[RFC-PATCH] net: stmmac: Add KR port support.

[RFC-PATCH] net: stmmac: Add KR port support.

[Daniel Walker]

From: Balamurugan Selvarajan <balamsel@cisco.com>

For KR port the mii interface is a chip-to-chip
interface without a mechanical connector. So PHY
inits are not applicable. In this case MAC is
configured to operate at forced speed(1000Mbps)
and full duplex. Modified driver to accommodate
PHY and NON-PHY mode.

Cc: xe-linux-external@cisco.com
Signed-off-by: Balamurugan Selvarajan <balamsel@cisco.com>
Signed-off-by: Daniel Walker <danielwa@cisco.com>
---
 drivers/net/ethernet/stmicro/stmmac/common.h  |  2 +
 .../net/ethernet/stmicro/stmmac/stmmac_main.c | 89 +++++++++++++------
 2 files changed, 65 insertions(+), 26 deletions(-)

diff --git a/drivers/net/ethernet/stmicro/stmmac/common.h b/drivers/net/ethernet/stmicro/stmmac/common.h
index 5fecc83f175b..6458ce5ec0bd 100644
--- a/drivers/net/ethernet/stmicro/stmmac/common.h
+++ b/drivers/net/ethernet/stmicro/stmmac/common.h
@@ -435,6 +435,8 @@ struct dma_features {
 #define MAC_CTRL_REG		0x00000000	/* MAC Control */
 #define MAC_ENABLE_TX		0x00000008	/* Transmitter Enable */
 #define MAC_ENABLE_RX		0x00000004	/* Receiver Enable */
+#define MAC_FULL_DUPLEX		0x00000800
+#define MAC_PORT_SELECT		0x00008000
 
 /* Default LPI timers */
 #define STMMAC_DEFAULT_LIT_LS	0x3E8
diff --git a/drivers/net/ethernet/stmicro/stmmac/stmmac_main.c b/drivers/net/ethernet/stmicro/stmmac/stmmac_main.c
index 7b8404a21544..0b31aae65f3a 100644
--- a/drivers/net/ethernet/stmicro/stmmac/stmmac_main.c
+++ b/drivers/net/ethernet/stmicro/stmmac/stmmac_main.c
@@ -3212,6 +3212,7 @@ static int stmmac_hw_setup(struct net_device *dev, bool init_ptp)
 	bool sph_en;
 	u32 chan;
 	int ret;
+	int speed;
 
 	/* DMA initialization and SW reset */
 	ret = stmmac_init_dma_engine(priv);
@@ -3226,7 +3227,9 @@ static int stmmac_hw_setup(struct net_device *dev, bool init_ptp)
 
 	/* PS and related bits will be programmed according to the speed */
 	if (priv->hw->pcs) {
-		int speed = priv->plat->mac_port_sel_speed;
+		speed = priv->plat->mac_port_sel_speed;
+		if (priv->plat->phy_interface == PHY_INTERFACE_MODE_NA)
+			speed = SPEED_1000;
 
 		if ((speed == SPEED_10) || (speed == SPEED_100) ||
 		    (speed == SPEED_1000)) {
@@ -3256,6 +3259,17 @@ static int stmmac_hw_setup(struct net_device *dev, bool init_ptp)
 	/* Enable the MAC Rx/Tx */
 	stmmac_mac_set(priv, priv->ioaddr, true);
 
+	if (priv->plat->phy_interface == PHY_INTERFACE_MODE_NA) {
+		/*
+		 * Force MAC PORT SPEED to 1000Mbps and Full Duplex.
+		 */
+		u32 ctrl = readl(priv->ioaddr + MAC_CTRL_REG);
+		ctrl |= MAC_FULL_DUPLEX;
+		ctrl &= ~(MAC_PORT_SELECT);
+		writel(ctrl, priv->ioaddr + MAC_CTRL_REG);
+	}
+
+
 	/* Set the HW DMA mode and the COE */
 	stmmac_dma_operation_mode(priv);
 
@@ -3291,8 +3305,14 @@ static int stmmac_hw_setup(struct net_device *dev, bool init_ptp)
 		}
 	}
 
-	if (priv->hw->pcs)
-		stmmac_pcs_ctrl_ane(priv, priv->ioaddr, 1, priv->hw->ps, 0);
+	if (priv->hw->pcs) {
+		if (priv->plat->phy_interface != PHY_INTERFACE_MODE_NA) {
+			stmmac_pcs_ctrl_ane(priv, priv->ioaddr, 1, priv->hw->ps, 0);
+		} else {
+			/* Disable Autoneg */
+			writel(0x0, priv->ioaddr + GMAC_PCS_BASE);
+		}
+	}
 
 	/* set TX and RX rings length */
 	stmmac_set_rings_length(priv);
@@ -3644,12 +3664,14 @@ int stmmac_open(struct net_device *dev)
 	    priv->hw->pcs != STMMAC_PCS_RTBI &&
 	    (!priv->hw->xpcs ||
 	     xpcs_get_an_mode(priv->hw->xpcs, mode) != DW_AN_C73)) {
-		ret = stmmac_init_phy(dev);
-		if (ret) {
-			netdev_err(priv->dev,
-				   "%s: Cannot attach to PHY (error: %d)\n",
-				   __func__, ret);
-			goto init_phy_error;
+		if (priv->plat->phy_interface != PHY_INTERFACE_MODE_NA) {
+			ret = stmmac_init_phy(dev);
+			if (ret) {
+				netdev_err(priv->dev,
+					   "%s: Cannot attach to PHY (error: %d)\n",
+					   __func__, ret);
+				goto init_phy_error;
+			}
 		}
 	}
 
@@ -3705,7 +3727,8 @@ int stmmac_open(struct net_device *dev)
 
 	stmmac_init_coalesce(priv);
 
-	phylink_start(priv->phylink);
+	if (priv->plat->phy_interface != PHY_INTERFACE_MODE_NA)
+		phylink_start(priv->phylink);
 	/* We may have called phylink_speed_down before */
 	phylink_speed_up(priv->phylink);
 
@@ -3716,10 +3739,14 @@ int stmmac_open(struct net_device *dev)
 	stmmac_enable_all_queues(priv);
 	netif_tx_start_all_queues(priv->dev);
 
+	if (priv->plat->phy_interface == PHY_INTERFACE_MODE_NA)
+		netif_carrier_on(dev);
+
 	return 0;
 
 irq_error:
-	phylink_stop(priv->phylink);
+	if (priv->plat->phy_interface != PHY_INTERFACE_MODE_NA)
+		phylink_stop(priv->phylink);
 
 	for (chan = 0; chan < priv->plat->tx_queues_to_use; chan++)
 		hrtimer_cancel(&priv->tx_queue[chan].txtimer);
@@ -3728,7 +3755,8 @@ int stmmac_open(struct net_device *dev)
 init_error:
 	free_dma_desc_resources(priv);
 dma_desc_error:
-	phylink_disconnect_phy(priv->phylink);
+	if (priv->plat->phy_interface != PHY_INTERFACE_MODE_NA)
+		phylink_disconnect_phy(priv->phylink);
 init_phy_error:
 	pm_runtime_put(priv->device);
 	return ret;
@@ -3758,8 +3786,10 @@ int stmmac_release(struct net_device *dev)
 	if (device_may_wakeup(priv->device))
 		phylink_speed_down(priv->phylink, false);
 	/* Stop and disconnect the PHY */
-	phylink_stop(priv->phylink);
-	phylink_disconnect_phy(priv->phylink);
+	if (priv->plat->phy_interface != PHY_INTERFACE_MODE_NA) {
+		phylink_stop(priv->phylink);
+		phylink_disconnect_phy(priv->phylink);
+	}
 
 	stmmac_disable_all_queues(priv);
 
@@ -6986,12 +7016,15 @@ int stmmac_dvr_probe(struct device *device,
 	if (priv->hw->pcs != STMMAC_PCS_TBI &&
 	    priv->hw->pcs != STMMAC_PCS_RTBI) {
 		/* MDIO bus Registration */
-		ret = stmmac_mdio_register(ndev);
-		if (ret < 0) {
-			dev_err(priv->device,
-				"%s: MDIO bus (id: %d) registration failed",
-				__func__, priv->plat->bus_id);
-			goto error_mdio_register;
+
+		if (priv->plat->phy_interface != PHY_INTERFACE_MODE_NA) {
+			ret = stmmac_mdio_register(ndev);
+			if (ret < 0) {
+				dev_err(priv->device,
+					"%s: MDIO bus (id: %d) registration failed",
+					__func__, priv->plat->bus_id);
+				goto error_mdio_register;
+			}
 		}
 	}
 
@@ -7004,10 +7037,12 @@ int stmmac_dvr_probe(struct device *device,
 			goto error_xpcs_setup;
 	}
 
-	ret = stmmac_phy_setup(priv);
-	if (ret) {
-		netdev_err(ndev, "failed to setup phy (%d)\n", ret);
-		goto error_phy_setup;
+	if (priv->plat->phy_interface != PHY_INTERFACE_MODE_NA) {
+		ret = stmmac_phy_setup(priv);
+		if (ret) {
+			netdev_err(ndev, "failed to setup phy (%d)\n", ret);
+			goto error_phy_setup;
+		}
 	}
 
 	ret = register_netdev(ndev);
@@ -7039,11 +7074,13 @@ int stmmac_dvr_probe(struct device *device,
 error_serdes_powerup:
 	unregister_netdev(ndev);
 error_netdev_register:
-	phylink_destroy(priv->phylink);
+	if (priv->plat->phy_interface != PHY_INTERFACE_MODE_NA)
+		phylink_destroy(priv->phylink);
 error_xpcs_setup:
 error_phy_setup:
 	if (priv->hw->pcs != STMMAC_PCS_TBI &&
-	    priv->hw->pcs != STMMAC_PCS_RTBI)
+	    priv->hw->pcs != STMMAC_PCS_RTBI &&
+	    priv->plat->phy_interface != PHY_INTERFACE_MODE_NA)
 		stmmac_mdio_unregister(ndev);
 error_mdio_register:
 	stmmac_napi_del(ndev);
-- 
2.25.1

Re: [RFC-PATCH] net: stmmac: Add KR port support.

[Florian Fainelli]

On 7/29/21 4:44 PM, Daniel Walker wrote:
> From: Balamurugan Selvarajan <balamsel@cisco.com>
> 
> For KR port the mii interface is a chip-to-chip
> interface without a mechanical connector. So PHY
> inits are not applicable. In this case MAC is
> configured to operate at forced speed(1000Mbps)
> and full duplex. Modified driver to accommodate
> PHY and NON-PHY mode.
> 
> Cc: xe-linux-external@cisco.com
> Signed-off-by: Balamurugan Selvarajan <balamsel@cisco.com>
> Signed-off-by: Daniel Walker <danielwa@cisco.com>

You are not adding KR support per-se, you are just hacking the driver so
it is happy with an unspecified phy_interface_t value and assuming
1000Mbits/sec, this is not going to work.

Just add KR/backplane properly or use a fixed-link property hardcoded
for 1000Mbits/sec and be done with it with no hacking, which would be
just way better than what is proposed here.
-- 
Florian

Re: [RFC-PATCH] net: stmmac: Add KR port support.

[Andrew Lunn]

On Thu, Jul 29, 2021 at 04:44:42PM -0700, Daniel Walker wrote:
> From: Balamurugan Selvarajan <balamsel@cisco.com>
> 
> For KR port the mii interface is a chip-to-chip
> interface without a mechanical connector. So PHY
> inits are not applicable. In this case MAC is
> configured to operate at forced speed(1000Mbps)
> and full duplex. Modified driver to accommodate
> PHY and NON-PHY mode.

I agree with Florian here. Look at all the in kernel examples of a SoC
MAC connected to an Ethernet switch. Some use rgmii, others 1000BaseX
or higher. But they all follow the same scheme, and don't need
invasive MAC driver changes.

    Andrew

Re: [RFC-PATCH] net: stmmac: Add KR port support.

[Daniel Walker]

On Fri, Jul 30, 2021 at 05:01:44AM +0200, Andrew Lunn wrote:
> On Thu, Jul 29, 2021 at 04:44:42PM -0700, Daniel Walker wrote:
> > From: Balamurugan Selvarajan <balamsel@cisco.com>
> > 
> > For KR port the mii interface is a chip-to-chip
> > interface without a mechanical connector. So PHY
> > inits are not applicable. In this case MAC is
> > configured to operate at forced speed(1000Mbps)
> > and full duplex. Modified driver to accommodate
> > PHY and NON-PHY mode.
> 
> I agree with Florian here. Look at all the in kernel examples of a SoC
> MAC connected to an Ethernet switch. Some use rgmii, others 1000BaseX
> or higher. But they all follow the same scheme, and don't need
> invasive MAC driver changes.


Can you provide the examples which you looked at ?

Daniel

Re: [RFC-PATCH] net: stmmac: Add KR port support.

[Russell King (Oracle)]

On Thu, Jul 29, 2021 at 04:44:42PM -0700, Daniel Walker wrote:
> From: Balamurugan Selvarajan <balamsel@cisco.com>
> 
> For KR port the mii interface is a chip-to-chip
> interface without a mechanical connector. So PHY
> inits are not applicable. In this case MAC is
> configured to operate at forced speed(1000Mbps)
> and full duplex. Modified driver to accommodate
> PHY and NON-PHY mode.

Can we clarify exactly what you are talking about here. What does
"KR port" refer to? What protocol is spoken by this port? Is it
1000BASE-KX (1000BASE-X over backplane)? Does it include 10GBASE-KR?

Thanks.

-- 
RMK's Patch system: https://www.armlinux.org.uk/developer/patches/
FTTP is here! 40Mbps down 10Mbps up. Decent connectivity at last!

Re: [RFC-PATCH] net: stmmac: Add KR port support.

[Andrew Lunn]

On Fri, Jul 30, 2021 at 07:48:30AM -0700, Daniel Walker wrote:
> On Fri, Jul 30, 2021 at 05:01:44AM +0200, Andrew Lunn wrote:
> > On Thu, Jul 29, 2021 at 04:44:42PM -0700, Daniel Walker wrote:
> > > From: Balamurugan Selvarajan <balamsel@cisco.com>
> > > 
> > > For KR port the mii interface is a chip-to-chip
> > > interface without a mechanical connector. So PHY
> > > inits are not applicable. In this case MAC is
> > > configured to operate at forced speed(1000Mbps)
> > > and full duplex. Modified driver to accommodate
> > > PHY and NON-PHY mode.
> > 
> > I agree with Florian here. Look at all the in kernel examples of a SoC
> > MAC connected to an Ethernet switch. Some use rgmii, others 1000BaseX
> > or higher. But they all follow the same scheme, and don't need
> > invasive MAC driver changes.
> 
> 
> Can you provide the examples which you looked at ?

There are plenty of examples using Freescale FEC and Marvell Ethernet
switches:

arch/arm/boot/dts/vf610-zii-dev-rev-b.dts

Or the Mavell based

arch/arm/boot/dts/armada-xp-linksys-mamba.dts

	Andrew