dt-bindings: imx6q-pcie: Handle various clock configurations
Subsystems (5)
PATCH v4 1/3 dt-bindings: imx6q-pcie: Handle various clock configurations
The i.MX SoCs have various clock configurations routed into the PCIe IP, the list of clock is below. Document all those configurations in the DT binding document.
All SoCs: pcie, pcie_bus
6QDL, 7D: + pcie_phy
6SX: + pcie_phy pcie_inbound_axi
8MQ: + pcie_phy pcie_aux
8MM, 8MP: + pcie_aux
---
Cc: Fabio Estevam <[email protected]>
Cc: Krzysztof Kozlowski <[email protected]>
Cc: Lucas Stach <[email protected]>
Cc: Richard Zhu <[email protected]>
Cc: Rob Herring <[email protected]>
Cc: Shawn Guo <[email protected]>
Cc: [email protected]
Cc: NXP Linux Team <[email protected]>
To: [email protected]
---
V2: - Add AB from Alex
V3: - Duplicate clock-names maxItems to mx6sx and mx8mq compatibles
- Flatten the if-else structure
- The validation no longer works and introduces errors like these:
arch/arm64/boot/dts/freescale/imx8mm-verdin-wifi-dahlia.dtb: pcie@33800000: clock-names:2: 'pcie_phy' was expected
V4: - Reinstate minItems: for clock-names in main section, turn the
last two clock-names items into enums to cover all IP variants.
- Add another allOf entry for mx6q/mx6qp/mx7d clock-names list.
- Adjust clock maxItems in the allOf section.
--- .../bindings/pci/fsl,imx6q-pcie.yaml | 73 +++++++++++++++++-- 1 file changed, 68 insertions(+), 5 deletions(-)
Documentation/devicetree/bindings/pci/fsl,imx6q-pcie.yaml +68 -5
@@ -14,9 +14,6 @@ description: |+ This PCIe host controller is based on the Synopsys DesignWare PCIe IP and thus inherits all the common properties defined in snps,dw-pcie.yaml. -allOf: - - $ref: /schemas/pci/snps,dw-pcie.yaml# - properties: compatible: enum:
@@ -60,8 +57,8 @@ properties: items: - const: pcie - const: pcie_bus - - const: pcie_phy - - const: pcie_inbound_axi for imx6sx-pcie, pcie_aux for imx8mq-pcie + - enum: [ pcie_phy, pcie_aux ] + - enum: [ pcie_aux, pcie_inbound_axi ] num-lanes: const: 1
@@ -177,6 +174,72 @@ required: unevaluatedProperties: false +allOf: + - $ref: /schemas/pci/snps,dw-pcie.yaml# + + - if: + properties: + compatible: + contains: + enum: + - fsl,imx6q-pcie + - fsl,imx6qp-pcie + - fsl,imx7d-pcie + then: + properties: + clocks: + maxItems: 3 + clock-names: + items: + - const: pcie + - const: pcie_bus + - const: pcie_phy + + - if: + properties: + compatible: + contains: + const: fsl,imx6sx-pcie + then: + properties: + clock-names: + items: + - const: pcie + - const: pcie_bus + - const: pcie_phy + - const: pcie_inbound_axi + + - if: + properties: + compatible: + contains: + const: fsl,imx8mq-pcie + then: + properties: + clock-names: + items: + - const: pcie + - const: pcie_bus + - const: pcie_phy + - const: pcie_aux + + - if: + properties: + compatible: + contains: + enum: + - fsl,imx8mm-pcie + - fsl,imx8mp-pcie + then: + properties: + clocks: + maxItems: 3 + clock-names: + items: + - const: pcie + - const: pcie_bus + - const: pcie_aux + examples: - | #include <dt-bindings/clock/imx6qdl-clock.h>
Signature · 1 line
2.35.1
PATCH v4 2/3 dt-bindings: imx6q-pcie: Handle various PD configurations
The i.MX SoCs have various power domain configurations routed into the PCIe IP. MX6SX is the only one which contains 2 domains and also uses power-domain-names. MX6QDL do not use any domains. All the rest uses one domain and does not use power-domain-names anymore.
Document all those configurations in the DT binding document.
---
Cc: Fabio Estevam <[email protected]>
Cc: Krzysztof Kozlowski <[email protected]>
Cc: Lucas Stach <[email protected]>
Cc: Richard Zhu <[email protected]>
Cc: Rob Herring <[email protected]>
Cc: Shawn Guo <[email protected]>
Cc: [email protected]
Cc: NXP Linux Team <[email protected]>
To: [email protected]
---
V2: - Keep the power-domains description in the main section
V3: - Move power-domains back where they were originally (fixes V2)
- Do not use else: in allOf section
V4: - Special-case MX6Q and MX6QP in allOf section since they dont use PDs
- Drop minItems: from power-domains main section
- Handle anything which is not MX6SX,MX6Q,MX6QP as having one PD
--- .../bindings/pci/fsl,imx6q-pcie.yaml | 43 +++++++++++++++++++ 1 file changed, 43 insertions(+)
Documentation/devicetree/bindings/pci/fsl,imx6q-pcie.yaml +43 -0
@@ -69,6 +69,7 @@ properties: required properties for imx7d-pcie and imx8mq-pcie. power-domains: + minItems: 1 items: - description: The phandle pointing to the DISPLAY domain for imx6sx-pcie, to PCIE_PHY power domain for imx7d-pcie and
@@ -77,6 +78,7 @@ properties: for imx6sx-pcie. power-domain-names: + minItems: 1 items: - const: pcie - const: pcie_phy
@@ -240,6 +242,47 @@ allOf: - const: pcie_bus - const: pcie_aux + - if: + properties: + compatible: + contains: + const: fsl,imx6sx-pcie + then: + properties: + power-domains: + minItems: 2 + maxItems: 2 + power-domain-names: + minItems: 2 + maxItems: 2 + + - if: + properties: + compatible: + contains: + enum: + - fsl,imx6q-pcie + - fsl,imx6qp-pcie + then: + properties: + power-domains: false + power-domain-names: false + + - if: + not: + properties: + compatible: + contains: + enum: + - fsl,imx6sx-pcie + - fsl,imx6q-pcie + - fsl,imx6qp-pcie + then: + properties: + power-domains: + maxItems: 1 + power-domain-names: false + examples: - | #include <dt-bindings/clock/imx6qdl-clock.h>
Signature · 1 line
2.35.1
Re: [PATCH v4 2/3] dt-bindings: imx6q-pcie: Handle various PD configurations
Quoted 6 lines · Marek Vasut
The i.MX SoCs have various power domain configurations routed into the PCIe IP. MX6SX is the only one which contains 2 domains and also uses power-domain-names. MX6QDL do not use any domains. All the rest uses one domain and does not use power-domain-names anymore.
Document all those configurations in the DT binding document.
Best regards,
Krzysztof
PATCH v4 3/3 dt-bindings: imx6q-pcie: Handle more resets on legacy platforms
In reply to Marek VasutThe i.MX6 and i.MX7D does not use block controller to toggle PCIe reset, hence the PCIe DT description contains three reset entries on these older SoCs. Add this exception into the binding document.
---
Cc: Fabio Estevam <[email protected]>
Cc: Krzysztof Kozlowski <[email protected]>
Cc: Lucas Stach <[email protected]>
Cc: Richard Zhu <[email protected]>
Cc: Rob Herring <[email protected]>
Cc: Shawn Guo <[email protected]>
Cc: [email protected]
Cc: NXP Linux Team <[email protected]>
To: [email protected]
---
V2: - Add mx8mq to 3-reset PCIe core variant
- Handle the resets in allOf section
V3: - Reinstate reset: maxItems:3 and add minItems:2
- Move reset-names back to main section
- The validation no longer works and introduces errors like these:
arch/arm64/boot/dts/freescale/imx8mm-verdin-wifi-dahlia.dtb: pcie@33800000: reset-names:0: 'pciephy' was expected
V4: - Reinstate reset minItems and maxItems
- Turn the first two reset-names items into enums to cover all
the various name combinations, sort the rest in allOf section
--- .../bindings/pci/fsl,imx6q-pcie.yaml | 34 +++++++++++++++++-- 1 file changed, 32 insertions(+), 2 deletions(-)
Documentation/devicetree/bindings/pci/fsl,imx6q-pcie.yaml +32 -2
@@ -84,14 +84,16 @@ properties: - const: pcie_phy resets: + minItems: 2 maxItems: 3 description: Phandles to PCIe-related reset lines exposed by SRC IP block. Additional required by imx7d-pcie and imx8mq-pcie. reset-names: + minItems: 2 items: - - const: pciephy - - const: apps + - enum: [ pciephy, apps ] + - enum: [ apps, turnoff ] - const: turnoff fsl,tx-deemph-gen1:
@@ -283,6 +285,34 @@ allOf: maxItems: 1 power-domain-names: false + - if: + properties: + compatible: + contains: + enum: + - fsl,imx6q-pcie + - fsl,imx6sx-pcie + - fsl,imx6qp-pcie + - fsl,imx7d-pcie + - fsl,imx8mq-pcie + then: + properties: + resets: + minItems: 3 + reset-names: + items: + - const: pciephy + - const: apps + - const: turnoff + else: + properties: + resets: + maxItems: 2 + reset-names: + items: + - const: apps + - const: turnoff + examples: - | #include <dt-bindings/clock/imx6qdl-clock.h>
Signature · 1 line
2.35.1
Re: [PATCH v4 3/3] dt-bindings: imx6q-pcie: Handle more resets on legacy platforms
Quoted 40 lines · Marek Vasut
The i.MX6 and i.MX7D does not use block controller to toggle PCIe reset, hence the PCIe DT description contains three reset entries on these older SoCs. Add this exception into the binding document.
Signed-off-by Marek Vasut <[email protected]>---
Cc: Fabio Estevam <[email protected]>
Cc: Krzysztof Kozlowski <[email protected]>
Cc: Lucas Stach <[email protected]>
Cc: Richard Zhu <[email protected]>
Cc: Rob Herring <[email protected]>
Cc: Shawn Guo <[email protected]>
Cc: [email protected]
Cc: NXP Linux Team <[email protected]>
To: [email protected]
---
V2: - Add mx8mq to 3-reset PCIe core variant
- Handle the resets in allOf section
V3: - Reinstate reset: maxItems:3 and add minItems:2
- Move reset-names back to main section
- The validation no longer works and introduces errors like these:
arch/arm64/boot/dts/freescale/imx8mm-verdin-wifi-dahlia.dtb: pcie@33800000: reset-names:0: 'pciephy' was expected
V4: - Reinstate reset minItems and maxItems
- Turn the first two reset-names items into enums to cover all
the various name combinations, sort the rest in allOf section--- .../bindings/pci/fsl,imx6q-pcie.yaml | 34 +++++++++++++++++-- 1 file changed, 32 insertions(+), 2 deletions(-)Documentation/devicetree/bindings/pci/fsl,imx6q-pcie.yaml +1
-0@@ -84,14 +84,16 @@ properties: - const: pcie_phy resets: + minItems: 2 maxItems: 3 description: Phandles to PCIe-related reset lines exposed by SRC
Documentation/devicetree/bindings/pci/fsl,imx6q-pcie.yaml properties: · line 97
Quoted 10 lines
Documentation/devicetree/bindings/pci/fsl,imx6q-pcie.yaml +3
-2IP block. Additional required by imx7d-pcie and imx8mq-pcie. reset-names: + minItems: 2 items: - - const: pciephy - - const: apps + - enum: [ pciephy, apps ] + - enum: [ apps, turnoff ] - const: turnoff
I would expect to remove all these entries. I asked to keep reset-names with minIetms and maxItems. It works fine with your approach, but why having the items in multiple places?
Best regards,
Krzysztof
Re: [PATCH v4 3/3] dt-bindings: imx6q-pcie: Handle more resets on legacy platforms
Quoted 43 lines · Krzysztof Kozlowski
Quoted 42 lines · Marek Vasut
The i.MX6 and i.MX7D does not use block controller to toggle PCIe reset, hence the PCIe DT description contains three reset entries on these older SoCs. Add this exception into the binding document.
Signed-off-by Marek Vasut <[email protected]>---
Cc: Fabio Estevam <[email protected]>
Cc: Krzysztof Kozlowski <[email protected]>
Cc: Lucas Stach <[email protected]>
Cc: Richard Zhu <[email protected]>
Cc: Rob Herring <[email protected]>
Cc: Shawn Guo <[email protected]>
Cc: [email protected]
Cc: NXP Linux Team <[email protected]>
To: [email protected]
---
V2: - Add mx8mq to 3-reset PCIe core variant
- Handle the resets in allOf section
V3: - Reinstate reset: maxItems:3 and add minItems:2
- Move reset-names back to main section
- The validation no longer works and introduces errors like these:
arch/arm64/boot/dts/freescale/imx8mm-verdin-wifi-dahlia.dtb: pcie@33800000: reset-names:0: 'pciephy' was expected
V4: - Reinstate reset minItems and maxItems
- Turn the first two reset-names items into enums to cover all
the various name combinations, sort the rest in allOf section--- .../bindings/pci/fsl,imx6q-pcie.yaml | 34 +++++++++++++++++-- 1 file changed, 32 insertions(+), 2 deletions(-)Documentation/devicetree/bindings/pci/fsl,imx6q-pcie.yaml +1
-0@@ -84,14 +84,16 @@ properties: - const: pcie_phy resets: + minItems: 2 maxItems: 3 description: Phandles to PCIe-related reset lines exposed by SRC IP block. Additional required by imx7d-pcie and imx8mq-pcie.
Quoted 12 lines
Documentation/devicetree/bindings/pci/fsl,imx6q-pcie.yaml properties: · line 97
Quoted 8 lines
Documentation/devicetree/bindings/pci/fsl,imx6q-pcie.yaml +3
-2reset-names: + minItems: 2 items: - - const: pciephy - - const: apps + - enum: [ pciephy, apps ] + - enum: [ apps, turnoff ] - const: turnoffI would expect to remove all these entries. I asked to keep reset-names with minIetms and maxItems. It works fine with your approach, but why having the items in multiple places?
I feel like maybe I'm getting a bit lost in the complexity. I did that ^ and it does seem to work, so V5 is coming.