Skip to contentSkip to conversation

net: stmmac: don't create the MDIO bus if there's no mdio node on DT

PATCH In review Bartosz Golaszewski 3 participants 4 messages Aug 8, 2023 – Aug 10, 2023

Subsystems (4)
lore
Share & tools
lore
#0 lore raw

PATCH 1/1 net: stmmac: don't create the MDIO bus if there's no mdio node on DT

From: Bartosz Golaszewski <[email protected]>

The stmmac_dt_phy() function that parses the device-tree node of the MAC and allocates the MDIO and PHY resources misses one use-case: when the MAC doesn't have a fixed link but also doesn't define its own mdio bus on the device tree and instead shares the MDIO lines with a different MAC with its PHY phandle reaching over into a different node.

As this function could also use some more readability, rework it to handle this use-case and simplify the code.

---
 .../ethernet/stmicro/stmmac/stmmac_platform.c | 26 +++++++++----------
 1 file changed, 13 insertions(+), 13 deletions(-)
drivers/net/ethernet/stmicro/stmmac/stmmac_platform.c +13 -13
index be8e79c7aa34..91844673df43 100644
@@ -320,12 +320,14 @@ static int stmmac_mtl_setup(struct platform_device *pdev,
 static int stmmac_dt_phy(struct plat_stmmacenet_data *plat,
 			 struct device_node *np, struct device *dev)
 {
-	bool mdio = !of_phy_is_fixed_link(np);
 	static const struct of_device_id need_mdio_ids[] = {
 		{ .compatible = "snps,dwc-qos-ethernet-4.10" },
 		{},
 	};
 
+	if (of_phy_is_fixed_link(np))
+		return 0;
+
 	if (of_match_node(need_mdio_ids, np)) {
 		plat->mdio_node = of_get_child_by_name(np, "mdio");
 	} else {
@@ -340,20 +342,18 @@ static int stmmac_dt_phy(struct plat_stmmacenet_data *plat,
 		}
 	}
 
-	if (plat->mdio_node) {
-		dev_dbg(dev, "Found MDIO subnode\n");
-		mdio = true;
-	}
+	if (!plat->mdio_node)
+		return 0;
 
-	if (mdio) {
-		plat->mdio_bus_data =
-			devm_kzalloc(dev, sizeof(struct stmmac_mdio_bus_data),
-				     GFP_KERNEL);
-		if (!plat->mdio_bus_data)
-			return -ENOMEM;
+	dev_dbg(dev, "Found MDIO subnode\n");
 
-		plat->mdio_bus_data->needs_reset = true;
-	}
+	plat->mdio_bus_data = devm_kzalloc(dev,
+					   sizeof(struct stmmac_mdio_bus_data),
+					   GFP_KERNEL);
+	if (!plat->mdio_bus_data)
+		return -ENOMEM;
+
+	plat->mdio_bus_data->needs_reset = true;
 
 	return 0;
 }
Signature · 7 lines
2.39.2


_______________________________________________
linux-arm-kernel mailing list
[email protected]
http://lists.infradead.org/mailman/listinfo/linux-arm-kernel
#1 lore raw
Quoted 7 lines · Bartosz Golaszewski

From: Bartosz Golaszewski <[email protected]>

The stmmac_dt_phy() function that parses the device-tree node of the MAC and allocates the MDIO and PHY resources misses one use-case: when the MAC doesn't have a fixed link but also doesn't define its own mdio bus on the device tree and instead shares the MDIO lines with a different MAC with its PHY phandle reaching over into a different node.

It does not share the MDIO lines. The other MDIO bus master happens to have two PHYs and there are no PHYs on this MDIO bus, so no point instantiating it.

drivers/net/ethernet/stmicro/stmmac/stmmac_platform.c static int stmmac_mtl_setup(struct platform_device *pdev, · line 330

Quoted 12 lines

static int stmmac_dt_phy(struct plat_stmmacenet_data *plat,

drivers/net/ethernet/stmicro/stmmac/stmmac_platform.c +3 -1
 			 struct device_node *np, struct device *dev)
 {
-	bool mdio = !of_phy_is_fixed_link(np);
 	static const struct of_device_id need_mdio_ids[] = {
 		{ .compatible = "snps,dwc-qos-ethernet-4.10" },
 		{},
 	};
 
+	if (of_phy_is_fixed_link(np))
+		return 0;
+
		/**
		 * If snps,dwmac-mdio is passed from DT, always register
		 * the MDIO
		 */
		for_each_child_of_node(np, plat->mdio_node) {
			if (of_device_is_compatible(plat->mdio_node,
						    "snps,dwmac-mdio"))
				break;
		}

The comment suggests it should always be registered. This MAC might have a fixed-phy, but that does not mean there is not an Ethernet switch on the bus, or a PHY for some other MAC etc. MDIO busses masters should be considered fully independent devices.

https://elixir.bootlin.com/linux/v6.5-rc5/source/arch/arm/boot/dts/nxp/vf/vf610-zii-ssmb-dtu.dts

&fec1 {
	phy-mode = "rmii";
	pinctrl-names = "default";
	pinctrl-0 = <&pinctrl_fec1>;
	status = "okay";

	fixed-link {
		speed = <100>;
		full-duplex;
	};

	mdio1: mdio {
		#address-cells = <1>;
		#size-cells = <0>;
		clock-frequency = <12500000>;
		suppress-preamble;
		status = "okay";

		switch0: switch0@0 {
			compatible = "marvell,mv88e6190";
			pinctrl-0 = <&pinctrl_gpio_switch0>;
			pinctrl-names = "default";
			reg = <0>;
			eeprom-length = <65536>;
			interrupt-parent = <&gpio3>;
			interrupts = <2 IRQ_TYPE_LEVEL_LOW>;
			interrupt-controller;
			#interrupt-cells = <2>;

Both a fixed link, and something on the MDIO bus....

Andrew

_______________________________________________
linux-arm-kernel mailing list
[email protected]
http://lists.infradead.org/mailman/listinfo/linux-arm-kernel

#2 lore raw
Quoted 10 lines · Andrew Lunn
Quoted 7 lines · Bartosz Golaszewski

From: Bartosz Golaszewski <[email protected]>

The stmmac_dt_phy() function that parses the device-tree node of the MAC and allocates the MDIO and PHY resources misses one use-case: when the MAC doesn't have a fixed link but also doesn't define its own mdio bus on the device tree and instead shares the MDIO lines with a different MAC with its PHY phandle reaching over into a different node.

Quoted 3 lines

It does not share the MDIO lines. The other MDIO bus master happens to have two PHYs and there are no PHYs on this MDIO bus, so no point instantiating it.

Yes, I sent it before we established that thanks to Andrew's input.

Quoted 60 lines

drivers/net/ethernet/stmicro/stmmac/stmmac_platform.c static int stmmac_mtl_setup(struct platform_device *pdev, · line 330

Quoted 13 lines

static int stmmac_dt_phy(struct plat_stmmacenet_data *plat,
struct device_node *np, struct device *dev)
{

drivers/net/ethernet/stmicro/stmmac/stmmac_platform.c +3 -1
-     bool mdio = !of_phy_is_fixed_link(np);
      static const struct of_device_id need_mdio_ids[] = {
              { .compatible = "snps,dwc-qos-ethernet-4.10" },
              {},
      };
 
+     if (of_phy_is_fixed_link(np))
+             return 0;
+
                /**
                 * If snps,dwmac-mdio is passed from DT, always register
                 * the MDIO
                 */
                for_each_child_of_node(np, plat->mdio_node) {
                        if (of_device_is_compatible(plat->mdio_node,
                                                    "snps,dwmac-mdio"))
                                break;
                }

The comment suggests it should always be registered. This MAC might have a fixed-phy, but that does not mean there is not an Ethernet switch on the bus, or a PHY for some other MAC etc. MDIO busses masters should be considered fully independent devices.

https://elixir.bootlin.com/linux/v6.5-rc5/source/arch/arm/boot/dts/nxp/vf/vf610-zii-ssmb-dtu.dts

&fec1 {
        phy-mode = "rmii";
        pinctrl-names = "default";
        pinctrl-0 = <&pinctrl_fec1>;
        status = "okay";

        fixed-link {
                speed = <100>;
                full-duplex;
        };

        mdio1: mdio {
                #address-cells = <1>;
                #size-cells = <0>;
                clock-frequency = <12500000>;
                suppress-preamble;
                status = "okay";

                switch0: switch0@0 {
                        compatible = "marvell,mv88e6190";
                        pinctrl-0 = <&pinctrl_gpio_switch0>;
                        pinctrl-names = "default";
                        reg = <0>;
                        eeprom-length = <65536>;
                        interrupt-parent = <&gpio3>;
                        interrupts = <2 IRQ_TYPE_LEVEL_LOW>;
                        interrupt-controller;
                        #interrupt-cells = <2>;
Quoted 3 lines

Both a fixed link, and something on the MDIO bus....

Andrew

Makes sense, we can drop all my stmmac patches from today, I need to rethink it in detail.

Bart

_______________________________________________
linux-arm-kernel mailing list
[email protected]
http://lists.infradead.org/mailman/listinfo/linux-arm-kernel

maintainer Acked-by on v1 1/1
#3 lore raw
In reply to Bartosz Golaszewski
Quoted 57 lines · Bartosz Golaszewski

From: Bartosz Golaszewski <[email protected]>

The stmmac_dt_phy() function that parses the device-tree node of the MAC and allocates the MDIO and PHY resources misses one use-case: when the MAC doesn't have a fixed link but also doesn't define its own mdio bus on the device tree and instead shares the MDIO lines with a different MAC with its PHY phandle reaching over into a different node.

As this function could also use some more readability, rework it to handle this use-case and simplify the code.

---
  .../ethernet/stmicro/stmmac/stmmac_platform.c | 26 +++++++++----------
  1 file changed, 13 insertions(+), 13 deletions(-)
drivers/net/ethernet/stmicro/stmmac/stmmac_platform.c +6 -13
index be8e79c7aa34..91844673df43 100644
@@ -320,12 +320,14 @@ static int stmmac_mtl_setup(struct platform_device *pdev,
  static int stmmac_dt_phy(struct plat_stmmacenet_data *plat,
  			 struct device_node *np, struct device *dev)
  {
-	bool mdio = !of_phy_is_fixed_link(np);
  	static const struct of_device_id need_mdio_ids[] = {
  		{ .compatible = "snps,dwc-qos-ethernet-4.10" },
  		{},
  	};
 
+	if (of_phy_is_fixed_link(np))
+		return 0;
+
  	if (of_match_node(need_mdio_ids, np)) {
  		plat->mdio_node = of_get_child_by_name(np, "mdio");
  	} else {
@@ -340,20 +342,18 @@ static int stmmac_dt_phy(struct plat_stmmacenet_data *plat,
  		}
  	}
 
-	if (plat->mdio_node) {
-		dev_dbg(dev, "Found MDIO subnode\n");
-		mdio = true;
-	}
+	if (!plat->mdio_node)
+		return 0;
 
-	if (mdio) {
-		plat->mdio_bus_data =
-			devm_kzalloc(dev, sizeof(struct stmmac_mdio_bus_data),
-				     GFP_KERNEL);
-		if (!plat->mdio_bus_data)
-			return -ENOMEM;
+	dev_dbg(dev, "Found MDIO subnode\n");
 
-		plat->mdio_bus_data->needs_reset = true;
-	}

drivers/net/ethernet/stmicro/stmmac/stmmac_platform.c static int stmmac_dt_phy(struct plat_stmmacenet_data *plat, · line 359

Quoted 10 lines
drivers/net/ethernet/stmicro/stmmac/stmmac_platform.c +7 -0
+	plat->mdio_bus_data = devm_kzalloc(dev,
+					   sizeof(struct stmmac_mdio_bus_data),
+					   GFP_KERNEL);
+	if (!plat->mdio_bus_data)
+		return -ENOMEM;
+
+	plat->mdio_bus_data->needs_reset = true;
 
  	return 0;
  }

Regards
Alex