On Fri, Sep 06, 2024 at 03:39:51PM +0200, Niklas Söderlund wrote: > The initial marvell-88q2xxx driver only supported the Marvell 88Q2110 > PHY without auto negotiation support. The reason documented states that > the provided initialization sequence did not to work. Now a method to > enable auto negotiation have been found by comparing the initialization > of other supported devices and an out-of-tree PHY driver. > > Perform the minimal needed initialization of the PHY to get auto > negotiation working and remove the limitation that disables the auto > negotiation feature for the mv88q2110 device. > > With this change a 1000Mbps full duplex link is able to be negotiated > between two mv88q2110 and the link works perfectly. The other side also > reflects the manually configure settings of the master device. > > # ethtool eth0 > Settings for eth0: > Supported ports: [ ] > Supported link modes: 100baseT1/Full > 1000baseT1/Full > Supported pause frame use: Symmetric Receive-only > Supports auto-negotiation: Yes My understanding is that most automotive applications using T1 don't actually want auto-neg, because it is slow. Given the static use case, everything can be statically configured. Is there a danger this change is going to cause regressions? There are users who are happy for it to use 100BaseT1 without negotiation, and the link partner is not offering any sort of negotiation. But with this change, autoneg is now the default, and the link fails to come up? To me, this actually seems like a generic problem for automotive. We want to indicate the device does support autoneg, but we don't want it on by default? I don't know if we can express that at the moment? Andrew
