Re: [PATCH RFC 12/18] net: dsa: mt7530: Call phylib for set_eee and get_eee

[Andrew Lunn <andrew@lunn.ch>]

 On Fri, Feb 17, 2023 at 11:48:01AM +0000, Russell King (Oracle) wrote:
> On Fri, Feb 17, 2023 at 04:42:24AM +0100, Andrew Lunn wrote:
> > phylib should be called in order to manage the EEE settings in the
> > PHY, and to return EEE status such are supported link modes, and what
> > the link peer supports.
> > 
> > Signed-off-by: Andrew Lunn <andrew@lunn.ch>
> > ---
> >  drivers/net/dsa/mt7530.c | 6 ++++++
> >  1 file changed, 6 insertions(+)
> > 
> > diff --git a/drivers/net/dsa/mt7530.c b/drivers/net/dsa/mt7530.c
> > index 214450378978..a472353f14f8 100644
> > --- a/drivers/net/dsa/mt7530.c
> > +++ b/drivers/net/dsa/mt7530.c
> > @@ -3124,10 +3124,13 @@ static int mt753x_get_mac_eee(struct dsa_switch *ds, int port,
> >  {
> >  	struct mt7530_priv *priv = ds->priv;
> >  	u32 eeecr = mt7530_read(priv, MT7530_PMEEECR_P(port));
> > +	struct dsa_port *dp = dsa_to_port(ds, port);
> >  
> >  	e->tx_lpi_enabled = !(eeecr & LPI_MODE_EN);
> >  	e->tx_lpi_timer = GET_LPI_THRESH(eeecr);
> >  
> > +	if (dp->slave->phydev)
> > +		return phy_ethtool_get_eee(dp->slave->phydev, e);
> 
> Given that DSA makes use of phylink, is there a reason why we can't use
> the phylink wrappers here (which may allow for future EEE improvements,
> e.g. moving the gating of EEE with the TX LPI enable) ?

Turns out this is pointless. dsa_slave_get_eee() and
dsa_slave_set_eee() already make calls to phylink_ethtool_get_eee()
and phylink_ethtool_set_eee() after calling into the DSA driver.

So i will drop this patch, and the b53 change.

   Andrew