Skip to contentSkip to conversation

This series adds ethernet support for RPi4.

PATCH In review [email protected] 4 participants 14 messages Oct 11, 2019 – Oct 15, 2019

Subsystems (5)
lore
Share & tools
lore
#0 lore raw

cover This series adds ethernet support for RPi4.

From: Matthias Brugger <[email protected]>

Raspberry Pi 4 uses the broadcom genet chip in version five. This chip has a dma controller integrated. Up to now the maximal burst size was hard-coded to 0x10. But it turns out that Raspberry Pi 4 does only work with the smaller maximal burst size of 0x8.

This series adds a new optional property to the driver, dma-burst-sz. The very same property is already used by another drivers in the kernel.

Matthias Brugger (3):
dt-bindings: net: bcmgenet add property for max DMA burst size
net: bcmgenet: use optional max DMA burst size property
ARM: dts: bcm2711: Enable GENET support for the RPi4

 .../devicetree/bindings/net/brcm,bcmgenet.txt |  2 ++
 arch/arm/boot/dts/bcm2711-rpi-4-b.dts         | 22 +++++++++++++++++++
 arch/arm/boot/dts/bcm2711.dtsi                | 18 +++++++++++++++
 .../net/ethernet/broadcom/genet/bcmgenet.c    | 13 +++++++++--
 .../net/ethernet/broadcom/genet/bcmgenet.h    |  1 +
 5 files changed, 54 insertions(+), 2 deletions(-)
Signature · 1 line
2.23.0
#1 lore raw

PATCH 1/3 dt-bindings: net: bcmgenet add property for max DMA burst size

From: Matthias Brugger <[email protected]>

The maximal usable DMA burst size can vary in different SoCs. Add a optional property to configure the DMA channels properly.

---

 Documentation/devicetree/bindings/net/brcm,bcmgenet.txt | 2 ++
 1 file changed, 2 insertions(+)
Documentation/devicetree/bindings/net/brcm,bcmgenet.txt +2 -0
index 3956af1d30f3..10a7169ec902 100644
@@ -30,6 +30,8 @@ Optional properties:
   See Documentation/devicetree/bindings/net/fixed-link.txt for information on
   the property specifics
 
+- dma-burst-sz: Maximal length of a burst.
+
 Required child nodes:
 
 - mdio bus node: this node should always be present regardless of the PHY
Signature · 1 line
2.23.0
#2 lore raw

Re: [PATCH v1 1/3] dt-bindings: net: bcmgenet add property for max DMA burst size

Quoted 6 lines · On 10/11/19 11:48 AM, [email protected] wrote:

From: Matthias Brugger <[email protected]>

The maximal usable DMA burst size can vary in different SoCs. Add a optional property to configure the DMA channels properly.

There are enough differences in the integration of GENET in 2711 (versus any other STB chip where it has been used) that a brcm,bcm2711-genet-v5 compatible string would be in order. From there, you can derive the DMA burst size without using additional properties.

Signature · 1 line
Florian
#3 lore raw

PATCH 2/3 net: bcmgenet: use optional max DMA burst size property

In reply to [email protected]

From: Matthias Brugger <[email protected]>

Depending on the HW, the maximal usable DMA burst size can vary. If not set accordingly a timeout in the transmit queue happens and no package can be sent. Read to optional max-burst-sz property, if not present, fallback to the standard value.

---

 drivers/net/ethernet/broadcom/genet/bcmgenet.c | 13 +++++++++++--
 drivers/net/ethernet/broadcom/genet/bcmgenet.h |  1 +
 2 files changed, 12 insertions(+), 2 deletions(-)
drivers/net/ethernet/broadcom/genet/bcmgenet.c +11 -2
index 12cb77ef1081..a7bb822a6d83 100644
@@ -2576,7 +2576,8 @@ static int bcmgenet_init_dma(struct bcmgenet_priv *priv)
 	}
 
 	/* Init rDma */
-	bcmgenet_rdma_writel(priv, DMA_MAX_BURST_LENGTH, DMA_SCB_BURST_SIZE);
+	bcmgenet_rdma_writel(priv, priv->dma_max_burst_length,
+			     DMA_SCB_BURST_SIZE);
 
 	/* Initialize Rx queues */
 	ret = bcmgenet_init_rx_queues(priv->dev);
@@ -2589,7 +2590,8 @@ static int bcmgenet_init_dma(struct bcmgenet_priv *priv)
 	}
 
 	/* Init tDma */
-	bcmgenet_tdma_writel(priv, DMA_MAX_BURST_LENGTH, DMA_SCB_BURST_SIZE);
+	bcmgenet_tdma_writel(priv, priv->dma_max_burst_length,
+			     DMA_SCB_BURST_SIZE);
 
 	/* Initialize Tx queues */
 	bcmgenet_init_tx_queues(priv->dev);
@@ -3522,6 +3524,13 @@ static int bcmgenet_probe(struct platform_device *pdev)
 
 	clk_prepare_enable(priv->clk);
 
+	if (dn) {
+		of_property_read_u32(dn, "dma-burst-sz",
+				     &priv->dma_max_burst_length);
+	} else {
+		priv->dma_max_burst_length = DMA_MAX_BURST_LENGTH;
+	}
+
 	bcmgenet_set_hw_params(priv);
 
 	/* Mii wait queue */
drivers/net/ethernet/broadcom/genet/bcmgenet.h +1 -0
index 4a8fc03d82fd..897f356eb376 100644
@@ -663,6 +663,7 @@ struct bcmgenet_priv {
 	bool crc_fwd_en;
 
 	unsigned int dma_rx_chk_bit;
+	unsigned int dma_max_burst_length;
 
 	u32 msg_enable;
 
Signature · 1 line
2.23.0
#4 lore raw

Re: [PATCH v1 2/3] net: bcmgenet: use optional max DMA burst size property

Hi Matthias!
One small comment, I'll test everything later on.

Quoted 40 lines · On Fri, 2019-10-11 at 20:48 +0200, [email protected] wrote:

From: Matthias Brugger <[email protected]>

Depending on the HW, the maximal usable DMA burst size can vary. If not set accordingly a timeout in the transmit queue happens and no package can be sent. Read to optional max-burst-sz property, if not present, fallback to the standard value.

---

 drivers/net/ethernet/broadcom/genet/bcmgenet.c | 13 +++++++++++--
 drivers/net/ethernet/broadcom/genet/bcmgenet.h |  1 +
 2 files changed, 12 insertions(+), 2 deletions(-)
drivers/net/ethernet/broadcom/genet/bcmgenet.c +4 -2
index 12cb77ef1081..a7bb822a6d83 100644
@@ -2576,7 +2576,8 @@ static int bcmgenet_init_dma(struct bcmgenet_priv *priv)
 	}
 
 	/* Init rDma */
-	bcmgenet_rdma_writel(priv, DMA_MAX_BURST_LENGTH, DMA_SCB_BURST_SIZE);
+	bcmgenet_rdma_writel(priv, priv->dma_max_burst_length,
+			     DMA_SCB_BURST_SIZE);
 
 	/* Initialize Rx queues */
 	ret = bcmgenet_init_rx_queues(priv->dev);
@@ -2589,7 +2590,8 @@ static int bcmgenet_init_dma(struct bcmgenet_priv *priv)
 	}
 
 	/* Init tDma */
-	bcmgenet_tdma_writel(priv, DMA_MAX_BURST_LENGTH, DMA_SCB_BURST_SIZE);
+	bcmgenet_tdma_writel(priv, priv->dma_max_burst_length,
+			     DMA_SCB_BURST_SIZE);
 
 	/* Initialize Tx queues */
 	bcmgenet_init_tx_queues(priv->dev);

drivers/net/ethernet/broadcom/genet/bcmgenet.c static int bcmgenet_probe(struct platform_device *pdev) · line 3529

Quoted 6 lines
drivers/net/ethernet/broadcom/genet/bcmgenet.c +3 -0
@@ -3522,6 +3524,13 @@ static int bcmgenet_probe(struct platform_device *pdev)
 
 	clk_prepare_enable(priv->clk);
 
+	if (dn) {
+		of_property_read_u32(dn, "dma-burst-sz",
+				     &priv->dma_max_burst_length);

You set the 'dma-burst-sz' binding as optional, though this assumes that if a device node is available dma_max_burst_length comes from the device tree. I think you should check of_property_read_u32's return value and on failure default to DMA_MAX_BURST_LENGTH.

Quoted 13 lines
drivers/net/ethernet/broadcom/genet/bcmgenet.c +4 -0
+	} else {
+		priv->dma_max_burst_length = DMA_MAX_BURST_LENGTH;
+	}
+
 	bcmgenet_set_hw_params(priv);
 
 	/* Mii wait queue */
drivers/net/ethernet/broadcom/genet/bcmgenet.h +0 -0
index 4a8fc03d82fd..897f356eb376 100644

drivers/net/ethernet/broadcom/genet/bcmgenet.h struct bcmgenet_priv { · line 669

Quoted 7 lines
drivers/net/ethernet/broadcom/genet/bcmgenet.h +1 -0
@@ -663,6 +663,7 @@ struct bcmgenet_priv {
 	bool crc_fwd_en;
 
 	unsigned int dma_rx_chk_bit;
+	unsigned int dma_max_burst_length;
 
 	u32 msg_enable;
 

Regards,
Nicolas

#5 lore raw

PATCH 3/3 ARM: dts: bcm2711: Enable GENET support for the RPi4

In reply to [email protected]

From: Matthias Brugger <[email protected]>

Enable Gigabit Ethernet support on the Raspberry Pi 4 Model B.

---

 arch/arm/boot/dts/bcm2711-rpi-4-b.dts | 22 ++++++++++++++++++++++
 arch/arm/boot/dts/bcm2711.dtsi        | 18 ++++++++++++++++++
 2 files changed, 40 insertions(+)
arch/arm/boot/dts/bcm2711-rpi-4-b.dts +22 -0
index cccc1ccd19be..958553d62670 100644
@@ -97,6 +97,28 @@
 	status = "okay";
 };
 
+&genet {
+	phy-handle = <&phy1>;
+	phy-mode = "rgmii";
+	status = "okay";
+	dma-burst-sz = <0x08>;
+
+	mdio@e14 {
+		compatible = "brcm,genet-mdio-v5";
+		reg = <0xe14 0x8>;
+		reg-names = "mdio";
+		#address-cells = <0x0>;
+		#size-cells = <0x1>;
+
+		phy1: ethernet-phy@1 {
+			compatible = "ethernet-phy-ieee802.3-c22";
+			/* No PHY interrupt */
+			max-speed = <1000>;
+			reg = <0x1>;
+		};
+	};
+};
+
 /* uart0 communicates with the BT module */
 &uart0 {
 	pinctrl-names = "default";
arch/arm/boot/dts/bcm2711.dtsi +18 -0
index ac83dac2e6ba..e2e837fcad59 100644
@@ -305,6 +305,24 @@
 			cpu-release-addr = <0x0 0x000000f0>;
 		};
 	};
+
+	scb {
+		compatible = "simple-bus";
+		#address-cells = <2>;
+		#size-cells = <1>;
+
+		ranges = <0x0 0x7c000000  0x0 0xfc000000  0x03800000>;
+
+		genet: ethernet@7d580000 {
+			compatible = "brcm,genet-v5";
+			reg = <0x0 0x7d580000 0x10000>;
+			#address-cells = <0x1>;
+			#size-cells = <0x1>;
+			interrupts = <GIC_SPI 157 IRQ_TYPE_LEVEL_HIGH>,
+				     <GIC_SPI 158 IRQ_TYPE_LEVEL_HIGH>;
+			status = "disabled";
+		};
+	};
 };
 
 &clk_osc {
Signature · 1 line
2.23.0
#6 lore raw

Re: [PATCH v1 3/3] ARM: dts: bcm2711: Enable GENET support for the RPi4

Quoted 18 lines · On 10/11/19 11:48 AM, [email protected] wrote:

From: Matthias Brugger <[email protected]>

Enable Gigabit Ethernet support on the Raspberry Pi 4 Model B.

---

 arch/arm/boot/dts/bcm2711-rpi-4-b.dts | 22 ++++++++++++++++++++++
 arch/arm/boot/dts/bcm2711.dtsi        | 18 ++++++++++++++++++
 2 files changed, 40 insertions(+)
arch/arm/boot/dts/bcm2711-rpi-4-b.dts +0 -0
index cccc1ccd19be..958553d62670 100644

arch/arm/boot/dts/bcm2711-rpi-4-b.dts line 102

Quoted 6 lines
arch/arm/boot/dts/bcm2711-rpi-4-b.dts +3 -0
@@ -97,6 +97,28 @@
 	status = "okay";
 };
 
+&genet {
+	phy-handle = <&phy1>;
+	phy-mode = "rgmii";

Can you check that things still work against David Miller's net-next? Tree, in particular the BCM54213PE PHY might be matched by the BCM54210E entry in drivers/net/phy/broadcom.c and I just fixed an issue in how RGMII delays were configured:

https://git.kernel.org/pub/scm/linux/kernel/git/davem/net-next.git/commit/?id=fea7fda7f50a6059220f83251e70709e45cc8040

This might require you to change the 'phy-mode' property to what is appropriate.

arch/arm/boot/dts/bcm2711-rpi-4-b.dts line 114

Quoted 12 lines
arch/arm/boot/dts/bcm2711-rpi-4-b.dts +12 -0
+	status = "okay";
+	dma-burst-sz = <0x08>;
+
+	mdio@e14 {
+		compatible = "brcm,genet-mdio-v5";
+		reg = <0xe14 0x8>;
+		reg-names = "mdio";
+		#address-cells = <0x0>;
+		#size-cells = <0x1>;
+
+		phy1: ethernet-phy@1 {
+			compatible = "ethernet-phy-ieee802.3-c22";

This does not hurt, but this compatibility string is not required.

arch/arm/boot/dts/bcm2711-rpi-4-b.dts line 116

Quoted 2 lines
arch/arm/boot/dts/bcm2711-rpi-4-b.dts +2 -0
+			/* No PHY interrupt */
+			max-speed = <1000>;

And this property is not required either, since the PHY library will determine the PHY's capabilities.

Other than those patches, I believe you also need something like this (inspired by the Rpi downstream patch):

(quoted diff) +0 -0

b/drivers/net/ethernet/broadcom/genet/bcmmii.c
index 970e478a9017..94d1dd5d56bf 100644

drivers/net/ethernet/broadcom/genet/bcmmii.c +3 -2
@@ -273,11 +273,12 @@ int bcmgenet_mii_probe(struct net_device *dev)
        struct bcmgenet_priv *priv = netdev_priv(dev);
        struct device_node *dn = priv->pdev->dev.of_node;
        struct phy_device *phydev;
-       u32 phy_flags;
+       u32 phy_flags = 0;
        int ret;
 
        /* Communicate the integrated PHY revision */
-       phy_flags = priv->gphy_rev;
+       if (priv->internal_phy)
+               phy_flags = priv->gphy_rev;
 
        /* Initialize link state variables that bcmgenet_mii_setup() uses */
        priv->old_link = -1;

to prevent the internal PHY revision, which we stick into phydev->dev_flags from incorrectly setting bits in drivers/net/phy/broadcom.c. I will probably send this as a fix in the next few hours.

Signature · 1 line
Florian
#7 lore raw

Re: [PATCH v1 3/3] ARM: dts: bcm2711: Enable GENET support for the RPi4

Hi Florian,

Am 11.10.19 um 21:31 schrieb Florian Fainelli:

Quoted 30 lines · On 10/11/19 11:48 AM, [email protected] wrote:
Quoted 18 lines

From: Matthias Brugger <[email protected]>

Enable Gigabit Ethernet support on the Raspberry Pi 4 Model B.

---

 arch/arm/boot/dts/bcm2711-rpi-4-b.dts | 22 ++++++++++++++++++++++
 arch/arm/boot/dts/bcm2711.dtsi        | 18 ++++++++++++++++++
 2 files changed, 40 insertions(+)
arch/arm/boot/dts/bcm2711-rpi-4-b.dts +0 -0
index cccc1ccd19be..958553d62670 100644

arch/arm/boot/dts/bcm2711-rpi-4-b.dts line 102

Quoted 6 lines
arch/arm/boot/dts/bcm2711-rpi-4-b.dts +3 -0
@@ -97,6 +97,28 @@
 	status = "okay";
 };
 
+&genet {
+	phy-handle = <&phy1>;
+	phy-mode = "rgmii";

Can you check that things still work against David Miller's net-next? Tree, in particular the BCM54213PE PHY might be matched by the BCM54210E entry in drivers/net/phy/broadcom.c and I just fixed an issue in how RGMII delays were configured:

Quoted 4 lines

https://git.kernel.org/pub/scm/linux/kernel/git/davem/net-next.git/commit/?id=fea7fda7f50a6059220f83251e70709e45cc8040

This might require you to change the 'phy-mode' property to what is appropriate.

i applied your changes, kept the phy-mode above and the interfaces cames up. But there is a lot of packet loss using ping. After applying this downstream patch [1] the packet loss doesn't occur.

Is the packet loss a possible cause of the wrong phy-mode and mentioned patch only a workaround?

[1] -
https://github.com/raspberrypi/linux/commit/360c8e98883f9cd075564be8a7fc25ac0785dee4

#8 lore raw

Re: [PATCH v1 3/3] ARM: dts: bcm2711: Enable GENET support for the RPi4

Quoted 33 lines · Stefan Wahren

Hi Florian,

Am 11.10.19 um 21:31 schrieb Florian Fainelli:

Quoted 30 lines · On 10/11/19 11:48 AM, [email protected] wrote:
Quoted 18 lines

From: Matthias Brugger <[email protected]>

Enable Gigabit Ethernet support on the Raspberry Pi 4 Model B.

---

 arch/arm/boot/dts/bcm2711-rpi-4-b.dts | 22 ++++++++++++++++++++++
 arch/arm/boot/dts/bcm2711.dtsi        | 18 ++++++++++++++++++
 2 files changed, 40 insertions(+)
arch/arm/boot/dts/bcm2711-rpi-4-b.dts +0 -0
index cccc1ccd19be..958553d62670 100644

arch/arm/boot/dts/bcm2711-rpi-4-b.dts line 102

Quoted 6 lines
arch/arm/boot/dts/bcm2711-rpi-4-b.dts +3 -0
@@ -97,6 +97,28 @@
 	status = "okay";
 };
 
+&genet {
+	phy-handle = <&phy1>;
+	phy-mode = "rgmii";

Can you check that things still work against David Miller's net-next? Tree, in particular the BCM54213PE PHY might be matched by the BCM54210E entry in drivers/net/phy/broadcom.c and I just fixed an issue in how RGMII delays were configured:

Quoted 8 lines
Quoted 4 lines

https://git.kernel.org/pub/scm/linux/kernel/git/davem/net-next.git/commit/?id=fea7fda7f50a6059220f83251e70709e45cc8040

This might require you to change the 'phy-mode' property to what is appropriate.

i applied your changes, kept the phy-mode above and the interfaces cames up. But there is a lot of packet loss using ping. After applying this downstream patch [1] the packet loss doesn't occur.

Packet loss is symptomatic of a mis-configured RGMII interface between the MAC and the PHY.

Quoted 3 lines

Is the packet loss a possible cause of the wrong phy-mode and mentioned patch only a workaround?

The patch at [1] is not doing much with respect to RGMII delays, so it will just keep whatever was configured prior to Linux taking over the PHY. The addition of the BCM54213PE entry makes use of the bcm54xx_config_init() callback, which does not call bcm54xx_config_clock_delay() for the BCM54213PE PHY model, which is why it "solves" the packet loss by preserving whatever was already configured.

I would suggest checking with the Pi folks whether 'rgmii' is really the right mode of operation here (that is, the PHY is not adding TXC or RXC delays at all), or it just happens to work by virtue of the PHY device defaulting to a certain mode *and* the PHY device driver in Linux not attempting to correctly change the RX/TX clock delays based on the phy_interface_t value (aka: maintain the status quo).

Thanks for checking!

Quoted 4 lines

[1] -
https://github.com/raspberrypi/linux/commit/360c8e98883f9cd075564be8a7fc25ac0785dee4

Signature · 1 line
Florian
#9 lore raw

Re: [PATCH v1 3/3] ARM: dts: bcm2711: Enable GENET support for the RPi4

In reply to Florian Fainelli

Hi Florian,

Am 11.10.19 um 21:31 schrieb Florian Fainelli:

Quoted 30 lines · On 10/11/19 11:48 AM, [email protected] wrote:
Quoted 18 lines

From: Matthias Brugger <[email protected]>

Enable Gigabit Ethernet support on the Raspberry Pi 4 Model B.

---

 arch/arm/boot/dts/bcm2711-rpi-4-b.dts | 22 ++++++++++++++++++++++
 arch/arm/boot/dts/bcm2711.dtsi        | 18 ++++++++++++++++++
 2 files changed, 40 insertions(+)
arch/arm/boot/dts/bcm2711-rpi-4-b.dts +0 -0
index cccc1ccd19be..958553d62670 100644

arch/arm/boot/dts/bcm2711-rpi-4-b.dts line 102

Quoted 6 lines
arch/arm/boot/dts/bcm2711-rpi-4-b.dts +3 -0
@@ -97,6 +97,28 @@
 	status = "okay";
 };
 
+&genet {
+	phy-handle = <&phy1>;
+	phy-mode = "rgmii";

Can you check that things still work against David Miller's net-next? Tree, in particular the BCM54213PE PHY might be matched by the BCM54210E entry in drivers/net/phy/broadcom.c and I just fixed an issue in how RGMII delays were configured:

Quoted 4 lines

https://git.kernel.org/pub/scm/linux/kernel/git/davem/net-next.git/commit/?id=fea7fda7f50a6059220f83251e70709e45cc8040

This might require you to change the 'phy-mode' property to what is appropriate.

since i didn't get a reply from the Pi folks yet, i tried all the other rgmii variants on top of this branch [1].

But none of them worked.

In case of "rgmii-id" i'm getting:

unknown phy mode: 9

and for "rgmii-rxid":

unknown phy mode: 10

In those cases the PHY didn't even probe. Did i missed something?

[1] - https://github.com/lategoodbye/rpi-zero/commits/bcm2711-genet

#10 lore raw

Re: [PATCH v1 3/3] ARM: dts: bcm2711: Enable GENET support for the RPi4

Quoted 45 lines · Stefan Wahren

Hi Florian,

Am 11.10.19 um 21:31 schrieb Florian Fainelli:

Quoted 30 lines · On 10/11/19 11:48 AM, [email protected] wrote:
Quoted 18 lines

From: Matthias Brugger <[email protected]>

Enable Gigabit Ethernet support on the Raspberry Pi 4 Model B.

---

 arch/arm/boot/dts/bcm2711-rpi-4-b.dts | 22 ++++++++++++++++++++++
 arch/arm/boot/dts/bcm2711.dtsi        | 18 ++++++++++++++++++
 2 files changed, 40 insertions(+)
arch/arm/boot/dts/bcm2711-rpi-4-b.dts +0 -0
index cccc1ccd19be..958553d62670 100644

arch/arm/boot/dts/bcm2711-rpi-4-b.dts line 102

Quoted 6 lines
arch/arm/boot/dts/bcm2711-rpi-4-b.dts +3 -0
@@ -97,6 +97,28 @@
 	status = "okay";
 };
 
+&genet {
+	phy-handle = <&phy1>;
+	phy-mode = "rgmii";

Can you check that things still work against David Miller's net-next? Tree, in particular the BCM54213PE PHY might be matched by the BCM54210E entry in drivers/net/phy/broadcom.c and I just fixed an issue in how RGMII delays were configured:

Quoted 4 lines

https://git.kernel.org/pub/scm/linux/kernel/git/davem/net-next.git/commit/?id=fea7fda7f50a6059220f83251e70709e45cc8040

This might require you to change the 'phy-mode' property to what is appropriate.

since i didn't get a reply from the Pi folks yet, i tried all the other rgmii variants on top of this branch [1].

But none of them worked.

In case of "rgmii-id" i'm getting:

Quoted 7 lines

unknown phy mode: 9

and for "rgmii-rxid":

unknown phy mode: 10

In those cases the PHY didn't even probe. Did i missed something?

Well the GENET driver can only deal with two specific RGMII modes, so you would need something like this:

(quoted diff) +0 -0

b/drivers/net/ethernet/broadcom/genet/bcmmii.c
index 970e478a9017..a12656cccf89 100644

drivers/net/ethernet/broadcom/genet/bcmmii.c +0 -0
@@ -241,6 +241,8 @@ int bcmgenet_mii_config(struct net_device *dev, bool

init)

                id_mode_dis = BIT(16);
                /* fall through */
        case PHY_INTERFACE_MODE_RGMII_TXID:
+       case PHY_INTERFACE_MODE_RGMII_RXID:
+       case PHY_INTERFACE_MODE_RGMII_ID:
                if (id_mode_dis)
                        phy_name = "external RGMII (no delay)";
                else
Quoted 3 lines

[1] - https://github.com/lategoodbye/rpi-zero/commits/bcm2711-genet

Signature · 6 lines
Florian

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

Re: [PATCH v1 3/3] ARM: dts: bcm2711: Enable GENET support for the RPi4

In reply to [email protected]

Am 11.10.19 um 20:48 schrieb [email protected]:

Quoted 4 lines

From: Matthias Brugger <[email protected]>

Enable Gigabit Ethernet support on the Raspberry Pi 4 Model B.

I asked some questions about genet to the RPi guys [1] which are relevant to this patch (missing clocks and interrupt, MAC address assignment) but i didn't get an answer yet.

During my tests with a similiar patch series i noticed that the driver won't probe without a MAC address.

How does it get into DT (via U-Boot)?

[1] -
https://github.com/raspberrypi/linux/issues/3101#issuecomment-534665860

#12 lore raw

Re: [PATCH v1 3/3] ARM: dts: bcm2711: Enable GENET support for the RPi4

Quoted 6 lines · Stefan Wahren

Am 11.10.19 um 20:48 schrieb [email protected]:

Quoted 4 lines

From: Matthias Brugger <[email protected]>

Enable Gigabit Ethernet support on the Raspberry Pi 4 Model B.

Quoted 6 lines

I asked some questions about genet to the RPi guys [1] which are relevant to this patch (missing clocks and interrupt, MAC address assignment) but i didn't get an answer yet.

During my tests with a similiar patch series i noticed that the driver won't probe without a MAC address.

Yes, that is true, we have an internal patch for that that we just recently did, let me submit it so this does not block you.

Quoted 6 lines

How does it get into DT (via U-Boot)?

[1] -
https://github.com/raspberrypi/linux/issues/3101#issuecomment-534665860

Signature · 1 line
Florian
#13 lore raw

Re: [PATCH v1 3/3] ARM: dts: bcm2711: Enable GENET support for the RPi4

In reply to Stefan Wahren

Am 12.10.19 um 01:09 schrieb Stefan Wahren:

Quoted 5 lines

Am 11.10.19 um 20:48 schrieb [email protected]:

Quoted 4 lines

From: Matthias Brugger <[email protected]>

Enable Gigabit Ethernet support on the Raspberry Pi 4 Model B.

Quoted 8 lines

I asked some questions about genet to the RPi guys [1] which are relevant to this patch (missing clocks and interrupt, MAC address assignment) but i didn't get an answer yet.

During my tests with a similiar patch series i noticed that the driver won't probe without a MAC address.

How does it get into DT (via U-Boot)?

Okay, the bootloader uses the ethernet0 alias for the genet DT node. So this should also be added to the RPi 4 DTS. But i consider the MAC randomize patch still helpful.

Quoted 4 lines

[1] -
https://github.com/raspberrypi/linux/issues/3101#issuecomment-534665860