[PATCH net] net: phy: Fix the issue that netif always links up after resuming

Even though the link is down before entering hibernation,
there is an issue that the network interface always links up after resuming
from hibernation.

If the link is still down before enabling the network interface,
and after resuming from hibernation, the phydev->state is forcibly set
to PHY_UP in mdio_bus_phy_restore(), and the link becomes up.

In suspend sequence, only if the PHY is attached, mdio_bus_phy_suspend()
calls phy_stop_machine(), and mdio_bus_phy_resume() calls
phy_start_machine().
In resume sequence, it's enough to do the same as mdio_bus_phy_resume()
because the state has been preserved.

This patch fixes the issue by calling phy_start_machine() in
mdio_bus_phy_restore() in the same way as mdio_bus_phy_resume().

Suggested-by: Heiner Kallweit <hkallweit1@gmail.com>
Signed-off-by: Kunihiko Hayashi <hayashi.kunihiko@socionext.com>
---
 drivers/net/phy/phy_device.c | 7 ++-----
 1 file changed, 2 insertions(+), 5 deletions(-)

This patch is based on the RFC patch discussion [1].
[1] https://www.spinics.net/lists/netdev/msg537326.html

diff --git a/drivers/net/phy/phy_device.c b/drivers/net/phy/phy_device.c
index 7d5d698..3685be4 100644
--- a/drivers/net/phy/phy_device.c
+++ b/drivers/net/phy/phy_device.c

@@ -315,11 +315,8 @@ static int mdio_bus_phy_restore(struct device *dev)
 	if (ret < 0)
 		return ret;
 
-	/* The PHY needs to renegotiate. */
-	phydev->link = 0;
-	phydev->state = PHY_UP;
-
-	phy_start_machine(phydev);
+	if (phydev->attached_dev && phydev->adjust_link)
+		phy_start_machine(phydev);
 
 	return 0;
 }

-- 
2.7.4

On 18.12.2018 08:57, Kunihiko Hayashi wrote:

Even though the link is down before entering hibernation,
there is an issue that the network interface always links up after resuming
from hibernation.

If the link is still down before enabling the network interface,
and after resuming from hibernation, the phydev->state is forcibly set
to PHY_UP in mdio_bus_phy_restore(), and the link becomes up.

In suspend sequence, only if the PHY is attached, mdio_bus_phy_suspend()
calls phy_stop_machine(), and mdio_bus_phy_resume() calls
phy_start_machine().
In resume sequence, it's enough to do the same as mdio_bus_phy_resume()
because the state has been preserved.

This patch fixes the issue by calling phy_start_machine() in
mdio_bus_phy_restore() in the same way as mdio_bus_phy_resume().

The patch itself is fine and also that you tagged it as "net".
What's missing is a "Fixes:" line (before the Suggested-by:).
I think it should be:

Fixes: bc87922ff59d ("phy: Move PHY PM operations into phy_device")

This commit didn't actually introduce the issue, but moved it to today's
place.

Hi Heiner,

On Tue, 18 Dec 2018 19:15:39 +0100 [off-list ref] wrote:

On 18.12.2018 08:57, Kunihiko Hayashi wrote:

quoted

Even though the link is down before entering hibernation,
there is an issue that the network interface always links up after resuming
from hibernation.

If the link is still down before enabling the network interface,
and after resuming from hibernation, the phydev->state is forcibly set
to PHY_UP in mdio_bus_phy_restore(), and the link becomes up.

In suspend sequence, only if the PHY is attached, mdio_bus_phy_suspend()
calls phy_stop_machine(), and mdio_bus_phy_resume() calls
phy_start_machine().
In resume sequence, it's enough to do the same as mdio_bus_phy_resume()
because the state has been preserved.

This patch fixes the issue by calling phy_start_machine() in
mdio_bus_phy_restore() in the same way as mdio_bus_phy_resume().

The patch itself is fine and also that you tagged it as "net".
What's missing is a "Fixes:" line (before the Suggested-by:).
I think it should be:

Fixes: bc87922ff59d ("phy: Move PHY PM operations into phy_device")

This commit didn't actually introduce the issue, but moved it to today's
place.

Thanks for pointing out.
I was concerned about that and which commit would be fixed if added.

Thank you,

quoted

Suggested-by: Heiner Kallweit <hkallweit1@gmail.com>
Signed-off-by: Kunihiko Hayashi <hayashi.kunihiko@socionext.com>
---
 drivers/net/phy/phy_device.c | 7 ++-----
 1 file changed, 2 insertions(+), 5 deletions(-)

This patch is based on the RFC patch discussion [1].
[1] https://www.spinics.net/lists/netdev/msg537326.html

diff --git a/drivers/net/phy/phy_device.c b/drivers/net/phy/phy_device.c
index 7d5d698..3685be4 100644
--- a/drivers/net/phy/phy_device.c
+++ b/drivers/net/phy/phy_device.c

@@ -315,11 +315,8 @@ static int mdio_bus_phy_restore(struct device *dev)
 	if (ret < 0)
 		return ret;
 
-	/* The PHY needs to renegotiate. */
-	phydev->link = 0;
-	phydev->state = PHY_UP;
-
-	phy_start_machine(phydev);
+	if (phydev->attached_dev && phydev->adjust_link)
+		phy_start_machine(phydev);
 
 	return 0;
 }

---
Best Regards,
Kunihiko Hayashi

From: Kunihiko Hayashi <hayashi.kunihiko@socionext.com>
Date: Tue, 18 Dec 2018 16:57:04 +0900

Even though the link is down before entering hibernation,
there is an issue that the network interface always links up after resuming
from hibernation.

If the link is still down before enabling the network interface,
and after resuming from hibernation, the phydev->state is forcibly set
to PHY_UP in mdio_bus_phy_restore(), and the link becomes up.

In suspend sequence, only if the PHY is attached, mdio_bus_phy_suspend()
calls phy_stop_machine(), and mdio_bus_phy_resume() calls
phy_start_machine().
In resume sequence, it's enough to do the same as mdio_bus_phy_resume()
because the state has been preserved.

This patch fixes the issue by calling phy_start_machine() in
mdio_bus_phy_restore() in the same way as mdio_bus_phy_resume().

Suggested-by: Heiner Kallweit <hkallweit1@gmail.com>
Signed-off-by: Kunihiko Hayashi <hayashi.kunihiko@socionext.com>

Applied and queued up for -stable.