| Message ID | E1tXhIJ-000me6-SA@rmk-PC.armlinux.org.uk (mailing list archive) |
|---|---|
| State | Superseded |
| Delegated to: | Netdev Maintainers |
| Headers | show |
| Series | [RFC,net-next] net: phylink: always do a major config when attaching a SFP PHY | expand |
Hi Eric, I'd like to get a tested-by from you before sending this for merging, as this should fix your reported issue. Thanks. On Tue, Jan 14, 2025 at 01:49:23PM +0000, Russell King (Oracle) wrote: > Background: https://lore.kernel.org/r/20250107123615.161095-1-ericwouds@gmail.com > > Since adding negotiation of in-band capabilities, it is no longer > sufficient to just look at the MLO_AN_xxx mode and PHY interface to > decide whether to do a major configuration, since the result now > depends on the capabilities of the attaching PHY. > > Always trigger a major configuration in this case. > > Reported-by: Eric Woudstra <ericwouds@gmail.com> > Signed-off-by: Russell King (Oracle) <rmk+kernel@armlinux.org.uk> > --- > This follows on from https://lore.kernel.org/r/Z4TbR93B-X8A8iHe@shell.armlinux.org.uk "net: phylink: fix PCS without autoneg" > > drivers/net/phy/phylink.c | 11 +++++------ > 1 file changed, 5 insertions(+), 6 deletions(-) > > diff --git a/drivers/net/phy/phylink.c b/drivers/net/phy/phylink.c > index ff0efb52189f..7d64182cf802 100644 > --- a/drivers/net/phy/phylink.c > +++ b/drivers/net/phy/phylink.c > @@ -3412,12 +3412,11 @@ static phy_interface_t phylink_choose_sfp_interface(struct phylink *pl, > return interface; > } > > -static void phylink_sfp_set_config(struct phylink *pl, > - unsigned long *supported, > - struct phylink_link_state *state) > +static void phylink_sfp_set_config(struct phylink *pl, unsigned long *supported, > + struct phylink_link_state *state, > + bool changed) > { > u8 mode = MLO_AN_INBAND; > - bool changed = false; > > phylink_dbg(pl, "requesting link mode %s/%s with support %*pb\n", > phylink_an_mode_str(mode), phy_modes(state->interface), > @@ -3494,7 +3493,7 @@ static int phylink_sfp_config_phy(struct phylink *pl, struct phy_device *phy) > > pl->link_port = pl->sfp_port; > > - phylink_sfp_set_config(pl, support, &config); > + phylink_sfp_set_config(pl, support, &config, true); > > return 0; > } > @@ -3569,7 +3568,7 @@ static int phylink_sfp_config_optical(struct phylink *pl) > > pl->link_port = pl->sfp_port; > > - phylink_sfp_set_config(pl, pl->sfp_support, &config); > + phylink_sfp_set_config(pl, pl->sfp_support, &config, false); > > return 0; > } > -- > 2.30.2 > >
On 1/16/25 12:19 AM, Russell King (Oracle) wrote: > Hi Eric, > > I'd like to get a tested-by from you before sending this for merging, > as this should fix your reported issue. Tested-by: Eric Woudstra <ericwouds@gmail.com> It works ok, thanks. The kernel log shows: [root@bpir3 ~]# dmesg | grep eth1 [ 2.515546] mtk_soc_eth 15100000.ethernet eth1: mediatek frame engine at 0xffff800082300000, irq 123 [ 41.406446] mtk_soc_eth 15100000.ethernet eth1: configuring for inband/2500base-x link mode [ 41.414819] mtk_soc_eth 15100000.ethernet eth1: major config, requested inband/2500base-x [ 41.422983] mtk_soc_eth 15100000.ethernet eth1: interface 2500base-x inband modes: pcs=01 phy=00 [ 41.431936] mtk_soc_eth 15100000.ethernet eth1: major config, active inband/inband,an-disabled/2500base-x [ 41.441523] mtk_soc_eth 15100000.ethernet eth1: phylink_mac_config: mode=inband/2500base-x/none adv=00,00000000,00008000,0000e240 pause=04 [ 42.370423] mtk_soc_eth 15100000.ethernet eth1: interface 2 (mii) rate match none supports 0-3,6-7,13-14 [ 42.380034] mtk_soc_eth 15100000.ethernet eth1: interface 3 (gmii) rate match none supports 0-3,5-7,13-14 [ 42.389698] mtk_soc_eth 15100000.ethernet eth1: interface 4 (sgmii) rate match none supports 0-3,5-7,13-14 [ 42.399441] mtk_soc_eth 15100000.ethernet eth1: interface 22 (1000base-x) rate match none supports 5-7,13-14 [ 42.409348] mtk_soc_eth 15100000.ethernet eth1: interface 23 (2500base-x) rate match none supports 6-7,13-14,47 [ 42.744938] mtk_soc_eth 15100000.ethernet eth1: PHY i2c:sfp-1:11 uses interfaces 4,23, validating 4,23 [ 42.754256] mtk_soc_eth 15100000.ethernet eth1: interface 4 (sgmii) rate match none supports 0-3,5-7,13-14 [ 42.800352] mtk_soc_eth 15100000.ethernet eth1: interface 23 (2500base-x) rate match none supports 6-7,13-14,47 [ 42.810541] mtk_soc_eth 15100000.ethernet eth1: PHY [i2c:sfp-1:11] driver [RTL8221B-VB-CG 2.5Gbps PHY (C45)] (irq=POLL) [ 42.821312] mtk_soc_eth 15100000.ethernet eth1: phy: 2500base-x setting supported 00,00000000,00008000,000060ef advertising 00,00000000,00008000,000060ef [ 42.835035] mtk_soc_eth 15100000.ethernet eth1: requesting link mode inband/2500base-x with support 00,00000000,00008000,000060ef [ 42.846675] mtk_soc_eth 15100000.ethernet eth1: major config, requested inband/2500base-x [ 42.854847] mtk_soc_eth 15100000.ethernet eth1: interface 2500base-x inband modes: pcs=01 phy=00 [ 42.863616] mtk_soc_eth 15100000.ethernet eth1: major config, active phy/outband/2500base-x [ 42.871966] mtk_soc_eth 15100000.ethernet eth1: phylink_mac_config: mode=phy/2500base-x/none adv=00,00000000,00008000,000060ef pause=04 [ 43.590342] mtk_soc_eth 15100000.ethernet eth1: phy link down 2500base-x/Unknown/Unknown/none/off [ 43.607350] brlan: port 5(eth1) entered blocking state [ 43.612481] brlan: port 5(eth1) entered disabled state [ 43.617707] mtk_soc_eth 15100000.ethernet eth1: entered allmulticast mode [ 43.624679] mtk_soc_eth 15100000.ethernet eth1: entered promiscuous mode [ 47.700338] mtk_soc_eth 15100000.ethernet eth1: phy link up 2500base-x/2.5Gbps/Full/none/rx/tx [ 47.700357] mtk_soc_eth 15100000.ethernet eth1: Link is Up - 2.5Gbps/Full - flow control rx/tx [ 47.717579] mtk_soc_eth 15100000.ethernet eth1: No phy led trigger registered for speed(2500) [ 47.734637] brlan: port 5(eth1) entered blocking state [ 47.739782] brlan: port 5(eth1) entered forwarding state
diff --git a/drivers/net/phy/phylink.c b/drivers/net/phy/phylink.c index ff0efb52189f..7d64182cf802 100644 --- a/drivers/net/phy/phylink.c +++ b/drivers/net/phy/phylink.c @@ -3412,12 +3412,11 @@ static phy_interface_t phylink_choose_sfp_interface(struct phylink *pl, return interface; } -static void phylink_sfp_set_config(struct phylink *pl, - unsigned long *supported, - struct phylink_link_state *state) +static void phylink_sfp_set_config(struct phylink *pl, unsigned long *supported, + struct phylink_link_state *state, + bool changed) { u8 mode = MLO_AN_INBAND; - bool changed = false; phylink_dbg(pl, "requesting link mode %s/%s with support %*pb\n", phylink_an_mode_str(mode), phy_modes(state->interface), @@ -3494,7 +3493,7 @@ static int phylink_sfp_config_phy(struct phylink *pl, struct phy_device *phy) pl->link_port = pl->sfp_port; - phylink_sfp_set_config(pl, support, &config); + phylink_sfp_set_config(pl, support, &config, true); return 0; } @@ -3569,7 +3568,7 @@ static int phylink_sfp_config_optical(struct phylink *pl) pl->link_port = pl->sfp_port; - phylink_sfp_set_config(pl, pl->sfp_support, &config); + phylink_sfp_set_config(pl, pl->sfp_support, &config, false); return 0; }
Background: https://lore.kernel.org/r/20250107123615.161095-1-ericwouds@gmail.com Since adding negotiation of in-band capabilities, it is no longer sufficient to just look at the MLO_AN_xxx mode and PHY interface to decide whether to do a major configuration, since the result now depends on the capabilities of the attaching PHY. Always trigger a major configuration in this case. Reported-by: Eric Woudstra <ericwouds@gmail.com> Signed-off-by: Russell King (Oracle) <rmk+kernel@armlinux.org.uk> --- This follows on from https://lore.kernel.org/r/Z4TbR93B-X8A8iHe@shell.armlinux.org.uk "net: phylink: fix PCS without autoneg" drivers/net/phy/phylink.c | 11 +++++------ 1 file changed, 5 insertions(+), 6 deletions(-)