The Marvell switches are lacking DT bindings.
I need proper schema checking to add LED support to the
Marvell switch. Just how it is, it can't go on like this.
Some Device Tree fixes are included in the series, these
remove the major and most annoying warnings fallout noise:
some warnings remain, and these are of more serious nature,
such as missing phy-mode. They can be applied individually,
or to the networking tree with the rest of the patches.
This series requires Rob Herrings series
"dt-bindings: net: Child node schema cleanups" to be applied
first.
Signed-off-by: Linus Walleij <[email protected]>
---
Changes in v4:
- Rebase the series on top of Rob's series
"dt-bindings: net: Child node schema cleanups" (or the hex numbered
ports will not work)
- Fix up a whitespacing error corrupting v3...
- Add a new patch making the generic DSA binding require ports or
ethernet-ports in the switch node.
- Drop any corrections of port@a in the patches.
- Drop oneOf in the compatible enum for mv88e6xxx
- Use ethernet-switch, ethernet-ports and ethernet-phy in the examples
- Transclude the dsa.yaml#/$defs/ethernet-ports define for ports
- Move the DTS and binding fixes first, before the actual bindings,
so they apply without (too many) warnings as fallout.
- Drop stray colon in text.
- Drop example port in the mveusb binding.
- Link to v3: https://lore.kernel.org/r/[email protected]
Changes in v3:
- Fix up a related mvusb example in a different binding that
the scripts were complaining about.
- Fix up the wording on internal vs external MDIO buses in the
mv88e6xxx binding document.
- Remove pointless label and put the right rev-mii into the
MV88E6060 schema.
- Link to v2: https://lore.kernel.org/r/[email protected]
Changes in v2:
- Break out a separate Marvell MV88E6060 binding file. I stand corrected.
- Drop the idea to rely on nodename mdio-external for the external
MDIO bus, keep the compatible, drop patch for the driver.
- Fix more Marvell DT mistakes.
- Fix NXP DT mistakes in a separate patch.
- Fix Marvell ARM64 mistakes in a separate patch.
- Link to v1: https://lore.kernel.org/r/[email protected]
---
Linus Walleij (7):
dt-bindings: net: dsa: Require ports or ethernet-ports
dt-bindings: net: mvusb: Fix up DSA example
ARM: dts: marvell: Fix some common switch mistakes
ARM: dts: nxp: Fix some common switch mistakes
ARM64: dts: marvell: Fix some common switch mistakes
dt-bindings: marvell: Rewrite MV88E6xxx in schema
dt-bindings: marvell: Add Marvell MV88E6060 DSA schema
Documentation/devicetree/bindings/net/dsa/dsa.yaml | 6 +
.../bindings/net/dsa/marvell,mv88e6060.yaml | 90 +++++++++
.../bindings/net/dsa/marvell,mv88e6xxx.yaml | 225 +++++++++++++++++++++
.../devicetree/bindings/net/dsa/marvell.txt | 109 ----------
.../devicetree/bindings/net/marvell,mvusb.yaml | 7 +-
MAINTAINERS | 3 +-
arch/arm/boot/dts/marvell/armada-370-rd.dts | 2 -
.../dts/marvell/armada-381-netgear-gs110emx.dts | 2 -
.../dts/marvell/armada-385-clearfog-gtr-l8.dts | 2 +-
.../dts/marvell/armada-385-clearfog-gtr-s4.dts | 2 +-
arch/arm/boot/dts/marvell/armada-385-linksys.dtsi | 2 -
.../boot/dts/marvell/armada-385-turris-omnia.dts | 16 +-
arch/arm/boot/dts/marvell/armada-388-clearfog.dts | 2 -
.../boot/dts/marvell/armada-xp-linksys-mamba.dts | 2 -
arch/arm/boot/dts/nxp/vf/vf610-zii-cfu1.dts | 2 +-
arch/arm/boot/dts/nxp/vf/vf610-zii-scu4-aib.dts | 8 +-
arch/arm/boot/dts/nxp/vf/vf610-zii-spb4.dts | 2 +-
arch/arm/boot/dts/nxp/vf/vf610-zii-ssmb-dtu.dts | 4 +-
arch/arm/boot/dts/nxp/vf/vf610-zii-ssmb-spu3.dts | 2 +-
.../boot/dts/marvell/armada-3720-espressobin.dtsi | 4 +-
.../boot/dts/marvell/armada-3720-gl-mv1000.dts | 4 +-
.../boot/dts/marvell/armada-3720-turris-mox.dts | 12 +-
.../boot/dts/marvell/armada-7040-mochabin.dts | 2 -
.../dts/marvell/armada-8040-clearfog-gt-8k.dts | 2 +-
arch/arm64/boot/dts/marvell/cn9130-crb.dtsi | 4 +-
25 files changed, 356 insertions(+), 160 deletions(-)
---
base-commit: 1c9be5fea84e409542a186893d219bf7cff22f5a
change-id: 20231008-marvell-88e6152-wan-led-88c43b7fd2fd
Best regards,
--
Linus Walleij <[email protected]>
When adding a proper schema for the Marvell mx88e6xxx switch,
the scripts start complaining about this embedded example:
dtschema/dtc warnings/errors:
net/marvell,mvusb.example.dtb: switch@0: ports: '#address-cells'
is a required property
from schema $id: http://devicetree.org/schemas/net/dsa/marvell,mv88e6xxx.yaml#
net/marvell,mvusb.example.dtb: switch@0: ports: '#size-cells'
is a required property
from schema $id: http://devicetree.org/schemas/net/dsa/marvell,mv88e6xxx.yaml#
Fix this up by extending the example with those properties in
the ports node.
While we are at it, rename "ports" to "ethernet-ports" and rename
"switch" to "ethernet-switch" as this is recommended practice.
Signed-off-by: Linus Walleij <[email protected]>
---
Documentation/devicetree/bindings/net/marvell,mvusb.yaml | 7 +++++--
1 file changed, 5 insertions(+), 2 deletions(-)
diff --git a/Documentation/devicetree/bindings/net/marvell,mvusb.yaml b/Documentation/devicetree/bindings/net/marvell,mvusb.yaml
index 3a3325168048..ab838c1ffeed 100644
--- a/Documentation/devicetree/bindings/net/marvell,mvusb.yaml
+++ b/Documentation/devicetree/bindings/net/marvell,mvusb.yaml
@@ -50,11 +50,14 @@ examples:
#address-cells = <1>;
#size-cells = <0>;
- switch@0 {
+ ethernet-switch@0 {
compatible = "marvell,mv88e6190";
reg = <0x0>;
- ports {
+ ethernet-ports {
+ #address-cells = <1>;
+ #size-cells = <0>;
+
/* Port definitions */
};
--
2.34.1
Fix some errors in the Marvell MV88E6xxx switch descriptions:
- The top node had no address size or cells.
- switch0@0 is not OK, should be switch@0.
Signed-off-by: Linus Walleij <[email protected]>
---
arch/arm64/boot/dts/marvell/armada-3720-espressobin.dtsi | 4 +---
arch/arm64/boot/dts/marvell/armada-3720-gl-mv1000.dts | 4 +---
arch/arm64/boot/dts/marvell/armada-3720-turris-mox.dts | 12 ++++++------
arch/arm64/boot/dts/marvell/armada-7040-mochabin.dts | 2 --
arch/arm64/boot/dts/marvell/armada-8040-clearfog-gt-8k.dts | 2 +-
arch/arm64/boot/dts/marvell/cn9130-crb.dtsi | 4 +---
6 files changed, 10 insertions(+), 18 deletions(-)
diff --git a/arch/arm64/boot/dts/marvell/armada-3720-espressobin.dtsi b/arch/arm64/boot/dts/marvell/armada-3720-espressobin.dtsi
index 5fc613d24151..b526efeee293 100644
--- a/arch/arm64/boot/dts/marvell/armada-3720-espressobin.dtsi
+++ b/arch/arm64/boot/dts/marvell/armada-3720-espressobin.dtsi
@@ -145,10 +145,8 @@ &usb2 {
};
&mdio {
- switch0: switch0@1 {
+ switch0: switch@1 {
compatible = "marvell,mv88e6085";
- #address-cells = <1>;
- #size-cells = <0>;
reg = <1>;
dsa,member = <0 0>;
diff --git a/arch/arm64/boot/dts/marvell/armada-3720-gl-mv1000.dts b/arch/arm64/boot/dts/marvell/armada-3720-gl-mv1000.dts
index b1b45b4fa9d4..5de4417f929c 100644
--- a/arch/arm64/boot/dts/marvell/armada-3720-gl-mv1000.dts
+++ b/arch/arm64/boot/dts/marvell/armada-3720-gl-mv1000.dts
@@ -152,10 +152,8 @@ &uart0 {
};
&mdio {
- switch0: switch0@1 {
+ switch0: switch@1 {
compatible = "marvell,mv88e6085";
- #address-cells = <1>;
- #size-cells = <0>;
reg = <1>;
dsa,member = <0 0>;
diff --git a/arch/arm64/boot/dts/marvell/armada-3720-turris-mox.dts b/arch/arm64/boot/dts/marvell/armada-3720-turris-mox.dts
index 9eab2bb22134..c69cb4e191e5 100644
--- a/arch/arm64/boot/dts/marvell/armada-3720-turris-mox.dts
+++ b/arch/arm64/boot/dts/marvell/armada-3720-turris-mox.dts
@@ -305,7 +305,7 @@ phy1: ethernet-phy@1 {
};
/* switch nodes are enabled by U-Boot if modules are present */
- switch0@10 {
+ switch@10 {
compatible = "marvell,mv88e6190";
reg = <0x10>;
dsa,member = <0 0>;
@@ -430,7 +430,7 @@ port-sfp@a {
};
};
- switch0@2 {
+ switch@2 {
compatible = "marvell,mv88e6085";
reg = <0x2>;
dsa,member = <0 0>;
@@ -497,7 +497,7 @@ port@5 {
};
};
- switch1@11 {
+ switch@11 {
compatible = "marvell,mv88e6190";
reg = <0x11>;
dsa,member = <0 1>;
@@ -622,7 +622,7 @@ port-sfp@a {
};
};
- switch1@2 {
+ switch@2 {
compatible = "marvell,mv88e6085";
reg = <0x2>;
dsa,member = <0 1>;
@@ -689,7 +689,7 @@ port@5 {
};
};
- switch2@12 {
+ switch@12 {
compatible = "marvell,mv88e6190";
reg = <0x12>;
dsa,member = <0 2>;
@@ -805,7 +805,7 @@ port-sfp@a {
};
};
- switch2@2 {
+ switch@2 {
compatible = "marvell,mv88e6085";
reg = <0x2>;
dsa,member = <0 2>;
diff --git a/arch/arm64/boot/dts/marvell/armada-7040-mochabin.dts b/arch/arm64/boot/dts/marvell/armada-7040-mochabin.dts
index 48202810bf78..3cc794fcf12e 100644
--- a/arch/arm64/boot/dts/marvell/armada-7040-mochabin.dts
+++ b/arch/arm64/boot/dts/marvell/armada-7040-mochabin.dts
@@ -303,8 +303,6 @@ eth2phy: ethernet-phy@1 {
/* 88E6141 Topaz switch */
switch: switch@3 {
compatible = "marvell,mv88e6085";
- #address-cells = <1>;
- #size-cells = <0>;
reg = <3>;
pinctrl-names = "default";
diff --git a/arch/arm64/boot/dts/marvell/armada-8040-clearfog-gt-8k.dts b/arch/arm64/boot/dts/marvell/armada-8040-clearfog-gt-8k.dts
index 4125202028c8..7a25ea36b565 100644
--- a/arch/arm64/boot/dts/marvell/armada-8040-clearfog-gt-8k.dts
+++ b/arch/arm64/boot/dts/marvell/armada-8040-clearfog-gt-8k.dts
@@ -497,7 +497,7 @@ ge_phy: ethernet-phy@0 {
reset-deassert-us = <10000>;
};
- switch0: switch0@4 {
+ switch0: switch@4 {
compatible = "marvell,mv88e6085";
reg = <4>;
pinctrl-names = "default";
diff --git a/arch/arm64/boot/dts/marvell/cn9130-crb.dtsi b/arch/arm64/boot/dts/marvell/cn9130-crb.dtsi
index 32cfb3e2efc3..110d4c9898bc 100644
--- a/arch/arm64/boot/dts/marvell/cn9130-crb.dtsi
+++ b/arch/arm64/boot/dts/marvell/cn9130-crb.dtsi
@@ -207,11 +207,9 @@ phy0: ethernet-phy@0 {
reg = <0>;
};
- switch6: switch0@6 {
+ switch6: switch@6 {
/* Actual device is MV88E6393X */
compatible = "marvell,mv88e6190";
- #address-cells = <1>;
- #size-cells = <0>;
reg = <6>;
interrupt-parent = <&cp0_gpio1>;
interrupts = <28 IRQ_TYPE_LEVEL_LOW>;
--
2.34.1
The Marvell MV88E6060 is one of the oldest DSA switches from
Marvell, and it has DT bindings used in the wild. Let's define
them properly.
It is different enough from the rest of the MV88E6xxx switches
that it deserves its own binding.
Signed-off-by: Linus Walleij <[email protected]>
---
.../bindings/net/dsa/marvell,mv88e6060.yaml | 90 ++++++++++++++++++++++
MAINTAINERS | 1 +
2 files changed, 91 insertions(+)
diff --git a/Documentation/devicetree/bindings/net/dsa/marvell,mv88e6060.yaml b/Documentation/devicetree/bindings/net/dsa/marvell,mv88e6060.yaml
new file mode 100644
index 000000000000..787f328551f6
--- /dev/null
+++ b/Documentation/devicetree/bindings/net/dsa/marvell,mv88e6060.yaml
@@ -0,0 +1,90 @@
+# SPDX-License-Identifier: (GPL-2.0-only OR BSD-2-Clause)
+%YAML 1.2
+---
+$id: http://devicetree.org/schemas/net/dsa/marvell,mv88e6060.yaml#
+$schema: http://devicetree.org/meta-schemas/core.yaml#
+
+title: Marvell MV88E6060 DSA switch
+
+maintainers:
+ - Andrew Lunn <[email protected]>
+
+description:
+ The Marvell MV88E6060 switch has been produced and sold by Marvell
+ since at least 2010. The switch has one pin ADDR4 that controls the
+ MDIO address of the switch to be 0x10 or 0x00, and on the MDIO bus
+ connected to the switch, the PHYs inside the switch appear as
+ independent devices on address 0x00-0x04 or 0x10-0x14, so in difference
+ from many other DSA switches this switch does not have an internal
+ MDIO bus for the PHY devices.
+
+properties:
+ compatible:
+ const: marvell,mv88e6060
+ description:
+ The MV88E6060 is the oldest Marvell DSA switch product, and
+ as such a bit limited in features compared to later hardware.
+
+ reg:
+ maxItems: 1
+
+ reset-gpios:
+ description:
+ GPIO to be used to reset the whole device
+ maxItems: 1
+
+$ref: dsa.yaml#
+
+allOf:
+ - $ref: dsa.yaml#/$defs/ethernet-ports
+
+required:
+ - compatible
+ - reg
+
+unevaluatedProperties: false
+
+examples:
+ - |
+ #include <dt-bindings/gpio/gpio.h>
+ #include <dt-bindings/interrupt-controller/irq.h>
+ mdio {
+ #address-cells = <1>;
+ #size-cells = <0>;
+
+ ethernet-switch@16 {
+ compatible = "marvell,mv88e6060";
+ reg = <16>;
+
+ ethernet-ports {
+ #address-cells = <1>;
+ #size-cells = <0>;
+
+ port@0 {
+ reg = <0>;
+ label = "lan1";
+ };
+ port@1 {
+ reg = <1>;
+ label = "lan2";
+ };
+ port@2 {
+ reg = <2>;
+ label = "lan3";
+ };
+ port@3 {
+ reg = <3>;
+ label = "lan4";
+ };
+ port@5 {
+ reg = <5>;
+ phy-mode = "rev-mii";
+ ethernet = <ðc>;
+ fixed-link {
+ speed = <100>;
+ full-duplex;
+ };
+ };
+ };
+ };
+ };
diff --git a/MAINTAINERS b/MAINTAINERS
index 1b4475254d27..4c933a2a56ad 100644
--- a/MAINTAINERS
+++ b/MAINTAINERS
@@ -12625,6 +12625,7 @@ MARVELL 88E6XXX ETHERNET SWITCH FABRIC DRIVER
M: Andrew Lunn <[email protected]>
L: [email protected]
S: Maintained
+F: Documentation/devicetree/bindings/net/dsa/marvell,mv88e6060.yaml
F: Documentation/devicetree/bindings/net/dsa/marvell,mv88e6xxx.yaml
F: Documentation/networking/devlink/mv88e6xxx.rst
F: drivers/net/dsa/mv88e6xxx/
--
2.34.1
Fix some errors in the Marvell MV88E6xxx switch descriptions:
- switch0@0 is not OK, should be switch@0
Signed-off-by: Linus Walleij <[email protected]>
---
arch/arm/boot/dts/nxp/vf/vf610-zii-cfu1.dts | 2 +-
arch/arm/boot/dts/nxp/vf/vf610-zii-scu4-aib.dts | 8 ++++----
arch/arm/boot/dts/nxp/vf/vf610-zii-spb4.dts | 2 +-
arch/arm/boot/dts/nxp/vf/vf610-zii-ssmb-dtu.dts | 4 ++--
arch/arm/boot/dts/nxp/vf/vf610-zii-ssmb-spu3.dts | 2 +-
5 files changed, 9 insertions(+), 9 deletions(-)
diff --git a/arch/arm/boot/dts/nxp/vf/vf610-zii-cfu1.dts b/arch/arm/boot/dts/nxp/vf/vf610-zii-cfu1.dts
index 1a19aec8957b..add47d8fb58a 100644
--- a/arch/arm/boot/dts/nxp/vf/vf610-zii-cfu1.dts
+++ b/arch/arm/boot/dts/nxp/vf/vf610-zii-cfu1.dts
@@ -162,7 +162,7 @@ mdio1: mdio {
suppress-preamble;
status = "okay";
- switch0: switch0@0 {
+ switch0: switch@0 {
compatible = "marvell,mv88e6085";
pinctrl-names = "default";
pinctrl-0 = <&pinctrl_switch>;
diff --git a/arch/arm/boot/dts/nxp/vf/vf610-zii-scu4-aib.dts b/arch/arm/boot/dts/nxp/vf/vf610-zii-scu4-aib.dts
index df1335492a19..50356bd87d04 100644
--- a/arch/arm/boot/dts/nxp/vf/vf610-zii-scu4-aib.dts
+++ b/arch/arm/boot/dts/nxp/vf/vf610-zii-scu4-aib.dts
@@ -47,7 +47,7 @@ mdio_mux_1: mdio@1 {
#address-cells = <1>;
#size-cells = <0>;
- switch0: switch0@0 {
+ switch0: switch@0 {
compatible = "marvell,mv88e6190";
reg = <0>;
dsa,member = <0 0>;
@@ -130,7 +130,7 @@ mdio_mux_2: mdio@2 {
#address-cells = <1>;
#size-cells = <0>;
- switch1: switch1@0 {
+ switch1: switch@0 {
compatible = "marvell,mv88e6190";
reg = <0>;
dsa,member = <0 1>;
@@ -188,7 +188,7 @@ mdio_mux_4: mdio@4 {
#address-cells = <1>;
#size-cells = <0>;
- switch2: switch2@0 {
+ switch2: switch@0 {
compatible = "marvell,mv88e6190";
reg = <0>;
dsa,member = <0 2>;
@@ -276,7 +276,7 @@ mdio_mux_8: mdio@8 {
#address-cells = <1>;
#size-cells = <0>;
- switch3: switch3@0 {
+ switch3: switch@0 {
compatible = "marvell,mv88e6190";
reg = <0>;
dsa,member = <0 3>;
diff --git a/arch/arm/boot/dts/nxp/vf/vf610-zii-spb4.dts b/arch/arm/boot/dts/nxp/vf/vf610-zii-spb4.dts
index 1461804ecaea..20e9e2dacbe6 100644
--- a/arch/arm/boot/dts/nxp/vf/vf610-zii-spb4.dts
+++ b/arch/arm/boot/dts/nxp/vf/vf610-zii-spb4.dts
@@ -123,7 +123,7 @@ mdio1: mdio {
suppress-preamble;
status = "okay";
- switch0: switch0@0 {
+ switch0: switch@0 {
compatible = "marvell,mv88e6190";
pinctrl-0 = <&pinctrl_gpio_switch0>;
pinctrl-names = "default";
diff --git a/arch/arm/boot/dts/nxp/vf/vf610-zii-ssmb-dtu.dts b/arch/arm/boot/dts/nxp/vf/vf610-zii-ssmb-dtu.dts
index 463c2452b9b7..aa53a60518c3 100644
--- a/arch/arm/boot/dts/nxp/vf/vf610-zii-ssmb-dtu.dts
+++ b/arch/arm/boot/dts/nxp/vf/vf610-zii-ssmb-dtu.dts
@@ -112,7 +112,7 @@ mdio1: mdio {
suppress-preamble;
status = "okay";
- switch0: switch0@0 {
+ switch0: switch@0 {
compatible = "marvell,mv88e6190";
pinctrl-0 = <&pinctrl_gpio_switch0>;
pinctrl-names = "default";
@@ -167,7 +167,7 @@ port@9 {
};
};
- mdio1 {
+ mdio-external {
compatible = "marvell,mv88e6xxx-mdio-external";
#address-cells = <1>;
#size-cells = <0>;
diff --git a/arch/arm/boot/dts/nxp/vf/vf610-zii-ssmb-spu3.dts b/arch/arm/boot/dts/nxp/vf/vf610-zii-ssmb-spu3.dts
index f5ae0d5de315..0b7063b74130 100644
--- a/arch/arm/boot/dts/nxp/vf/vf610-zii-ssmb-spu3.dts
+++ b/arch/arm/boot/dts/nxp/vf/vf610-zii-ssmb-spu3.dts
@@ -137,7 +137,7 @@ mdio1: mdio {
suppress-preamble;
status = "okay";
- switch0: switch0@0 {
+ switch0: switch@0 {
compatible = "marvell,mv88e6190";
pinctrl-0 = <&pinctrl_gpio_switch0>;
pinctrl-names = "default";
--
2.34.1
This is an attempt to rewrite the Marvell MV88E6xxx switch bindings
in YAML schema.
The current text binding says:
WARNING: This binding is currently unstable. Do not program it into a
FLASH never to be changed again. Once this binding is stable, this
warning will be removed.
Well that never happened before we switched to YAML markup,
we can't have it like this, what about fixing the mess?
Signed-off-by: Linus Walleij <[email protected]>
---
.../bindings/net/dsa/marvell,mv88e6xxx.yaml | 225 +++++++++++++++++++++
.../devicetree/bindings/net/dsa/marvell.txt | 109 ----------
MAINTAINERS | 2 +-
3 files changed, 226 insertions(+), 110 deletions(-)
diff --git a/Documentation/devicetree/bindings/net/dsa/marvell,mv88e6xxx.yaml b/Documentation/devicetree/bindings/net/dsa/marvell,mv88e6xxx.yaml
new file mode 100644
index 000000000000..8013ac411b15
--- /dev/null
+++ b/Documentation/devicetree/bindings/net/dsa/marvell,mv88e6xxx.yaml
@@ -0,0 +1,225 @@
+# SPDX-License-Identifier: (GPL-2.0-only OR BSD-2-Clause)
+%YAML 1.2
+---
+$id: http://devicetree.org/schemas/net/dsa/marvell,mv88e6xxx.yaml#
+$schema: http://devicetree.org/meta-schemas/core.yaml#
+
+title: Marvell MV88E6xxx DSA switch family
+
+maintainers:
+ - Andrew Lunn <[email protected]>
+
+description:
+ The Marvell MV88E6xxx switch series has been produced and sold
+ by Marvell since at least 2010. The switch has a few compatibles which
+ just indicate the base address of the switch, then operating systems
+ can investigate switch ID registers to find out which actual version
+ of the switch it is dealing with.
+
+properties:
+ compatible:
+ enum:
+ - marvell,mv88e6085
+ - marvell,mv88e6190
+ - marvell,mv88e6250
+ description: |
+ marvell,mv88e6085: This switch uses base address 0x10.
+ This switch and its siblings will be autodetected from
+ ID registers found in the switch, so only "marvell,mv88e6085" should be
+ specified. This includes the following list of MV88Exxxx switches:
+ 6085, 6095, 6097, 6123, 6131, 6141, 6161, 6165, 6171, 6172, 6175, 6176,
+ 6185, 6240, 6320, 6321, 6341, 6350, 6351, 6352
+ marvell,mv88e6190: This switch uses base address 0x00.
+ This switch and its siblings will be autodetected from
+ ID registers found in the switch, so only "marvell,mv88e6190" should be
+ specified. This includes the following list of MV88Exxxx switches:
+ 6190, 6190X, 6191, 6290, 6361, 6390, 6390X
+ marvell,mv88e6250: This switch uses base address 0x08 or 0x18.
+ This switch and its siblings will be autodetected from
+ ID registers found in the switch, so only "marvell,mv88e6250" should be
+ specified. This includes the following list of MV88Exxxx switches:
+ 6220, 6250
+
+ reg:
+ maxItems: 1
+
+ eeprom-length:
+ $ref: /schemas/types.yaml#/definitions/uint32
+ description: Set to the length of an EEPROM connected to the switch. Must be
+ set if the switch can not detect the presence and/or size of a connected
+ EEPROM, otherwise optional.
+
+ reset-gpios:
+ description:
+ GPIO to be used to reset the whole device
+ maxItems: 1
+
+ interrupts:
+ description: The switch provides an external interrupt line, but it is
+ not always used by target systems.
+ maxItems: 1
+
+ interrupt-controller:
+ description: The switch has an internal interrupt controller used by
+ the different sub-blocks.
+
+ '#interrupt-cells':
+ description: The internal interrupt controller only supports triggering
+ on active high level interrupts so the second cell must alway be set to
+ IRQ_TYPE_LEVEL_HIGH.
+ const: 2
+
+ mdio:
+ $ref: /schemas/net/mdio.yaml#
+ unevaluatedProperties: false
+ description: Marvell MV88E6xxx switches have an varying combination of
+ internal and external MDIO buses, in some cases a combined bus that
+ can be used both internally and externally. This node is for the
+ primary bus, used internally and sometimes also externally.
+
+ mdio-external:
+ $ref: /schemas/net/mdio.yaml#
+ unevaluatedProperties: false
+ description: Marvell MV88E6xxx switches that have a separate external
+ MDIO bus use this port to access external components on the MDIO bus.
+
+ properties:
+ compatible:
+ const: marvell,mv88e6xxx-mdio-external
+
+ required:
+ - compatible
+
+$ref: dsa.yaml#
+
+allOf:
+ - $ref: dsa.yaml#/$defs/ethernet-ports
+
+required:
+ - compatible
+ - reg
+
+unevaluatedProperties: false
+
+examples:
+ - |
+ #include <dt-bindings/gpio/gpio.h>
+ #include <dt-bindings/interrupt-controller/irq.h>
+ mdio {
+ #address-cells = <1>;
+ #size-cells = <0>;
+
+ switch0: ethernet-switch@0 {
+ compatible = "marvell,mv88e6085";
+ reg = <0>;
+ reset-gpios = <&gpio5 1 GPIO_ACTIVE_LOW>;
+ interrupts-extended = <&gpio0 27 IRQ_TYPE_LEVEL_LOW>;
+ interrupt-controller;
+ #interrupt-cells = <2>;
+
+ ethernet-ports {
+ #address-cells = <1>;
+ #size-cells = <0>;
+
+ port@0 {
+ reg = <0>;
+ label = "lan1";
+ };
+ port@1 {
+ reg = <1>;
+ label = "lan2";
+ };
+ port@2 {
+ reg = <2>;
+ label = "lan3";
+ };
+ port@3 {
+ reg = <3>;
+ label = "lan4";
+ };
+ port@4 {
+ reg = <4>;
+ label = "wan";
+ };
+
+ port@5 {
+ reg = <5>;
+ phy-mode = "sgmii";
+ ethernet = <ð2>;
+
+ fixed-link {
+ speed = <1000>;
+ full-duplex;
+ };
+ };
+ };
+
+ mdio {
+ #address-cells = <1>;
+ #size-cells = <0>;
+
+ switch0phy0: ethernet-phy@0 {
+ reg = <0>;
+ interrupts-extended = <&switch0 0 IRQ_TYPE_LEVEL_HIGH>;
+ };
+ };
+ };
+ };
+
+ - |
+ #include <dt-bindings/gpio/gpio.h>
+ #include <dt-bindings/interrupt-controller/irq.h>
+ mdio {
+ #address-cells = <1>;
+ #size-cells = <0>;
+
+ switch1: ethernet-switch@0 {
+ compatible = "marvell,mv88e6190";
+ reg = <0>;
+ reset-gpios = <&gpio5 1 GPIO_ACTIVE_LOW>;
+ interrupts-extended = <&gpio0 27 IRQ_TYPE_LEVEL_LOW>;
+ interrupt-controller;
+ #interrupt-cells = <2>;
+
+ ethernet-ports {
+ #address-cells = <1>;
+ #size-cells = <0>;
+
+ port@0 {
+ reg = <0>;
+ label = "lan1";
+ };
+ port@1 {
+ reg = <1>;
+ label = "lan2";
+ };
+ port@2 {
+ reg = <2>;
+ label = "lan3";
+ };
+ port@3 {
+ reg = <3>;
+ label = "lan4";
+ };
+ };
+
+ mdio {
+ #address-cells = <1>;
+ #size-cells = <0>;
+
+ switch1phy0: ethernet-phy@0 {
+ reg = <0>;
+ interrupts-extended = <&switch1 0 IRQ_TYPE_LEVEL_HIGH>;
+ };
+ };
+
+ mdio-external {
+ compatible = "marvell,mv88e6xxx-mdio-external";
+ #address-cells = <1>;
+ #size-cells = <0>;
+ switch1phy9: ethernet-phy@9 {
+ reg = <9>;
+ };
+ };
+ };
+ };
diff --git a/Documentation/devicetree/bindings/net/dsa/marvell.txt b/Documentation/devicetree/bindings/net/dsa/marvell.txt
deleted file mode 100644
index 6ec0c181b6db..000000000000
--- a/Documentation/devicetree/bindings/net/dsa/marvell.txt
+++ /dev/null
@@ -1,109 +0,0 @@
-Marvell DSA Switch Device Tree Bindings
----------------------------------------
-
-WARNING: This binding is currently unstable. Do not program it into a
-FLASH never to be changed again. Once this binding is stable, this
-warning will be removed.
-
-If you need a stable binding, use the old dsa.txt binding.
-
-Marvell Switches are MDIO devices. The following properties should be
-placed as a child node of an mdio device.
-
-The properties described here are those specific to Marvell devices.
-Additional required and optional properties can be found in dsa.txt.
-
-The compatibility string is used only to find an identification register,
-which is at a different MDIO base address in different switch families.
-- "marvell,mv88e6085" : Switch has base address 0x10. Use with models:
- 6085, 6095, 6097, 6123, 6131, 6141, 6161, 6165,
- 6171, 6172, 6175, 6176, 6185, 6240, 6320, 6321,
- 6341, 6350, 6351, 6352
-- "marvell,mv88e6190" : Switch has base address 0x00. Use with models:
- 6190, 6190X, 6191, 6290, 6361, 6390, 6390X
-- "marvell,mv88e6250" : Switch has base address 0x08 or 0x18. Use with model:
- 6220, 6250
-
-Required properties:
-- compatible : Should be one of "marvell,mv88e6085",
- "marvell,mv88e6190" or "marvell,mv88e6250" as
- indicated above
-- reg : Address on the MII bus for the switch.
-
-Optional properties:
-
-- reset-gpios : Should be a gpio specifier for a reset line
-- interrupts : Interrupt from the switch
-- interrupt-controller : Indicates the switch is itself an interrupt
- controller. This is used for the PHY interrupts.
-#interrupt-cells = <2> : Controller uses two cells, number and flag
-- eeprom-length : Set to the length of an EEPROM connected to the
- switch. Must be set if the switch can not detect
- the presence and/or size of a connected EEPROM,
- otherwise optional.
-- mdio : Container of PHY and devices on the switches MDIO
- bus.
-- mdio? : Container of PHYs and devices on the external MDIO
- bus. The node must contains a compatible string of
- "marvell,mv88e6xxx-mdio-external"
-
-Example:
-
- mdio {
- #address-cells = <1>;
- #size-cells = <0>;
- interrupt-parent = <&gpio0>;
- interrupts = <27 IRQ_TYPE_LEVEL_LOW>;
- interrupt-controller;
- #interrupt-cells = <2>;
-
- switch0: switch@0 {
- compatible = "marvell,mv88e6085";
- reg = <0>;
- reset-gpios = <&gpio5 1 GPIO_ACTIVE_LOW>;
-
- mdio {
- #address-cells = <1>;
- #size-cells = <0>;
- switch1phy0: switch1phy0@0 {
- reg = <0>;
- interrupt-parent = <&switch0>;
- interrupts = <0 IRQ_TYPE_LEVEL_HIGH>;
- };
- };
- };
- };
-
- mdio {
- #address-cells = <1>;
- #size-cells = <0>;
- interrupt-parent = <&gpio0>;
- interrupts = <27 IRQ_TYPE_LEVEL_LOW>;
- interrupt-controller;
- #interrupt-cells = <2>;
-
- switch0: switch@0 {
- compatible = "marvell,mv88e6190";
- reg = <0>;
- reset-gpios = <&gpio5 1 GPIO_ACTIVE_LOW>;
-
- mdio {
- #address-cells = <1>;
- #size-cells = <0>;
- switch1phy0: switch1phy0@0 {
- reg = <0>;
- interrupt-parent = <&switch0>;
- interrupts = <0 IRQ_TYPE_LEVEL_HIGH>;
- };
- };
-
- mdio1 {
- compatible = "marvell,mv88e6xxx-mdio-external";
- #address-cells = <1>;
- #size-cells = <0>;
- switch1phy9: switch1phy0@9 {
- reg = <9>;
- };
- };
- };
- };
diff --git a/MAINTAINERS b/MAINTAINERS
index 90f13281d297..1b4475254d27 100644
--- a/MAINTAINERS
+++ b/MAINTAINERS
@@ -12625,7 +12625,7 @@ MARVELL 88E6XXX ETHERNET SWITCH FABRIC DRIVER
M: Andrew Lunn <[email protected]>
L: [email protected]
S: Maintained
-F: Documentation/devicetree/bindings/net/dsa/marvell.txt
+F: Documentation/devicetree/bindings/net/dsa/marvell,mv88e6xxx.yaml
F: Documentation/networking/devlink/mv88e6xxx.rst
F: drivers/net/dsa/mv88e6xxx/
F: include/linux/dsa/mv88e6xxx.h
--
2.34.1
On Wed, 18 Oct 2023 11:03:45 +0200, Linus Walleij wrote:
> This is an attempt to rewrite the Marvell MV88E6xxx switch bindings
> in YAML schema.
>
> The current text binding says:
> WARNING: This binding is currently unstable. Do not program it into a
> FLASH never to be changed again. Once this binding is stable, this
> warning will be removed.
>
> Well that never happened before we switched to YAML markup,
> we can't have it like this, what about fixing the mess?
>
> Signed-off-by: Linus Walleij <[email protected]>
> ---
> .../bindings/net/dsa/marvell,mv88e6xxx.yaml | 225 +++++++++++++++++++++
> .../devicetree/bindings/net/dsa/marvell.txt | 109 ----------
> MAINTAINERS | 2 +-
> 3 files changed, 226 insertions(+), 110 deletions(-)
>
My bot found errors running 'make DT_CHECKER_FLAGS=-m dt_binding_check'
on your patch (DT_CHECKER_FLAGS is new in v5.13):
yamllint warnings/errors:
dtschema/dtc warnings/errors:
/builds/robherring/dt-review-ci/linux/Documentation/devicetree/bindings/net/marvell,mvusb.example.dtb: switch@0: ports: '#address-cells' is a required property
from schema $id: http://devicetree.org/schemas/net/dsa/marvell,mv88e6xxx.yaml#
/builds/robherring/dt-review-ci/linux/Documentation/devicetree/bindings/net/marvell,mvusb.example.dtb: switch@0: ports: '#size-cells' is a required property
from schema $id: http://devicetree.org/schemas/net/dsa/marvell,mv88e6xxx.yaml#
/builds/robherring/dt-review-ci/linux/Documentation/devicetree/bindings/net/marvell,mvusb.example.dtb: switch@0: ports: '#address-cells' is a required property
from schema $id: http://devicetree.org/schemas/net/dsa/marvell,mv88e6xxx.yaml#
/builds/robherring/dt-review-ci/linux/Documentation/devicetree/bindings/net/marvell,mvusb.example.dtb: switch@0: ports: '#size-cells' is a required property
from schema $id: http://devicetree.org/schemas/net/dsa/marvell,mv88e6xxx.yaml#
doc reference errors (make refcheckdocs):
See https://patchwork.ozlabs.org/project/devicetree-bindings/patch/[email protected]
The base for the series is generally the latest rc1. A different dependency
should be noted in *this* patch.
If you already ran 'make dt_binding_check' and didn't see the above
error(s), then make sure 'yamllint' is installed and dt-schema is up to
date:
pip3 install dtschema --upgrade
Please check and re-submit after running the above command yourself. Note
that DT_SCHEMA_FILES can be set to your schema file to speed up checking
your schema. However, it must be unset to test all examples with your schema.
On Wed, 18 Oct 2023 11:03:41 +0200, Linus Walleij wrote:
> When adding a proper schema for the Marvell mx88e6xxx switch,
> the scripts start complaining about this embedded example:
>
> dtschema/dtc warnings/errors:
> net/marvell,mvusb.example.dtb: switch@0: ports: '#address-cells'
> is a required property
> from schema $id: http://devicetree.org/schemas/net/dsa/marvell,mv88e6xxx.yaml#
> net/marvell,mvusb.example.dtb: switch@0: ports: '#size-cells'
> is a required property
> from schema $id: http://devicetree.org/schemas/net/dsa/marvell,mv88e6xxx.yaml#
>
> Fix this up by extending the example with those properties in
> the ports node.
>
> While we are at it, rename "ports" to "ethernet-ports" and rename
> "switch" to "ethernet-switch" as this is recommended practice.
>
> Signed-off-by: Linus Walleij <[email protected]>
> ---
> Documentation/devicetree/bindings/net/marvell,mvusb.yaml | 7 +++++--
> 1 file changed, 5 insertions(+), 2 deletions(-)
>
My bot found errors running 'make DT_CHECKER_FLAGS=-m dt_binding_check'
on your patch (DT_CHECKER_FLAGS is new in v5.13):
yamllint warnings/errors:
dtschema/dtc warnings/errors:
Documentation/devicetree/bindings/net/marvell,mvusb.example.dtb: /example-0/usb/mdio@1/ethernet-switch@0: failed to match any schema with compatible: ['marvell,mv88e6190']
doc reference errors (make refcheckdocs):
See https://patchwork.ozlabs.org/project/devicetree-bindings/patch/[email protected]
The base for the series is generally the latest rc1. A different dependency
should be noted in *this* patch.
If you already ran 'make dt_binding_check' and didn't see the above
error(s), then make sure 'yamllint' is installed and dt-schema is up to
date:
pip3 install dtschema --upgrade
Please check and re-submit after running the above command yourself. Note
that DT_SCHEMA_FILES can be set to your schema file to speed up checking
your schema. However, it must be unset to test all examples with your schema.
On Wed, Oct 18, 2023 at 12:32 PM Rob Herring <[email protected]> wrote:
> dtschema/dtc warnings/errors:
> Documentation/devicetree/bindings/net/marvell,mvusb.example.dtb: /example-0/usb/mdio@1/ethernet-switch@0: failed to match any schema with compatible: ['marvell,mv88e6190']
Isn't that just because the bindings now come last in the series.
Which is in response to a review comment, hence this warning
didn't appear before.
Yours,
Linus Walleij
On Wed, Oct 18, 2023 at 12:32 PM Rob Herring <[email protected]> wrote:
> yamllint warnings/errors:
>
> dtschema/dtc warnings/errors:
> /builds/robherring/dt-review-ci/linux/Documentation/devicetree/bindings/net/marvell,mvusb.example.dtb: switch@0: ports: '#address-cells' is a required property
> from schema $id: http://devicetree.org/schemas/net/dsa/marvell,mv88e6xxx.yaml#
> /builds/robherring/dt-review-ci/linux/Documentation/devicetree/bindings/net/marvell,mvusb.example.dtb: switch@0: ports: '#size-cells' is a required property
> from schema $id: http://devicetree.org/schemas/net/dsa/marvell,mv88e6xxx.yaml#
> /builds/robherring/dt-review-ci/linux/Documentation/devicetree/bindings/net/marvell,mvusb.example.dtb: switch@0: ports: '#address-cells' is a required property
> from schema $id: http://devicetree.org/schemas/net/dsa/marvell,mv88e6xxx.yaml#
> /builds/robherring/dt-review-ci/linux/Documentation/devicetree/bindings/net/marvell,mvusb.example.dtb: switch@0: ports: '#size-cells' is a required property
> from schema $id: http://devicetree.org/schemas/net/dsa/marvell,mv88e6xxx.yaml#
Fixed in patch 2/7?
Yours,
Linus Walleij
On Wed, Oct 18, 2023 at 11:03:41AM +0200, Linus Walleij wrote:
> When adding a proper schema for the Marvell mx88e6xxx switch,
> the scripts start complaining about this embedded example:
>
> dtschema/dtc warnings/errors:
> net/marvell,mvusb.example.dtb: switch@0: ports: '#address-cells'
> is a required property
> from schema $id: http://devicetree.org/schemas/net/dsa/marvell,mv88e6xxx.yaml#
> net/marvell,mvusb.example.dtb: switch@0: ports: '#size-cells'
> is a required property
> from schema $id: http://devicetree.org/schemas/net/dsa/marvell,mv88e6xxx.yaml#
>
> Fix this up by extending the example with those properties in
> the ports node.
>
> While we are at it, rename "ports" to "ethernet-ports" and rename
> "switch" to "ethernet-switch" as this is recommended practice.
>
> Signed-off-by: Linus Walleij <[email protected]>
Reviewed-by: Andrew Lunn <[email protected]>
Andrew
On Wed, Oct 18, 2023 at 11:03:43AM +0200, Linus Walleij wrote:
> Fix some errors in the Marvell MV88E6xxx switch descriptions:
> - switch0@0 is not OK, should be switch@0
>
> Signed-off-by: Linus Walleij <[email protected]>
Reviewed-by: Andrew Lunn <[email protected]>
Andrew
On Wed, Oct 18, 2023 at 11:03:44AM +0200, Linus Walleij wrote:
> Fix some errors in the Marvell MV88E6xxx switch descriptions:
> - The top node had no address size or cells.
> - switch0@0 is not OK, should be switch@0.
>
> Signed-off-by: Linus Walleij <[email protected]>
Reviewed-by: Andrew Lunn <[email protected]>
Andrew
On Wed, Oct 18, 2023 at 11:03:45AM +0200, Linus Walleij wrote:
> This is an attempt to rewrite the Marvell MV88E6xxx switch bindings
> in YAML schema.
>
> The current text binding says:
> WARNING: This binding is currently unstable. Do not program it into a
> FLASH never to be changed again. Once this binding is stable, this
> warning will be removed.
>
> Well that never happened before we switched to YAML markup,
> we can't have it like this, what about fixing the mess?
>
> Signed-off-by: Linus Walleij <[email protected]>
For the text describing the properties:
Reviewed-by: Andrew Lunn <[email protected]>
Andrew
On Wed, Oct 18, 2023 at 11:03:46AM +0200, Linus Walleij wrote:
> The Marvell MV88E6060 is one of the oldest DSA switches from
> Marvell, and it has DT bindings used in the wild. Let's define
> them properly.
>
> It is different enough from the rest of the MV88E6xxx switches
> that it deserves its own binding.
>
> Signed-off-by: Linus Walleij <[email protected]>
Reviewed-by: Andrew Lunn <[email protected]>
Andrew
On Wed, Oct 18, 2023 at 01:37:10PM +0200, Linus Walleij wrote:
> On Wed, Oct 18, 2023 at 12:32 PM Rob Herring <[email protected]> wrote:
>
> > dtschema/dtc warnings/errors:
> > Documentation/devicetree/bindings/net/marvell,mvusb.example.dtb: /example-0/usb/mdio@1/ethernet-switch@0: failed to match any schema with compatible: ['marvell,mv88e6190']
>
> Isn't that just because the bindings now come last in the series.
> Which is in response to a review comment, hence this warning
> didn't appear before.
Yes. The only option that avoids this is squashing the 2 patches. I
think it is fine to leave this as-is.
Rob
On Wed, Oct 18, 2023 at 01:39:45PM +0200, Linus Walleij wrote:
> On Wed, Oct 18, 2023 at 12:32 PM Rob Herring <[email protected]> wrote:
>
> > yamllint warnings/errors:
> >
> > dtschema/dtc warnings/errors:
> > /builds/robherring/dt-review-ci/linux/Documentation/devicetree/bindings/net/marvell,mvusb.example.dtb: switch@0: ports: '#address-cells' is a required property
> > from schema $id: http://devicetree.org/schemas/net/dsa/marvell,mv88e6xxx.yaml#
> > /builds/robherring/dt-review-ci/linux/Documentation/devicetree/bindings/net/marvell,mvusb.example.dtb: switch@0: ports: '#size-cells' is a required property
> > from schema $id: http://devicetree.org/schemas/net/dsa/marvell,mv88e6xxx.yaml#
> > /builds/robherring/dt-review-ci/linux/Documentation/devicetree/bindings/net/marvell,mvusb.example.dtb: switch@0: ports: '#address-cells' is a required property
> > from schema $id: http://devicetree.org/schemas/net/dsa/marvell,mv88e6xxx.yaml#
> > /builds/robherring/dt-review-ci/linux/Documentation/devicetree/bindings/net/marvell,mvusb.example.dtb: switch@0: ports: '#size-cells' is a required property
> > from schema $id: http://devicetree.org/schemas/net/dsa/marvell,mv88e6xxx.yaml#
>
> Fixed in patch 2/7?
Yes. If one patch has errors we drop it. I should probably just give up
on the rest of the series instead.
Rob
On Wed, Oct 18, 2023 at 11:03:45AM +0200, Linus Walleij wrote:
> This is an attempt to rewrite the Marvell MV88E6xxx switch bindings
> in YAML schema.
>
> The current text binding says:
> WARNING: This binding is currently unstable. Do not program it into a
> FLASH never to be changed again. Once this binding is stable, this
> warning will be removed.
>
> Well that never happened before we switched to YAML markup,
> we can't have it like this, what about fixing the mess?
>
> Signed-off-by: Linus Walleij <[email protected]>
> ---
> .../bindings/net/dsa/marvell,mv88e6xxx.yaml | 225 +++++++++++++++++++++
> .../devicetree/bindings/net/dsa/marvell.txt | 109 ----------
> MAINTAINERS | 2 +-
> 3 files changed, 226 insertions(+), 110 deletions(-)
>
> diff --git a/Documentation/devicetree/bindings/net/dsa/marvell,mv88e6xxx.yaml b/Documentation/devicetree/bindings/net/dsa/marvell,mv88e6xxx.yaml
> new file mode 100644
> index 000000000000..8013ac411b15
> --- /dev/null
> +++ b/Documentation/devicetree/bindings/net/dsa/marvell,mv88e6xxx.yaml
> @@ -0,0 +1,225 @@
> +# SPDX-License-Identifier: (GPL-2.0-only OR BSD-2-Clause)
> +%YAML 1.2
> +---
> +$id: http://devicetree.org/schemas/net/dsa/marvell,mv88e6xxx.yaml#
> +$schema: http://devicetree.org/meta-schemas/core.yaml#
> +
> +title: Marvell MV88E6xxx DSA switch family
> +
> +maintainers:
> + - Andrew Lunn <[email protected]>
> +
> +description:
> + The Marvell MV88E6xxx switch series has been produced and sold
> + by Marvell since at least 2010. The switch has a few compatibles which
> + just indicate the base address of the switch, then operating systems
> + can investigate switch ID registers to find out which actual version
> + of the switch it is dealing with.
> +
> +properties:
> + compatible:
> + enum:
> + - marvell,mv88e6085
> + - marvell,mv88e6190
> + - marvell,mv88e6250
> + description: |
> + marvell,mv88e6085: This switch uses base address 0x10.
> + This switch and its siblings will be autodetected from
> + ID registers found in the switch, so only "marvell,mv88e6085" should be
> + specified. This includes the following list of MV88Exxxx switches:
> + 6085, 6095, 6097, 6123, 6131, 6141, 6161, 6165, 6171, 6172, 6175, 6176,
> + 6185, 6240, 6320, 6321, 6341, 6350, 6351, 6352
> + marvell,mv88e6190: This switch uses base address 0x00.
> + This switch and its siblings will be autodetected from
> + ID registers found in the switch, so only "marvell,mv88e6190" should be
> + specified. This includes the following list of MV88Exxxx switches:
> + 6190, 6190X, 6191, 6290, 6361, 6390, 6390X
> + marvell,mv88e6250: This switch uses base address 0x08 or 0x18.
> + This switch and its siblings will be autodetected from
> + ID registers found in the switch, so only "marvell,mv88e6250" should be
> + specified. This includes the following list of MV88Exxxx switches:
> + 6220, 6250
> +
> + reg:
> + maxItems: 1
> +
> + eeprom-length:
> + $ref: /schemas/types.yaml#/definitions/uint32
> + description: Set to the length of an EEPROM connected to the switch. Must be
> + set if the switch can not detect the presence and/or size of a connected
> + EEPROM, otherwise optional.
> +
> + reset-gpios:
> + description:
> + GPIO to be used to reset the whole device
> + maxItems: 1
> +
> + interrupts:
> + description: The switch provides an external interrupt line, but it is
> + not always used by target systems.
> + maxItems: 1
> +
> + interrupt-controller:
> + description: The switch has an internal interrupt controller used by
> + the different sub-blocks.
> +
> + '#interrupt-cells':
> + description: The internal interrupt controller only supports triggering
> + on active high level interrupts so the second cell must alway be set to
> + IRQ_TYPE_LEVEL_HIGH.
> + const: 2
> +
> + mdio:
> + $ref: /schemas/net/mdio.yaml#
> + unevaluatedProperties: false
> + description: Marvell MV88E6xxx switches have an varying combination of
> + internal and external MDIO buses, in some cases a combined bus that
> + can be used both internally and externally. This node is for the
> + primary bus, used internally and sometimes also externally.
> +
> + mdio-external:
> + $ref: /schemas/net/mdio.yaml#
> + unevaluatedProperties: false
> + description: Marvell MV88E6xxx switches that have a separate external
> + MDIO bus use this port to access external components on the MDIO bus.
> +
> + properties:
> + compatible:
> + const: marvell,mv88e6xxx-mdio-external
> +
> + required:
> + - compatible
> +
> +$ref: dsa.yaml#
Drop this. Covered by the line below.
> +
> +allOf:
> + - $ref: dsa.yaml#/$defs/ethernet-ports
On Thu, Oct 19, 2023 at 08:45:14AM -0500, Rob Herring wrote:
> On Wed, Oct 18, 2023 at 01:37:10PM +0200, Linus Walleij wrote:
> > On Wed, Oct 18, 2023 at 12:32 PM Rob Herring <[email protected]> wrote:
> >
> > > dtschema/dtc warnings/errors:
> > > Documentation/devicetree/bindings/net/marvell,mvusb.example.dtb: /example-0/usb/mdio@1/ethernet-switch@0: failed to match any schema with compatible: ['marvell,mv88e6190']
> >
> > Isn't that just because the bindings now come last in the series.
> > Which is in response to a review comment, hence this warning
> > didn't appear before.
>
> Yes. The only option that avoids this is squashing the 2 patches. I
> think it is fine to leave this as-is.
>
> Rob
Anyway, I'm surprised that the bot would send this email, since the
warning existed prior to this patch, and I would expect that the bot
only notifies of newly introduced issues.
On Wed, Oct 18, 2023 at 11:03:41AM +0200, Linus Walleij wrote:
> When adding a proper schema for the Marvell mx88e6xxx switch,
> the scripts start complaining about this embedded example:
>
> dtschema/dtc warnings/errors:
> net/marvell,mvusb.example.dtb: switch@0: ports: '#address-cells'
> is a required property
> from schema $id: http://devicetree.org/schemas/net/dsa/marvell,mv88e6xxx.yaml#
> net/marvell,mvusb.example.dtb: switch@0: ports: '#size-cells'
> is a required property
> from schema $id: http://devicetree.org/schemas/net/dsa/marvell,mv88e6xxx.yaml#
>
> Fix this up by extending the example with those properties in
> the ports node.
>
> While we are at it, rename "ports" to "ethernet-ports" and rename
> "switch" to "ethernet-switch" as this is recommended practice.
>
> Signed-off-by: Linus Walleij <[email protected]>
> ---
Reviewed-by: Vladimir Oltean <[email protected]>
On Wed, Oct 18, 2023 at 11:03:43AM +0200, Linus Walleij wrote:
> Fix some errors in the Marvell MV88E6xxx switch descriptions:
> - switch0@0 is not OK, should be switch@0
>
> Signed-off-by: Linus Walleij <[email protected]>
> ---
Same comment as on patch 3, prefer ethernet-switch and ethernet-ports.
+Marek
On Wed, Oct 18, 2023 at 11:03:44AM +0200, Linus Walleij wrote:
> Fix some errors in the Marvell MV88E6xxx switch descriptions:
> - The top node had no address size or cells.
> - switch0@0 is not OK, should be switch@0.
>
> Signed-off-by: Linus Walleij <[email protected]>
> ---
> diff --git a/arch/arm64/boot/dts/marvell/armada-3720-turris-mox.dts b/arch/arm64/boot/dts/marvell/armada-3720-turris-mox.dts
> index 9eab2bb22134..c69cb4e191e5 100644
> --- a/arch/arm64/boot/dts/marvell/armada-3720-turris-mox.dts
> +++ b/arch/arm64/boot/dts/marvell/armada-3720-turris-mox.dts
> @@ -305,7 +305,7 @@ phy1: ethernet-phy@1 {
> };
>
> /* switch nodes are enabled by U-Boot if modules are present */
> - switch0@10 {
> + switch@10 {
As the comment says: U-Boot
(https://elixir.bootlin.com/u-boot/latest/source/board/CZ.NIC/turris_mox/turris_mox.c#L728)
sets up status = "okay" for these nodes depending on the MOXTET
configuration. It doesn't look as if it's doing that by alias, just by
path ("%s/switch%i@%x").
I have a Turris MOX, please allow me some time to test if the node name
change is going to be significant and cause regressions. I expect the
answer to be yes (sadly).
> compatible = "marvell,mv88e6190";
> reg = <0x10>;
> dsa,member = <0 0>;
> @@ -430,7 +430,7 @@ port-sfp@a {
> };
> };
>
> - switch0@2 {
> + switch@2 {
> compatible = "marvell,mv88e6085";
> reg = <0x2>;
> dsa,member = <0 0>;
> @@ -497,7 +497,7 @@ port@5 {
> };
> };
>
> - switch1@11 {
> + switch@11 {
> compatible = "marvell,mv88e6190";
> reg = <0x11>;
> dsa,member = <0 1>;
> @@ -622,7 +622,7 @@ port-sfp@a {
> };
> };
>
> - switch1@2 {
> + switch@2 {
> compatible = "marvell,mv88e6085";
> reg = <0x2>;
> dsa,member = <0 1>;
> @@ -689,7 +689,7 @@ port@5 {
> };
> };
>
> - switch2@12 {
> + switch@12 {
> compatible = "marvell,mv88e6190";
> reg = <0x12>;
> dsa,member = <0 2>;
> @@ -805,7 +805,7 @@ port-sfp@a {
> };
> };
>
> - switch2@2 {
> + switch@2 {
> compatible = "marvell,mv88e6085";
> reg = <0x2>;
> dsa,member = <0 2>;
On Thu, Oct 19, 2023 at 05:40:22PM +0300, Vladimir Oltean wrote:
> +Marek
>
> On Wed, Oct 18, 2023 at 11:03:44AM +0200, Linus Walleij wrote:
> > Fix some errors in the Marvell MV88E6xxx switch descriptions:
> > - The top node had no address size or cells.
> > - switch0@0 is not OK, should be switch@0.
> >
> > Signed-off-by: Linus Walleij <[email protected]>
> > ---
> > diff --git a/arch/arm64/boot/dts/marvell/armada-3720-turris-mox.dts b/arch/arm64/boot/dts/marvell/armada-3720-turris-mox.dts
> > index 9eab2bb22134..c69cb4e191e5 100644
> > --- a/arch/arm64/boot/dts/marvell/armada-3720-turris-mox.dts
> > +++ b/arch/arm64/boot/dts/marvell/armada-3720-turris-mox.dts
> > @@ -305,7 +305,7 @@ phy1: ethernet-phy@1 {
> > };
> >
> > /* switch nodes are enabled by U-Boot if modules are present */
> > - switch0@10 {
> > + switch@10 {
>
> As the comment says: U-Boot
> (https://elixir.bootlin.com/u-boot/latest/source/board/CZ.NIC/turris_mox/turris_mox.c#L728)
> sets up status = "okay" for these nodes depending on the MOXTET
> configuration. It doesn't look as if it's doing that by alias, just by
> path ("%s/switch%i@%x").
>
> I have a Turris MOX, please allow me some time to test if the node name
> change is going to be significant and cause regressions. I expect the
> answer to be yes (sadly).
Yeah, it's bad.
U-Boot 2018.11 (Dec 16 2018 - 12:50:19 +0000), Build: jenkins-turris-os-packages-kittens-mox-90
DRAM: 1 GiB
Enabling Armada 3720 wComphy-0: SGMII1 3.125 Gbps
Comphy-1: PEX0 5 Gbps
Comphy-2: USB3_HOST0 5 Gbps
MMC: sdhci@d8000: 0
Loading Environment from SPI Flash... SF: Detected w25q64dw with page size 256 Bytes, erase size 4 KiB, total 8 MiB
OK
Model: CZ.NIC Turris Mox Board
Net: eth0: neta@30000
Turris Mox:
Board version: 22
RAM size: 1024 MiB
SD/eMMC version: SD
Module Topology:
1: Peridot Switch Module (8-port)
2: Peridot Switch Module (8-port)
3: Peridot Switch Module (8-port)
4: SFP Module
Hit any key to stop autoboot: 0
=> run sd_tftp_boot
neta@30000 Waiting for PHY auto negotiation to complete....... done
BOOTP broadcast 1
BOOTP broadcast 2
DHCP client bound to address 10.0.0.117 (254 ms)
Using neta@30000 device
TFTP from server 10.0.0.1; our IP address is 10.0.0.117
Filename 'mox/armada-3720-turris-mox.dtb'.
Load address: 0x4f00000
Loading: ####
1.5 MiB/s
done
Bytes transferred = 19479 (4c17 hex)
Using neta@30000 device
TFTP from server 10.0.0.1; our IP address is 10.0.0.117
Filename 'mox/Image'.
Load address: 0x5000000
Loading: #################################################################
##########################################
6 MiB/s
done
Bytes transferred = 54069760 (3390a00 hex)
## Flattened Device Tree blob at 04f00000
Booting using the fdt blob at 0x4f00000
Loading Device Tree to 000000003bf16000, end 000000003bf1dc16 ... OK
ERROR: board-specific fdt fixup failed: FDT_ERR_NOTFOUND
- must RESET the board to recover.
FDT creation failed! hanging...### ERROR ### Please RESET the board ###
On Thu, 19 Oct 2023 17:49:35 +0300
Vladimir Oltean <[email protected]> wrote:
> On Thu, Oct 19, 2023 at 05:40:22PM +0300, Vladimir Oltean wrote:
> > +Marek
> >
> > On Wed, Oct 18, 2023 at 11:03:44AM +0200, Linus Walleij wrote:
> > > Fix some errors in the Marvell MV88E6xxx switch descriptions:
> > > - The top node had no address size or cells.
> > > - switch0@0 is not OK, should be switch@0.
> > >
> > > Signed-off-by: Linus Walleij <[email protected]>
> > > ---
> > > diff --git a/arch/arm64/boot/dts/marvell/armada-3720-turris-mox.dts b/arch/arm64/boot/dts/marvell/armada-3720-turris-mox.dts
> > > index 9eab2bb22134..c69cb4e191e5 100644
> > > --- a/arch/arm64/boot/dts/marvell/armada-3720-turris-mox.dts
> > > +++ b/arch/arm64/boot/dts/marvell/armada-3720-turris-mox.dts
> > > @@ -305,7 +305,7 @@ phy1: ethernet-phy@1 {
> > > };
> > >
> > > /* switch nodes are enabled by U-Boot if modules are present */
> > > - switch0@10 {
> > > + switch@10 {
> >
> > As the comment says: U-Boot
> > (https://elixir.bootlin.com/u-boot/latest/source/board/CZ.NIC/turris_mox/turris_mox.c#L728)
> > sets up status = "okay" for these nodes depending on the MOXTET
> > configuration. It doesn't look as if it's doing that by alias, just by
> > path ("%s/switch%i@%x").
> >
> > I have a Turris MOX, please allow me some time to test if the node name
> > change is going to be significant and cause regressions. I expect the
> > answer to be yes (sadly).
>
> Yeah, it's bad.
>
> U-Boot 2018.11 (Dec 16 2018 - 12:50:19 +0000), Build: jenkins-turris-os-packages-kittens-mox-90
>
> DRAM: 1 GiB
> Enabling Armada 3720 wComphy-0: SGMII1 3.125 Gbps
> Comphy-1: PEX0 5 Gbps
> Comphy-2: USB3_HOST0 5 Gbps
> MMC: sdhci@d8000: 0
> Loading Environment from SPI Flash... SF: Detected w25q64dw with page size 256 Bytes, erase size 4 KiB, total 8 MiB
> OK
> Model: CZ.NIC Turris Mox Board
> Net: eth0: neta@30000
> Turris Mox:
> Board version: 22
> RAM size: 1024 MiB
> SD/eMMC version: SD
> Module Topology:
> 1: Peridot Switch Module (8-port)
> 2: Peridot Switch Module (8-port)
> 3: Peridot Switch Module (8-port)
> 4: SFP Module
>
> Hit any key to stop autoboot: 0
> => run sd_tftp_boot
> neta@30000 Waiting for PHY auto negotiation to complete....... done
> BOOTP broadcast 1
> BOOTP broadcast 2
> DHCP client bound to address 10.0.0.117 (254 ms)
> Using neta@30000 device
> TFTP from server 10.0.0.1; our IP address is 10.0.0.117
> Filename 'mox/armada-3720-turris-mox.dtb'.
> Load address: 0x4f00000
> Loading: ####
> 1.5 MiB/s
> done
> Bytes transferred = 19479 (4c17 hex)
> Using neta@30000 device
> TFTP from server 10.0.0.1; our IP address is 10.0.0.117
> Filename 'mox/Image'.
> Load address: 0x5000000
> Loading: #################################################################
> ##########################################
> 6 MiB/s
> done
> Bytes transferred = 54069760 (3390a00 hex)
> ## Flattened Device Tree blob at 04f00000
> Booting using the fdt blob at 0x4f00000
> Loading Device Tree to 000000003bf16000, end 000000003bf1dc16 ... OK
> ERROR: board-specific fdt fixup failed: FDT_ERR_NOTFOUND
> - must RESET the board to recover.
>
> FDT creation failed! hanging...### ERROR ### Please RESET the board ###
Yes, unfortunately changing that node name will break booting.
Maybe we could add a comment into the DTS to describe this unfortunate
state of things? :)
Marek
On Wed, Oct 18, 2023 at 11:03:45AM +0200, Linus Walleij wrote:
> +examples:
> + - |
> + #include <dt-bindings/gpio/gpio.h>
> + #include <dt-bindings/interrupt-controller/irq.h>
> + mdio {
> + #address-cells = <1>;
> + #size-cells = <0>;
> +
> + switch0: ethernet-switch@0 {
> + compatible = "marvell,mv88e6085";
> + reg = <0>;
> + reset-gpios = <&gpio5 1 GPIO_ACTIVE_LOW>;
> + interrupts-extended = <&gpio0 27 IRQ_TYPE_LEVEL_LOW>;
> + interrupt-controller;
> + #interrupt-cells = <2>;
> +
> + ethernet-ports {
> + #address-cells = <1>;
> + #size-cells = <0>;
> +
> + port@0 {
> + reg = <0>;
> + label = "lan1";
> + };
> + port@1 {
> + reg = <1>;
> + label = "lan2";
> + };
> + port@2 {
> + reg = <2>;
> + label = "lan3";
> + };
> + port@3 {
> + reg = <3>;
> + label = "lan4";
> + };
> + port@4 {
> + reg = <4>;
> + label = "wan";
> + };
> +
> + port@5 {
> + reg = <5>;
> + phy-mode = "sgmii";
> + ethernet = <ð2>;
> +
> + fixed-link {
> + speed = <1000>;
> + full-duplex;
> + };
> + };
> + };
> +
> + mdio {
> + #address-cells = <1>;
> + #size-cells = <0>;
> +
> + switch0phy0: ethernet-phy@0 {
> + reg = <0>;
> + interrupts-extended = <&switch0 0 IRQ_TYPE_LEVEL_HIGH>;
> + };
> + };
> + };
> + };
> +
> + - |
> + #include <dt-bindings/gpio/gpio.h>
> + #include <dt-bindings/interrupt-controller/irq.h>
> + mdio {
> + #address-cells = <1>;
> + #size-cells = <0>;
> +
> + switch1: ethernet-switch@0 {
> + compatible = "marvell,mv88e6190";
> + reg = <0>;
> + reset-gpios = <&gpio5 1 GPIO_ACTIVE_LOW>;
> + interrupts-extended = <&gpio0 27 IRQ_TYPE_LEVEL_LOW>;
> + interrupt-controller;
> + #interrupt-cells = <2>;
> +
> + ethernet-ports {
> + #address-cells = <1>;
> + #size-cells = <0>;
> +
> + port@0 {
> + reg = <0>;
> + label = "lan1";
> + };
> + port@1 {
> + reg = <1>;
> + label = "lan2";
> + };
> + port@2 {
> + reg = <2>;
> + label = "lan3";
> + };
> + port@3 {
> + reg = <3>;
> + label = "lan4";
> + };
> + };
> +
> + mdio {
> + #address-cells = <1>;
> + #size-cells = <0>;
> +
> + switch1phy0: ethernet-phy@0 {
> + reg = <0>;
> + interrupts-extended = <&switch1 0 IRQ_TYPE_LEVEL_HIGH>;
> + };
> + };
> +
> + mdio-external {
> + compatible = "marvell,mv88e6xxx-mdio-external";
> + #address-cells = <1>;
> + #size-cells = <0>;
> + switch1phy9: ethernet-phy@9 {
> + reg = <9>;
> + };
> + };
> + };
> + };
Yikes, both these examples are actually broken, for a reason that was
extensively discussed with Arınç in the past, and that he tried to
automatically detect through dt-schema but was ultimately told it's too
complicated.
https://patchwork.kernel.org/project/netdevbpf/cover/[email protected]/
Long story short: the "mdio" node is also the ds->slave_mii_bus (soon to
be ds->user_mii_bus after Florian's inclusivity changes). Having a
slave_mii_bus makes DSA know what to do with port nodes like this, which
don't have an explicit phy-handle:
port@3 {
reg = <3>;
label = "lan4";
};
but actually, call phy_connect() on the ds->slave_mii_bus at address 3
(the port "reg").
The issue is that phy_connect() won't work if ds->slave_mii_bus has an
OF presence, and ethernet-phy@3 isn't defined under it (which it isn't,
you only put ethernet-phy@9). The super detailed reason is that the
OF-based __of_mdiobus_register() does this:
/* Mask out all PHYs from auto probing. Instead the PHYs listed in
* the device tree are populated after the bus has been registered */
mdio->phy_mask = ~0;
So either:
- you delete the "mdio" node and the ethernet-phys under it, or
- you add all ethernet-phys under the mdio node, and put phy-handles
from ports to each of them, and phy-modes of "internal"
What you have now is exactly what won't work, i.e. an OF-based
slave_mii_bus with a non-OF-based phy_connect().
I don't want to see DT examples that are structurally broken, sorry,
because then we wonder why users are confused.
Personally, I would opt for adding the more modern explicit phy-handle
and phy-mode everywhere. Those also work with the U-Boot DM_DSA driver.
Just because we tolerate the bindings defined in the dark ages of DT
doesn't mean we should make an example out of them.
Also, you seem to have duplicated some work also done by Arınç but not
finalized (the mv88e6xxx schema conversion, on which you were also
copied). Let me add Marek here too, to make sure he's aware of 2
previous attempts and doesn't start working on a 3rd one :)
One other thing I see as a deal breaker for this schema conversion is
that $nodename for Marvell needs to allow basically anything (invalidating
the constraint from ethernet-switch.yaml), because we can't change node
names in the case of some boards, otherwise we risk breaking them
(see MOX). If the schema starts emitting warnings for those node names,
then it's inevitable that some pixie in the future will eventually break
them by "fixing" the node name.
On Wed, Oct 18, 2023 at 11:03:46AM +0200, Linus Walleij wrote:
> The Marvell MV88E6060 is one of the oldest DSA switches from
> Marvell, and it has DT bindings used in the wild. Let's define
> them properly.
>
> It is different enough from the rest of the MV88E6xxx switches
> that it deserves its own binding.
>
> Signed-off-by: Linus Walleij <[email protected]>
> ---
> .../bindings/net/dsa/marvell,mv88e6060.yaml | 90 ++++++++++++++++++++++
> MAINTAINERS | 1 +
> 2 files changed, 91 insertions(+)
>
> diff --git a/Documentation/devicetree/bindings/net/dsa/marvell,mv88e6060.yaml b/Documentation/devicetree/bindings/net/dsa/marvell,mv88e6060.yaml
> new file mode 100644
> index 000000000000..787f328551f6
> --- /dev/null
> +++ b/Documentation/devicetree/bindings/net/dsa/marvell,mv88e6060.yaml
> @@ -0,0 +1,90 @@
> +# SPDX-License-Identifier: (GPL-2.0-only OR BSD-2-Clause)
> +%YAML 1.2
> +---
> +$id: http://devicetree.org/schemas/net/dsa/marvell,mv88e6060.yaml#
> +$schema: http://devicetree.org/meta-schemas/core.yaml#
> +
> +title: Marvell MV88E6060 DSA switch
> +
> +maintainers:
> + - Andrew Lunn <[email protected]>
> +
> +description:
> + The Marvell MV88E6060 switch has been produced and sold by Marvell
> + since at least 2010. The switch has one pin ADDR4 that controls the
> + MDIO address of the switch to be 0x10 or 0x00, and on the MDIO bus
> + connected to the switch, the PHYs inside the switch appear as
> + independent devices on address 0x00-0x04 or 0x10-0x14, so in difference
> + from many other DSA switches this switch does not have an internal
> + MDIO bus for the PHY devices.
Where does 2010 come from (both here and in the other Marvell schema)?
Lennert Buytenhek added Linux support for this switch family in 2008.
Anyway,
Reviewed-by: Vladimir Oltean <[email protected]>
On Thu, Oct 19, 2023 at 05:26:49PM +0200, Marek Beh?n wrote:
> Yes, unfortunately changing that node name will break booting.
>
> Maybe we could add a comment into the DTS to describe this unfortunate
> state of things? :)
Well, the fact that Linus didn't notice means that there are insufficient
signals currently, so I guess a more explicit comment would help. Could
you prepare a patch?
On Thu, Oct 19, 2023 at 8:49 AM Rob Herring <[email protected]> wrote:
>
> On Wed, Oct 18, 2023 at 01:39:45PM +0200, Linus Walleij wrote:
> > On Wed, Oct 18, 2023 at 12:32 PM Rob Herring <[email protected]> wrote:
> >
> > > yamllint warnings/errors:
> > >
> > > dtschema/dtc warnings/errors:
> > > /builds/robherring/dt-review-ci/linux/Documentation/devicetree/bindings/net/marvell,mvusb.example.dtb: switch@0: ports: '#address-cells' is a required property
> > > from schema $id: http://devicetree.org/schemas/net/dsa/marvell,mv88e6xxx.yaml#
> > > /builds/robherring/dt-review-ci/linux/Documentation/devicetree/bindings/net/marvell,mvusb.example.dtb: switch@0: ports: '#size-cells' is a required property
> > > from schema $id: http://devicetree.org/schemas/net/dsa/marvell,mv88e6xxx.yaml#
> > > /builds/robherring/dt-review-ci/linux/Documentation/devicetree/bindings/net/marvell,mvusb.example.dtb: switch@0: ports: '#address-cells' is a required property
> > > from schema $id: http://devicetree.org/schemas/net/dsa/marvell,mv88e6xxx.yaml#
> > > /builds/robherring/dt-review-ci/linux/Documentation/devicetree/bindings/net/marvell,mvusb.example.dtb: switch@0: ports: '#size-cells' is a required property
> > > from schema $id: http://devicetree.org/schemas/net/dsa/marvell,mv88e6xxx.yaml#
> >
> > Fixed in patch 2/7?
>
> Yes. If one patch has errors we drop it. I should probably just give up
> on the rest of the series instead.
The bot should work better now not dropping patches when there are
warnings. It will give incremental new warnings with each patch.
Rob
On Thu, Oct 19, 2023 at 05:26:46PM +0300, Vladimir Oltean wrote:
> On Thu, Oct 19, 2023 at 08:45:14AM -0500, Rob Herring wrote:
> > On Wed, Oct 18, 2023 at 01:37:10PM +0200, Linus Walleij wrote:
> > > On Wed, Oct 18, 2023 at 12:32 PM Rob Herring <[email protected]> wrote:
> > >
> > > > dtschema/dtc warnings/errors:
> > > > Documentation/devicetree/bindings/net/marvell,mvusb.example.dtb: /example-0/usb/mdio@1/ethernet-switch@0: failed to match any schema with compatible: ['marvell,mv88e6190']
> > >
> > > Isn't that just because the bindings now come last in the series.
> > > Which is in response to a review comment, hence this warning
> > > didn't appear before.
> >
> > Yes. The only option that avoids this is squashing the 2 patches. I
> > think it is fine to leave this as-is.
> >
> > Rob
>
> Anyway, I'm surprised that the bot would send this email, since the
> warning existed prior to this patch, and I would expect that the bot
> only notifies of newly introduced issues.
Oh right. Except that the node name changed in this patch and that
changed the error message.
Rob
On 19.10.2023 18:35, Vladimir Oltean wrote:
> Yikes, both these examples are actually broken, for a reason that was
> extensively discussed with Arınç in the past, and that he tried to
> automatically detect through dt-schema but was ultimately told it's too
> complicated.
> https://patchwork.kernel.org/project/netdevbpf/cover/[email protected]/
True, and then I've figured that that approach was wrong from the start.
Read below.
>
> Long story short: the "mdio" node is also the ds->slave_mii_bus (soon to
> be ds->user_mii_bus after Florian's inclusivity changes). Having a
> slave_mii_bus makes DSA know what to do with port nodes like this, which
> don't have an explicit phy-handle:
>
> port@3 {
> reg = <3>;
> label = "lan4";
> };
>
> but actually, call phy_connect() on the ds->slave_mii_bus at address 3
> (the port "reg").
>
> The issue is that phy_connect() won't work if ds->slave_mii_bus has an
> OF presence, and ethernet-phy@3 isn't defined under it (which it isn't,
> you only put ethernet-phy@9). The super detailed reason is that the
> OF-based __of_mdiobus_register() does this:
>
> /* Mask out all PHYs from auto probing. Instead the PHYs listed in
> * the device tree are populated after the bus has been registered */
> mdio->phy_mask = ~0;
>
> So either:
>
> - you delete the "mdio" node and the ethernet-phys under it, or
> - you add all ethernet-phys under the mdio node, and put phy-handles
> from ports to each of them, and phy-modes of "internal"
>
> What you have now is exactly what won't work, i.e. an OF-based
> slave_mii_bus with a non-OF-based phy_connect().
>
> I don't want to see DT examples that are structurally broken, sorry,
> because then we wonder why users are confused.
>
> Personally, I would opt for adding the more modern explicit phy-handle
> and phy-mode everywhere. Those also work with the U-Boot DM_DSA driver.
As can be seen on the conversation on the patch series you've referred to
above, this is ultimately what I've proposed to enforce on all ethernet
controllers (I hadn't mentioned phy-mode there yet). As long as Andrew
holds his stance, this won't happen. The whole conversation on that patch
series is a great read as to how a devicetree should be defined and how
drivers should interact with it.
But sure, they can at least be put on the examples for this specific
schema.
Arınç
On Thu, Oct 19, 2023 at 5:35 PM Vladimir Oltean <[email protected]> wrote:
> Yikes, both these examples are actually broken,
As you can see from the patch, they are just carried over from
Documentation/devicetree/bindings/net/dsa/marvell.txt
+/- fixes to make them pass schema checks.
> So either:
>
> - you delete the "mdio" node and the ethernet-phys under it, or
> - you add all ethernet-phys under the mdio node, and put phy-handles
> from ports to each of them, and phy-modes of "internal"
>
> What you have now is exactly what won't work, i.e. an OF-based
> slave_mii_bus with a non-OF-based phy_connect().
Yeah when I run check_dtbs I get a few (not many) warnings
like this on aarch64 and armv7_multi:
arch/arm/boot/dts/nxp/imx/imx6q-b450v3.dtb: switch@0: ports:port@4:
'phy-mode' is a required property
from schema $id:
http://devicetree.org/schemas/net/dsa/marvell,mv88e6xxx.yaml#
Isn't there some in-kernel DTS file with a *good* example of how
a Marvell mv88e6xxx switch is supposed to look I can just
copy instead? We shouldn't conjure synthetic examples.
> I don't want to see DT examples that are structurally broken, sorry,
> because then we wonder why users are confused.
These examples are already in the kernel. Migrating them
from marvell.txt to marvell,mv88e6xxx.yaml doesn't make
the situation worse, it's not like people magically start trusting
the examples more because they are in YAML than in .txt.
But sure let's try to put in better examples!
> Personally, I would opt for adding the more modern explicit phy-handle
> and phy-mode everywhere.
I'm game. Point out the DTS file and I will take that.
> Also, you seem to have duplicated some work also done by Arınç but not
> finalized (the mv88e6xxx schema conversion, on which you were also
> copied). Let me add Marek here too, to make sure he's aware of 2
> previous attempts and doesn't start working on a 3rd one :)
Haha I forgot :D
> One other thing I see as a deal breaker for this schema conversion is
> that $nodename for Marvell needs to allow basically anything (invalidating
> the constraint from ethernet-switch.yaml), because we can't change node
> names in the case of some boards, otherwise we risk breaking them
> (see MOX). If the schema starts emitting warnings for those node names,
> then it's inevitable that some pixie in the future will eventually break
> them by "fixing" the node name.
I already did a bit of hippo-in-china-porcelain store in the patches
in this series mostly renaming things like "switch0@0" to "switch@0"
(yeah that's all).
Is this part of the problem or something else?
I run the binding checks across all aarch64 and armv7_multi platforms
on this patch set without any major issues.
Yours,
Linus Walleij
On Thu, Oct 19, 2023 at 6:22 PM Vladimir Oltean <[email protected]> wrote:
> On Thu, Oct 19, 2023 at 05:26:49PM +0200, Marek Behún wrote:
> > Yes, unfortunately changing that node name will break booting.
> >
> > Maybe we could add a comment into the DTS to describe this unfortunate
> > state of things? :)
>
> Well, the fact that Linus didn't notice means that there are insufficient
> signals currently, so I guess a more explicit comment would help. Could
> you prepare a patch?
I can just include a blurb in my patch so we don't get colliding
changes.
Yours,
Linus Walleij
On Fri, Oct 20, 2023 at 02:47:20PM +0200, Linus Walleij wrote:
> On Thu, Oct 19, 2023 at 5:35 PM Vladimir Oltean <[email protected]> wrote:
>
> > Yikes, both these examples are actually broken,
>
> As you can see from the patch, they are just carried over from
> Documentation/devicetree/bindings/net/dsa/marvell.txt
>
> +/- fixes to make them pass schema checks.
(...)
> These examples are already in the kernel. Migrating them
> from marvell.txt to marvell,mv88e6xxx.yaml doesn't make
> the situation worse, it's not like people magically start trusting
> the examples more because they are in YAML than in .txt.
>
> But sure let's try to put in better examples!
You are not correct here. The examples from
Documentation/devicetree/bindings/net/dsa/marvell.txt don't have ports,
and the way in which you added the ports is wrong (at least relative to
the way in which you kept the mdio node).
> > What you have now is exactly what won't work, i.e. an OF-based
> > slave_mii_bus with a non-OF-based phy_connect().
>
> Yeah when I run check_dtbs I get a few (not many) warnings
> like this on aarch64 and armv7_multi:
>
> arch/arm/boot/dts/nxp/imx/imx6q-b450v3.dtb: switch@0: ports:port@4:
> 'phy-mode' is a required property
> from schema $id:
> http://devicetree.org/schemas/net/dsa/marvell,mv88e6xxx.yaml#
Ok, the warning is valid, but I don't know what phy-mode to put there.
It is unrelated anyway. Some warnings will be expected after the schema
conversion, and they are not all mechanical to fix. When we put schema
validation in place for checking that CPU ports have valid link
descriptions that phylink can use, we decided to be lax in the kernel,
but strict in the dt-schema. Hmm, not sure what were we thinking.
We didn't have a schema for Marvell, so we weren't even seeing many of
the validation errors that you're now uncovering.
> Isn't there some in-kernel DTS file with a *good* example of how
> a Marvell mv88e6xxx switch is supposed to look I can just
> copy instead? We shouldn't conjure synthetic examples.
(...)
> I'm game. Point out the DTS file and I will take that.
You can use https://elixir.bootlin.com/u-boot/latest/source/arch/arm/dts/imx6qdl-gw5904.dtsi#L211
(optionally renaming switch to ethernet-switch, and ports to ethernet-ports).
That uses the subset of the mv88e6xxx kernel bindings that U-Boot also
understands, so taking a U-Boot example is actually preferable.
> > One other thing I see as a deal breaker for this schema conversion is
> > that $nodename for Marvell needs to allow basically anything (invalidating
> > the constraint from ethernet-switch.yaml), because we can't change node
> > names in the case of some boards, otherwise we risk breaking them
> > (see MOX). If the schema starts emitting warnings for those node names,
> > then it's inevitable that some pixie in the future will eventually break
> > them by "fixing" the node name.
>
> I already did a bit of hippo-in-china-porcelain store in the patches
> in this series mostly renaming things like "switch0@0" to "switch@0"
> (yeah that's all).
>
> Is this part of the problem or something else?
Yes, for most of the switches, renaming their OF nodes should not be a problem.
For Marvell, I'd exercise extra caution and only rename those OF nodes
where I can confirm that doing so won't break anything. Marvell is one
of the oldest DSA drivers, and you can tell that the bindings have gone
through a lot before becoming more or less uniform.
Anyway, for the $nodename constraint, it _looks_ all mechanical and trivial
to fix (unlike the missing phy-mode that you point to, above), so someone
will jump to fix it. I would like to avoid that, because boot testing
will be key, and a board is not always available.
> Isn't there some in-kernel DTS file with a *good* example of how
> a Marvell mv88e6xxx switch is supposed to look I can just
> copy instead? We shouldn't conjure synthetic examples.
arch/arm/boot/dts/marvell/armada-381-netgear-gs110emx.dts is an
example of a 6390 with external PHYs. The nodes are in a somewhat
unconventional order, but nothing requires them to be any specific
order.
arch/arm/boot/dts/nxp/vf/vf610-zii-ssmb-spu3.dts is another 6390 which
only uses the internal PHYs, so there are no mdio busses listed or
needed.
arch/arm/boot/dts/marvell/armada-370-rd.dts is another example of how
it can be done. It lists the PHYs on the MDIO bus, but does not have
any phy-handles, it lets the DSA core link them. Some people might
consider this a bad example? But it works, i use this machine for
development work.
arch/arm/boot/dts/marvell/armada-385-linksys.dtsi is another
alternative, which does not have the MDIO bus.
Andrew
On Fri, Oct 20, 2023 at 02:59:43PM +0200, Linus Walleij wrote:
> On Thu, Oct 19, 2023 at 6:22 PM Vladimir Oltean <[email protected]> wrote:
> > On Thu, Oct 19, 2023 at 05:26:49PM +0200, Marek Behún wrote:
> > > Yes, unfortunately changing that node name will break booting.
> > >
> > > Maybe we could add a comment into the DTS to describe this unfortunate
> > > state of things? :)
> >
> > Well, the fact that Linus didn't notice means that there are insufficient
> > signals currently, so I guess a more explicit comment would help. Could
> > you prepare a patch?
>
> I can just include a blurb in my patch so we don't get colliding
> changes.
Colliding with what, if you shouldn't change the node names?