Re: [PATCH net-next v1 3/4] net: phy: Don't report advertised EEE modes if EEE is disabled

[Russell King (Oracle)]

On Thu, Apr 24, 2025 at 04:30:40PM +0200, Andrew Lunn wrote:
> On Thu, Apr 24, 2025 at 03:02:21PM +0200, Oleksij Rempel wrote:
> > Currently, `ethtool --show-eee` reports "Advertised EEE link modes" even when
> > EEE is disabled, which can be misleading. For example:
> > 
> >   EEE settings for lan1:
> >           EEE status: disabled
> >           Tx LPI: disabled
> >           Supported EEE link modes:  100baseT/Full
> >                                      1000baseT/Full
> >           Advertised EEE link modes:  100baseT/Full
> >                                       1000baseT/Full
> >           Link partner advertised EEE link modes:  Not reported
> 
> What is the behaviour for normal link mode advertisement? If i turn
> autoneg off, do the advertised link modes disappear? Do they reappear
> when i turn autoneg back on again?
> 
> I would expect EEE to follow what the normal link modes do. Assuming
> the Read/modify/write does not break this.

It's difficult to compare, because ethtool is implemented differently
between modifying the link modes and the EEE stuff. ethtool -s autoneg
on uses this:

                        if (autoneg_wanted == AUTONEG_ENABLE &&
                            advertising_wanted == NULL &&
                            full_advertising_wanted == NULL) {
                                unsigned int i;

                                /* Auto negotiation enabled, but with
                                 * unspecified speed and duplex: enable all
                                 * supported speeds and duplexes.
                                 */

whereas do_seee() has no special handling.

So, if we want ethtool --set-eee eee off; ethtool --set-eee eee on *not*
to end up with the advertising mask being cleared, then we have to
preserve it, or force the advertising mask to something if the
advertising mask is empty and eee_enabled is true.

Or we preserve the advertising mask when eee_enabled is cleared, which
is what we do today.

I think, given the different implementations in ethtool, we can't just
say "we want it to be have the same" by just modifying the kernel.

-- 
RMK's Patch system: https://www.armlinux.org.uk/developer/patches/
FTTP is here! 80Mbps down 10Mbps up. Decent connectivity at last!