Commit 52f95bbfcf72 ("stmmac: fix adjust link call in case of a switch
is attached") added some logic to avoid polling the fixed PHY and
therefore invoking the adjust_link callback more than once, since this
is a fixed PHY and link events won't be generated.
This works fine the first time, because we start with phydev->irq =
PHY_POLL, so we call adjust_link, then we set phydev->irq =
PHY_IGNORE_INTERRUPT and we stop polling the PHY.
Now, if we called ndo_close(), which calls both phy_stop() and does an
explicit netif_carrier_off(), we end up with a link down. Upon calling
ndo_open() again, despite starting the PHY state machine, we have
PHY_IGNORE_INTERRUPT set, and we generate no link event at all, so the
link is permanently down.
52f95bbfcf72 ("stmmac: fix adjust link call in case of a switch is attached")
Signed-off-by: Florian Fainelli <[email protected]>
---
Alexandre, Peppe,
The original patch is already a hack, but since this is a bugfix, I took the
same approach that you did here to backport this to -stable kernels.
drivers/net/ethernet/stmicro/stmmac/stmmac_main.c | 7 +++++++
1 file changed, 7 insertions(+)
diff --git a/drivers/net/ethernet/stmicro/stmmac/stmmac_main.c b/drivers/net/ethernet/stmicro/stmmac/stmmac_main.c
index 10909c9c0033..03dbf8e89c4c 100644
--- a/drivers/net/ethernet/stmicro/stmmac/stmmac_main.c
+++ b/drivers/net/ethernet/stmicro/stmmac/stmmac_main.c
@@ -882,6 +882,13 @@ static int stmmac_init_phy(struct net_device *dev)
return -ENODEV;
}
+ /* stmmac_adjust_link will change this to PHY_IGNORE_INTERRUPT to avoid
+ * subsequent PHY polling, make sure we force a link transition if
+ * we have a UP/DOWN/UP transition
+ */
+ if (phydev->is_pseudo_fixed_link)
+ phydev->irq = PHY_POLL;
+
pr_debug("stmmac_init_phy: %s: attached to PHY (UID 0x%x)"
" Link = %d\n", dev->name, phydev->phy_id, phydev->link);
--
2.9.3
On 11/14/2016 2:50 AM, Florian Fainelli wrote:
> Commit 52f95bbfcf72 ("stmmac: fix adjust link call in case of a switch
> is attached") added some logic to avoid polling the fixed PHY and
> therefore invoking the adjust_link callback more than once, since this
> is a fixed PHY and link events won't be generated.
>
> This works fine the first time, because we start with phydev->irq =
> PHY_POLL, so we call adjust_link, then we set phydev->irq =
> PHY_IGNORE_INTERRUPT and we stop polling the PHY.
>
> Now, if we called ndo_close(), which calls both phy_stop() and does an
> explicit netif_carrier_off(), we end up with a link down. Upon calling
> ndo_open() again, despite starting the PHY state machine, we have
> PHY_IGNORE_INTERRUPT set, and we generate no link event at all, so the
> link is permanently down.
>
> 52f95bbfcf72 ("stmmac: fix adjust link call in case of a switch is attached")
> Signed-off-by: Florian Fainelli <[email protected]>
> ---
> Alexandre, Peppe,
>
> The original patch is already a hack, but since this is a bugfix, I took the
> same approach that you did here to backport this to -stable kernels.
Acked-by: Giuseppe Cavallaro <[email protected]>
>
> drivers/net/ethernet/stmicro/stmmac/stmmac_main.c | 7 +++++++
> 1 file changed, 7 insertions(+)
>
> diff --git a/drivers/net/ethernet/stmicro/stmmac/stmmac_main.c b/drivers/net/ethernet/stmicro/stmmac/stmmac_main.c
> index 10909c9c0033..03dbf8e89c4c 100644
> --- a/drivers/net/ethernet/stmicro/stmmac/stmmac_main.c
> +++ b/drivers/net/ethernet/stmicro/stmmac/stmmac_main.c
> @@ -882,6 +882,13 @@ static int stmmac_init_phy(struct net_device *dev)
> return -ENODEV;
> }
>
> + /* stmmac_adjust_link will change this to PHY_IGNORE_INTERRUPT to avoid
> + * subsequent PHY polling, make sure we force a link transition if
> + * we have a UP/DOWN/UP transition
> + */
> + if (phydev->is_pseudo_fixed_link)
> + phydev->irq = PHY_POLL;
> +
> pr_debug("stmmac_init_phy: %s: attached to PHY (UID 0x%x)"
> " Link = %d\n", dev->name, phydev->phy_id, phydev->link);
>
>
From: Florian Fainelli <[email protected]>
Date: Sun, 13 Nov 2016 17:50:35 -0800
> Commit 52f95bbfcf72 ("stmmac: fix adjust link call in case of a switch
> is attached") added some logic to avoid polling the fixed PHY and
> therefore invoking the adjust_link callback more than once, since this
> is a fixed PHY and link events won't be generated.
>
> This works fine the first time, because we start with phydev->irq =
> PHY_POLL, so we call adjust_link, then we set phydev->irq =
> PHY_IGNORE_INTERRUPT and we stop polling the PHY.
>
> Now, if we called ndo_close(), which calls both phy_stop() and does an
> explicit netif_carrier_off(), we end up with a link down. Upon calling
> ndo_open() again, despite starting the PHY state machine, we have
> PHY_IGNORE_INTERRUPT set, and we generate no link event at all, so the
> link is permanently down.
>
> 52f95bbfcf72 ("stmmac: fix adjust link call in case of a switch is attached")
I added the missing "Fixes: " here.
> Signed-off-by: Florian Fainelli <[email protected]>
Applied and queued up for -stable, thanks Florian.