| Message ID | 20230326140818.246575-3-arinc.unal@arinc9.com (mailing list archive) |
|---|---|
| State | New, archived |
| Headers | show |
| Series | net: dsa: mt7530: fix port 5 phylink, phy muxing, and port 6 | expand |
On Sun, Mar 26, 2023 at 05:08:13PM +0300, arinc9.unal@gmail.com wrote: > From: Arınç ÜNAL <arinc.unal@arinc9.com> > > There're two call paths for setting up port 5: > > mt7530_setup() > -> mt7530_setup_port5() > > mt753x_phylink_mac_config() > -> mt753x_mac_config() > -> mt7530_mac_config() > -> mt7530_setup_port5() > > The first call path is supposed to run when phy muxing is being used. In > this case, port 5 is somewhat of a hidden port. It won't be defined on the > devicetree so phylink can't be used to manage the port. > > The second call path used to call mt7530_setup_port5() directly under case > 5 on mt7530_phylink_mac_config() before it was moved to mt7530_mac_config() > with 88bdef8be9f6 ("net: dsa: mt7530: Extend device data ready for adding a > new hardware"). mt7530_setup_port5() will never run through this call path > because the current code on mt7530_setup() bypasses phylink for all cases > of port 5. > > Leave it to phylink if port 5 is used as a CPU port or a user port. For the > cases of phy muxing or the port being disabled, call mt7530_setup_port5() > directly from mt7530_setup_port5() without involving phylink. You probably don't mean "call X() from X()" (that would make it recursive), but maybe from mt7530_setup(). But it was already called from mt7530_setup(), so I don't understand what is being transmitted here... > > Move setting the interface and P5_DISABLED mode to a more specific > location. They're supposed to be overwritten if phy muxing is detected. > > Add comments which explain the process. > > Fixes: 38f790a80560 ("net: dsa: mt7530: Add support for port 5") > Tested-by: Arınç ÜNAL <arinc.unal@arinc9.com> > Signed-off-by: Arınç ÜNAL <arinc.unal@arinc9.com> > --- Sorry, I didn't understand... so what was the problem, and how does the movement of the mt7530_setup_port5() call that isn't under phylink solve that problem?
On 27.03.2023 21:49, Vladimir Oltean wrote: > On Sun, Mar 26, 2023 at 05:08:13PM +0300, arinc9.unal@gmail.com wrote: >> From: Arınç ÜNAL <arinc.unal@arinc9.com> >> >> There're two call paths for setting up port 5: >> >> mt7530_setup() >> -> mt7530_setup_port5() >> >> mt753x_phylink_mac_config() >> -> mt753x_mac_config() >> -> mt7530_mac_config() >> -> mt7530_setup_port5() >> >> The first call path is supposed to run when phy muxing is being used. In >> this case, port 5 is somewhat of a hidden port. It won't be defined on the >> devicetree so phylink can't be used to manage the port. >> >> The second call path used to call mt7530_setup_port5() directly under case >> 5 on mt7530_phylink_mac_config() before it was moved to mt7530_mac_config() >> with 88bdef8be9f6 ("net: dsa: mt7530: Extend device data ready for adding a >> new hardware"). mt7530_setup_port5() will never run through this call path >> because the current code on mt7530_setup() bypasses phylink for all cases >> of port 5. >> >> Leave it to phylink if port 5 is used as a CPU port or a user port. For the >> cases of phy muxing or the port being disabled, call mt7530_setup_port5() >> directly from mt7530_setup_port5() without involving phylink. > > You probably don't mean "call X() from X()" (that would make it recursive), > but maybe from mt7530_setup(). But it was already called from mt7530_setup(), > so I don't understand what is being transmitted here... Oops, I meant to say call mt7530_setup_port5() directly from mt7530_setup() without involving phylink. Will fix. > >> >> Move setting the interface and P5_DISABLED mode to a more specific >> location. They're supposed to be overwritten if phy muxing is detected. >> >> Add comments which explain the process. >> >> Fixes: 38f790a80560 ("net: dsa: mt7530: Add support for port 5") >> Tested-by: Arınç ÜNAL <arinc.unal@arinc9.com> >> Signed-off-by: Arınç ÜNAL <arinc.unal@arinc9.com> >> --- > > Sorry, I didn't understand... so what was the problem, and how does the > movement of the mt7530_setup_port5() call that isn't under phylink solve > that problem? Port 5 being used as a CPU or user port was being set up from mt7530_setup() instead of using phylink. This patch fixes that. Arınç
diff --git a/drivers/net/dsa/mt7530.c b/drivers/net/dsa/mt7530.c index 62a4b899a961..eba356249ada 100644 --- a/drivers/net/dsa/mt7530.c +++ b/drivers/net/dsa/mt7530.c @@ -2247,16 +2247,18 @@ mt7530_setup(struct dsa_switch *ds) return ret; /* Setup port 5 */ - priv->p5_intf_sel = P5_DISABLED; - interface = PHY_INTERFACE_MODE_NA; - if (!dsa_is_unused_port(ds, 5)) { + /* Set the interface selection of port 5 to GMAC5 when it's used + * as a CPU port or a user port. Let phylink handle the rest. + */ priv->p5_intf_sel = P5_INTF_SEL_GMAC5; - ret = of_get_phy_mode(dsa_to_port(ds, 5)->dn, &interface); - if (ret && ret != -ENODEV) - return ret; } else { - /* Scan the ethernet nodes. look for GMAC1, lookup used phy */ + /* Scan the ethernet nodes. Look for GMAC1, lookup the used phy. + * Determine if phy muxing is defined and which phy to mux. + */ + priv->p5_intf_sel = P5_DISABLED; + interface = PHY_INTERFACE_MODE_NA; + for_each_child_of_node(dn, mac_np) { if (!of_device_is_compatible(mac_np, "mediatek,eth-mac")) @@ -2287,6 +2289,8 @@ mt7530_setup(struct dsa_switch *ds) of_node_put(phy_node); break; } + + mt7530_setup_port5(ds, interface); } #ifdef CONFIG_GPIOLIB @@ -2297,8 +2301,6 @@ mt7530_setup(struct dsa_switch *ds) } #endif /* CONFIG_GPIOLIB */ - mt7530_setup_port5(ds, interface); - /* Flush the FDB table */ ret = mt7530_fdb_cmd(priv, MT7530_FDB_FLUSH, NULL); if (ret < 0)