| Message ID | 20240429-b4-for-netnext-mt7530-use-switch-mdio-bus-for-phy-muxing-v1-1-1f775983e155@arinc9.com (mailing list archive) |
|---|---|
| State | New |
| Headers | show |
| Series | [net-next] net: dsa: mt7530: detect PHY muxing when PHY is defined on switch MDIO bus | expand |
On Mon, Apr 29, 2024 at 12:46:43PM +0300, Arınç ÜNAL via B4 Relay wrote: > From: Arınç ÜNAL <arinc.unal@arinc9.com> > > Currently, the MT7530 DSA subdriver configures the MT7530 switch to provide > direct access to switch PHYs, meaning, the switch PHYs listen on the MDIO > bus the switch listens on. The PHY muxing feature makes use of this. > > This is problematic as the PHY may be probed before the switch is > initialised, in which case attaching the PHY will fail. > > Since commit 91374ba537bd ("net: dsa: mt7530: support OF-based registration > of switch MDIO bus"), we can describe the switch PHYs on the MDIO bus of > the switch on the device tree. Extend the check to detect PHY muxing when > the PHY is defined on the MDIO bus of the switch on the device tree. > > When the PHY is described this way, the switch will be initialised first, > then the switch MDIO bus will be registered. Only after these steps, the > PHY will be probed. Looking at the commit description and the patch, I'm not sure whether you really mean "probed" or whether you mean "attached". PHY drivers will bind to PHY devices when they are detected on the MDIO bus (either by scanning or instantiating firmware description) and the devices are registered. As each device is registered, the drivers for the bus type are scanned and any matches will have their probe function called. This happens outside of any control of the DSA driver if the DSA device is on the same MDIO bus. This is separate from the process of looking up a PHY and attaching the PHY. So, I think there is probably a terminology issue with the patch description. I suspect you don't mean "probing" as in phy_probe() being called. Looking at the code, it looks like the driver is making decisions based on how PHYs are connected to the ethernet MACs in the device tree, and you're making decisions based on that. I wouldn't call that "probing" a PHY.
On 29.04.2024 12:58, Russell King (Oracle) wrote: > On Mon, Apr 29, 2024 at 12:46:43PM +0300, Arınç ÜNAL via B4 Relay wrote: >> From: Arınç ÜNAL <arinc.unal@arinc9.com> >> >> Currently, the MT7530 DSA subdriver configures the MT7530 switch to provide >> direct access to switch PHYs, meaning, the switch PHYs listen on the MDIO >> bus the switch listens on. The PHY muxing feature makes use of this. >> >> This is problematic as the PHY may be probed before the switch is >> initialised, in which case attaching the PHY will fail. >> >> Since commit 91374ba537bd ("net: dsa: mt7530: support OF-based registration >> of switch MDIO bus"), we can describe the switch PHYs on the MDIO bus of >> the switch on the device tree. Extend the check to detect PHY muxing when >> the PHY is defined on the MDIO bus of the switch on the device tree. >> >> When the PHY is described this way, the switch will be initialised first, >> then the switch MDIO bus will be registered. Only after these steps, the >> PHY will be probed. > > Looking at the commit description and the patch, I'm not sure whether > you really mean "probed" or whether you mean "attached". > > PHY drivers will bind to PHY devices when they are detected on the MDIO > bus (either by scanning or instantiating firmware description) and the > devices are registered. As each device is registered, the drivers for > the bus type are scanned and any matches will have their probe function > called. This happens outside of any control of the DSA driver if the > DSA device is on the same MDIO bus. > > This is separate from the process of looking up a PHY and attaching the > PHY. > > So, I think there is probably a terminology issue with the patch > description. I suspect you don't mean "probing" as in phy_probe() being > called. Looking at the code, it looks like the driver is making > decisions based on how PHYs are connected to the ethernet MACs in the > device tree, and you're making decisions based on that. I wouldn't call > that "probing" a PHY. Yes, I meant to say "looking up a PHY and attaching the PHY" by probing a PHY. I'll adjust the patch log to below, thanks. Currently, the MT7530 DSA subdriver configures the MT7530 switch to provide direct access to switch PHYs, meaning, the switch PHYs listen on the MDIO bus the switch listens on. The PHY muxing feature makes use of this. This is problematic as the PHY may be attached before the switch is initialised, in which case, the PHY will fail to be attached. Since commit 91374ba537bd ("net: dsa: mt7530: support OF-based registration of switch MDIO bus"), we can describe the switch PHYs on the MDIO bus of the switch on the device tree. Extend the check to detect PHY muxing when the PHY is defined on the MDIO bus of the switch on the device tree. When the PHY is described this way, the switch will be initialised first, then the switch MDIO bus will be registered. Only after these steps, the PHY will be attached. Arınç
diff --git a/drivers/net/dsa/mt7530.c b/drivers/net/dsa/mt7530.c index 2b9f904a98f0..6cf21c9d523b 100644 --- a/drivers/net/dsa/mt7530.c +++ b/drivers/net/dsa/mt7530.c @@ -2483,7 +2483,8 @@ mt7530_setup(struct dsa_switch *ds) if (!phy_node) continue; - if (phy_node->parent == priv->dev->of_node->parent) { + if (phy_node->parent == priv->dev->of_node->parent || + phy_node->parent->parent == priv->dev->of_node) { ret = of_get_phy_mode(mac_np, &interface); if (ret && ret != -ENODEV) { of_node_put(mac_np);