dt-bindings: Drop undocumented i.MX iomuxc-gpr bindings in examples
Subsystems (3)
Rob Herring
maintainerPATCH 1/1 dt-bindings: Drop undocumented i.MX iomuxc-gpr bindings in examples
The i.MX iomuxc-gpr bindings are undocumented and a mess. Drop their use from the examples.
The problem with the binding beyond the just random variations is that the iomuxc-gpr is not a separate block, but registers within the iomuxc block containing random leftover controls. As a separate DT node, it creates nodes with overlapping memory addresses.
--- Documentation/devicetree/bindings/mfd/syscon.yaml | 8 -------- Documentation/devicetree/bindings/mux/reg-mux.yaml | 1 - 2 files changed, 9 deletions(-)
Documentation/devicetree/bindings/mfd/syscon.yaml +0 -8
@@ -100,12 +100,4 @@ examples: compatible = "allwinner,sun8i-h3-system-controller", "syscon"; reg = <0x01c00000 0x1000>; }; - - - | - gpr: iomuxc-gpr@20e0000 { - compatible = "fsl,imx6q-iomuxc-gpr", "syscon"; - reg = <0x020e0000 0x38>; - hwlocks = <&hwlock1 1>; - }; - ...
Documentation/devicetree/bindings/mux/reg-mux.yaml +0 -1
@@ -96,7 +96,6 @@ examples: #include <dt-bindings/mux/mux.h> syscon@1000 { - compatible = "fsl,imx7d-iomuxc-gpr", "fsl,imx6q-iomuxc-gpr", "syscon", "simple-mfd"; reg = <0x1000 0x100>; mux2: mux-controller {
Signature · 1 line
2.32.0
Hi!
Quoted 36 lines · Rob Herring
The i.MX iomuxc-gpr bindings are undocumented and a mess. Drop their use from the examples.
The problem with the binding beyond the just random variations is that the iomuxc-gpr is not a separate block, but registers within the iomuxc block containing random leftover controls. As a separate DT node, it creates nodes with overlapping memory addresses.
Signed-off-by Rob Herring <[email protected]>--- Documentation/devicetree/bindings/mfd/syscon.yaml | 8 -------- Documentation/devicetree/bindings/mux/reg-mux.yaml | 1 - 2 files changed, 9 deletions(-)Documentation/devicetree/bindings/mfd/syscon.yaml +0
-8@@ -100,12 +100,4 @@ examples: compatible = "allwinner,sun8i-h3-system-controller", "syscon"; reg = <0x01c00000 0x1000>; }; - - - | - gpr: iomuxc-gpr@20e0000 { - compatible = "fsl,imx6q-iomuxc-gpr", "syscon"; - reg = <0x020e0000 0x38>; - hwlocks = <&hwlock1 1>; - }; -Documentation/devicetree/bindings/mux/reg-mux.yaml +0
-0
Documentation/devicetree/bindings/mux/reg-mux.yaml examples: · line 101
Quoted 7 lines
Documentation/devicetree/bindings/mux/reg-mux.yaml +0
-1@@ -96,7 +96,6 @@ examples: #include <dt-bindings/mux/mux.h> syscon@1000 { - compatible = "fsl,imx7d-iomuxc-gpr", "fsl,imx6q-iomuxc-gpr", "syscon", "simple-mfd"; reg = <0x1000 0x100>; mux2: mux-controller {
Hmm, the reg-mux file is "mine" and I have no strong objection since it doesn't really affect what the example is there for, but it does look a bit weird to simply remove the compatible, no? Is that ok?
Either way, fine by me as you're the expert.
Cheers,
Peter
Quoted 36 lines · Rob Herring
The i.MX iomuxc-gpr bindings are undocumented and a mess. Drop their use from the examples.
The problem with the binding beyond the just random variations is that the iomuxc-gpr is not a separate block, but registers within the iomuxc block containing random leftover controls. As a separate DT node, it creates nodes with overlapping memory addresses.
Signed-off-by Rob Herring <[email protected]>--- Documentation/devicetree/bindings/mfd/syscon.yaml | 8 -------- Documentation/devicetree/bindings/mux/reg-mux.yaml | 1 - 2 files changed, 9 deletions(-)Documentation/devicetree/bindings/mfd/syscon.yaml +0
-8@@ -100,12 +100,4 @@ examples: compatible = "allwinner,sun8i-h3-system-controller", "syscon"; reg = <0x01c00000 0x1000>; }; - - - | - gpr: iomuxc-gpr@20e0000 { - compatible = "fsl,imx6q-iomuxc-gpr", "syscon"; - reg = <0x020e0000 0x38>; - hwlocks = <&hwlock1 1>; - }; -Documentation/devicetree/bindings/mux/reg-mux.yaml +0
-0
Documentation/devicetree/bindings/mux/reg-mux.yaml examples: · line 99
Quoted 5 lines
Documentation/devicetree/bindings/mux/reg-mux.yaml +0
-1@@ -96,7 +96,6 @@ examples: #include <dt-bindings/mux/mux.h> syscon@1000 { - compatible = "fsl,imx7d-iomuxc-gpr", "fsl,imx6q-iomuxc-gpr", "syscon", "simple-mfd"; reg = <0x1000 0x100>;
Is leaving no compatible the correct solution here?
Do we have another (working) platform that we can use in its place?
Does it make sense to leave the "syscon" and "simple-mfd" entries?
Quoted 1 line
Documentation/devicetree/bindings/mux/reg-mux.yaml +0
-0mux2: mux-controller {
Signature · 4 lines
Lee Jones [李琼斯] Principal Technical Lead - Developer Services Linaro.org │ Open source software for Arm SoCs Follow Linaro: Facebook | Twitter | Blog
Rob Herring
maintainerQuoted 39 lines · Lee Jones
Quoted 37 lines · Rob Herring
The i.MX iomuxc-gpr bindings are undocumented and a mess. Drop their use from the examples.
The problem with the binding beyond the just random variations is that the iomuxc-gpr is not a separate block, but registers within the iomuxc block containing random leftover controls. As a separate DT node, it creates nodes with overlapping memory addresses.
Signed-off-by Rob Herring <[email protected]>--- Documentation/devicetree/bindings/mfd/syscon.yaml | 8 -------- Documentation/devicetree/bindings/mux/reg-mux.yaml | 1 - 2 files changed, 9 deletions(-)Documentation/devicetree/bindings/mfd/syscon.yaml +0
-8@@ -100,12 +100,4 @@ examples: compatible = "allwinner,sun8i-h3-system-controller", "syscon"; reg = <0x01c00000 0x1000>; }; - - - | - gpr: iomuxc-gpr@20e0000 { - compatible = "fsl,imx6q-iomuxc-gpr", "syscon"; - reg = <0x020e0000 0x38>; - hwlocks = <&hwlock1 1>; - }; -Documentation/devicetree/bindings/mux/reg-mux.yaml +0
-0@@ -96,7 +96,6 @@ examples:
Quoted 6 lines
Documentation/devicetree/bindings/mux/reg-mux.yaml examples: · line 99
Quoted 4 lines
Documentation/devicetree/bindings/mux/reg-mux.yaml +0
-1#include <dt-bindings/mux/mux.h> syscon@1000 { - compatible = "fsl,imx7d-iomuxc-gpr", "fsl,imx6q-iomuxc-gpr", "syscon", "simple-mfd"; reg = <0x1000 0x100>;Is leaving no compatible the correct solution here?
Documenting iomuxc-gpr is really the right one, but as I said it is a mess and I'm not touching that.
But compatible is not really important in terms of what the example shows.
Quoted 1 line
Do we have another (working) platform that we can use in its place?
Not one for video muxing that I'm aware of.
Quoted 1 line
Does it make sense to leave the "syscon" and "simple-mfd" entries?
No, because we don't allow those on their own.
Rob
Quoted 53 lines · Rob Herring
Quoted 39 lines · Lee Jones
Quoted 37 lines · Rob Herring
The i.MX iomuxc-gpr bindings are undocumented and a mess. Drop their use from the examples.
The problem with the binding beyond the just random variations is that the iomuxc-gpr is not a separate block, but registers within the iomuxc block containing random leftover controls. As a separate DT node, it creates nodes with overlapping memory addresses.
Signed-off-by Rob Herring <[email protected]>--- Documentation/devicetree/bindings/mfd/syscon.yaml | 8 -------- Documentation/devicetree/bindings/mux/reg-mux.yaml | 1 - 2 files changed, 9 deletions(-)Documentation/devicetree/bindings/mfd/syscon.yaml +0
-8@@ -100,12 +100,4 @@ examples: compatible = "allwinner,sun8i-h3-system-controller", "syscon"; reg = <0x01c00000 0x1000>; }; - - - | - gpr: iomuxc-gpr@20e0000 { - compatible = "fsl,imx6q-iomuxc-gpr", "syscon"; - reg = <0x020e0000 0x38>; - hwlocks = <&hwlock1 1>; - }; -Documentation/devicetree/bindings/mux/reg-mux.yaml +0
-0@@ -96,7 +96,6 @@ examples:Quoted 6 lines
Documentation/devicetree/bindings/mux/reg-mux.yaml examples: · line 99
Quoted 4 lines
Documentation/devicetree/bindings/mux/reg-mux.yaml +0
-1#include <dt-bindings/mux/mux.h> syscon@1000 { - compatible = "fsl,imx7d-iomuxc-gpr", "fsl,imx6q-iomuxc-gpr", "syscon", "simple-mfd"; reg = <0x1000 0x100>;Is leaving no compatible the correct solution here?
Documenting iomuxc-gpr is really the right one, but as I said it is a mess and I'm not touching that.
But compatible is not really important in terms of what the example shows.
Quoted 7 lines
Quoted 1 line
Do we have another (working) platform that we can use in its place?
Not one for video muxing that I'm aware of.
Quoted 1 line
Does it make sense to leave the "syscon" and "simple-mfd" entries?
No, because we don't allow those on their own.
Very well. I applied this for now.
... pending any fall-out from the Mux Maintainer. :)
Signature · 4 lines
Lee Jones [李琼斯] Principal Technical Lead - Developer Services Linaro.org │ Open source software for Arm SoCs Follow Linaro: Facebook | Twitter | Blog