4] dt-bindings: PCI: qcom,pcie-sm8150: Drop unrelated clocks from PCIe hosts
      commit: e1cb67ab82aab44cda410616498d4749399da217

Best regards,
-- 
Manivannan Sadhasivam <mani@kernel.org>