From 80662f4fd4771bc9c7cc4abdfbe866ebd1179621 Mon Sep 17 00:00:00 2001 From: "Russell King (Oracle)" Date: Tue, 23 Nov 2021 16:44:02 +0100 Subject: net: phylink: Force link down and retrigger resolve on interface change MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit On PHY state change the phylink_resolve() function can read stale information from the MAC and report incorrect link speed and duplex to the kernel message log. Example with a Marvell 88X3310 PHY connected to a SerDes port on Marvell 88E6393X switch: - PHY driver triggers state change due to PHY interface mode being changed from 10gbase-r to 2500base-x due to copper change in speed from 10Gbps to 2.5Gbps, but the PHY itself either hasn't yet changed its interface to the host, or the interrupt about loss of SerDes link hadn't arrived yet (there can be a delay of several milliseconds for this), so we still think that the 10gbase-r mode is up - phylink_resolve() - phylink_mac_pcs_get_state() - this fills in speed=10g link=up - interface mode is updated to 2500base-x but speed is left at 10Gbps - phylink_major_config() - interface is changed to 2500base-x - phylink_link_up() - mv88e6xxx_mac_link_up() - .port_set_speed_duplex() - speed is set to 10Gbps - reports "Link is Up - 10Gbps/Full" to dmesg Afterwards when the interrupt finally arrives for mv88e6xxx, another resolve is forced in which we get the correct speed from phylink_mac_pcs_get_state(), but since the interface is not being changed anymore, we don't call phylink_major_config() but only phylink_mac_config(), which does not set speed/duplex anymore. To fix this, we need to force the link down and trigger another resolve on PHY interface change event. Fixes: 9525ae83959b ("phylink: add phylink infrastructure") Signed-off-by: Russell King (Oracle) Signed-off-by: Marek Behún Signed-off-by: Jakub Kicinski --- drivers/net/phy/phylink.c | 13 ++++++++++++- 1 file changed, 12 insertions(+), 1 deletion(-) (limited to 'drivers/net/phy') diff --git a/drivers/net/phy/phylink.c b/drivers/net/phy/phylink.c index 3ad7397b8119..708d2a984621 100644 --- a/drivers/net/phy/phylink.c +++ b/drivers/net/phy/phylink.c @@ -710,6 +710,7 @@ static void phylink_resolve(struct work_struct *w) struct phylink_link_state link_state; struct net_device *ndev = pl->netdev; bool mac_config = false; + bool retrigger = false; bool cur_link_state; mutex_lock(&pl->state_mutex); @@ -723,6 +724,7 @@ static void phylink_resolve(struct work_struct *w) link_state.link = false; } else if (pl->mac_link_dropped) { link_state.link = false; + retrigger = true; } else { switch (pl->cur_link_an_mode) { case MLO_AN_PHY: @@ -747,6 +749,15 @@ static void phylink_resolve(struct work_struct *w) /* Only update if the PHY link is up */ if (pl->phydev && pl->phy_state.link) { + /* If the interface has changed, force a + * link down event if the link isn't already + * down, and re-resolve. + */ + if (link_state.interface != + pl->phy_state.interface) { + retrigger = true; + link_state.link = false; + } link_state.interface = pl->phy_state.interface; /* If we have a PHY, we need to update with @@ -789,7 +800,7 @@ static void phylink_resolve(struct work_struct *w) else phylink_link_up(pl, link_state); } - if (!link_state.link && pl->mac_link_dropped) { + if (!link_state.link && retrigger) { pl->mac_link_dropped = false; queue_work(system_power_efficient_wq, &pl->resolve); } -- cgit v1.2.3 From dbae3388ea9ca33bd1d5eabc3b0ef17e69c74677 Mon Sep 17 00:00:00 2001 From: "Russell King (Oracle)" Date: Tue, 23 Nov 2021 16:44:03 +0100 Subject: net: phylink: Force retrigger in case of latched link-fail indicator MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit On mv88e6xxx 1G/2.5G PCS, the SerDes register 4.2001.2 has the following description: This register bit indicates when link was lost since the last read. For the current link status, read this register back-to-back. Thus to get current link state, we need to read the register twice. But doing that in the link change interrupt handler would lead to potentially ignoring link down events, which we really want to avoid. Thus this needs to be solved in phylink's resolve, by retriggering another resolve in the event when PCS reports link down and previous link was up, and by re-reading PCS state if the previous link was down. The wrong value is read when phylink requests change from sgmii to 2500base-x mode, and link won't come up. This fixes the bug. Fixes: 9525ae83959b ("phylink: add phylink infrastructure") Signed-off-by: Russell King (Oracle) Signed-off-by: Marek Behún Signed-off-by: Jakub Kicinski --- drivers/net/phy/phylink.c | 13 +++++++++++++ 1 file changed, 13 insertions(+) (limited to 'drivers/net/phy') diff --git a/drivers/net/phy/phylink.c b/drivers/net/phy/phylink.c index 708d2a984621..5904546acae6 100644 --- a/drivers/net/phy/phylink.c +++ b/drivers/net/phy/phylink.c @@ -741,6 +741,19 @@ static void phylink_resolve(struct work_struct *w) case MLO_AN_INBAND: phylink_mac_pcs_get_state(pl, &link_state); + /* The PCS may have a latching link-fail indicator. + * If the link was up, bring the link down and + * re-trigger the resolve. Otherwise, re-read the + * PCS state to get the current status of the link. + */ + if (!link_state.link) { + if (cur_link_state) + retrigger = true; + else + phylink_mac_pcs_get_state(pl, + &link_state); + } + /* If we have a phy, the "up" state is the union of * both the PHY and the MAC */ -- cgit v1.2.3