net: phy: don't touch suspended flag if there's no suspend/resume callback
Subsystems (3)
Heiner Kallweit
maintainerPATCH 1/1 net: phy: don't touch suspended flag if there's no suspend/resume callback
So far we set phydev->suspended to true in phy_suspend() even if the PHY driver doesn't implement the suspend callback. This applies accordingly for the resume path. The current behavior doesn't cause any issue I'd be aware of, but it's not logical and misleading, especially considering the description of the flag:
"suspended: Set to true if this phy has been suspended successfully"
--- drivers/net/phy/phy_device.c | 32 +++++++++++++++----------------- 1 file changed, 15 insertions(+), 17 deletions(-)
drivers/net/phy/phy_device.c +15 -17
@@ -1519,23 +1519,22 @@ EXPORT_SYMBOL(phy_detach);int phy_suspend (struct phy_device *phydev) { -struct phy_driver *phydrv =to_phy_driver (phydev->mdio.dev.driver); -struct net_device *netdev = phydev->attached_dev;struct ethtool_wolinfo wol = { .cmd =ETHTOOL_GWOL }; -int ret =0 ; +struct net_device *netdev = phydev->attached_dev; +struct phy_driver *phydrv = phydev->drv; +int ret;/* If the device has WOL enabled, we cannot suspend the PHY */ phy_ethtool_get_wol (phydev, &wol);if (wol.wolopts || (netdev && netdev->wol_enabled))return -EBUSY ; -if (phydev->drv && phydrv->suspend) - ret = phydrv->suspend (phydev); - -if (ret) -return ret; +if (!phydrv || !phydrv->suspend) +return 0 ; - phydev->suspended = true; + ret = phydrv->suspend (phydev); +if (!ret) + phydev->suspended = true;return ret; }
@@ -1543,18 +1542,17 @@ EXPORT_SYMBOL(phy_suspend);int __phy_resume (struct phy_device *phydev) { -struct phy_driver *phydrv =to_phy_driver (phydev->mdio.dev.driver); -int ret =0 ; +struct phy_driver *phydrv = phydev->drv; +int ret;WARN_ON (!mutex_is_locked (&phydev->lock)); -if (phydev->drv && phydrv->resume) - ret = phydrv->resume (phydev); - -if (ret) -return ret; +if (!phydrv || !phydrv->resume) +return 0 ; - phydev->suspended = false; + ret = phydrv->resume (phydev); +if (!ret) + phydev->suspended = false;return ret; }
Signature · 1 line
2.26.0
Quoted 8 lines · Heiner Kallweit
So far we set
phydev->suspendedto true inphy_suspend()even if the PHY driver doesn't implement the suspend callback. This applies accordingly for the resume path. The current behavior doesn't cause any issue I'd be aware of, but it's not logical and misleading, especially considering the description of the flag:
"suspended: Set to true if this phy has been suspended successfully"Signed-off-by Heiner Kallweit <[email protected]>
Andrew
Quoted 8 lines · Heiner Kallweit
So far we set
phydev->suspendedto true inphy_suspend()even if the PHY driver doesn't implement the suspend callback. This applies accordingly for the resume path. The current behavior doesn't cause any issue I'd be aware of, but it's not logical and misleading, especially considering the description of the flag:
"suspended: Set to true if this phy has been suspended successfully"Signed-off-by Heiner Kallweit <[email protected]>
Signature · 1 line
Florian
From: Heiner Kallweit <[email protected]>
Date: Thu, 26 Mar 2020 18:58:24 +0100
Quoted 8 lines
So far we set
phydev->suspendedto true inphy_suspend()even if the PHY driver doesn't implement the suspend callback. This applies accordingly for the resume path. The current behavior doesn't cause any issue I'd be aware of, but it's not logical and misleading, especially considering the description of the flag:
"suspended: Set to true if this phy has been suspended successfully"Signed-off-by Heiner Kallweit <[email protected]>
Applied, thanks.