Re: [PATCH] phy: phy_ethtool_ksettings_set: Don't discard phy_start_aneg's return

Am 05.11.21 um 19:32 schrieb Andrew Lunn:

What PHY driver are you using this with? phy_start_aneg() generally
does not return errors, except for -EIO/-TIMEDOUT because
communication with the PHY has failed. All parameter validation should
of already happened before the call to phy_start_aneg(). So i'm
wondering if the PHY driver is doing something wrong.

I am modifying broadcom phy and will make sure that I will end up nut having
error checks in the wrong place.