Hi Grzegorz,
Grzegorz Jaszczyk <jaz@xxxxxxxxxxxx> wrote on Wed, 3 Apr 2019 11:02:15
+0200:
> Hi Miquel,
>
> pon., 1 kwi 2019 o 18:52 Miquel Raynal <miquel.raynal@xxxxxxxxxxx> napisał(a):
> > +static int mvebu_comphy_power_on(struct phy *phy)
> > +{
> > + struct mvebu_comphy_lane *lane = phy_get_drvdata(phy);
> > + struct mvebu_comphy_priv *priv = lane->priv;
> > + u32 fw_param = 0;
> > + int fw_mode;
> > + int ret;
> > +
> > + fw_mode = mvebu_comphy_get_fw_mode(lane->id, lane->port,
> > + lane->mode, lane->submode);
> > + if (fw_mode < 0)
> > + goto try_legacy;
> > +
> > + /* Try SMC flow first */
> > + switch (lane->mode) {
> > + case PHY_MODE_ETHERNET:
> > + dev_dbg(priv->dev, "set lane %d to Ethernet mode\n", lane->id);
> > + switch (lane->submode) {
> > + case PHY_INTERFACE_MODE_SGMII:
> > + case PHY_INTERFACE_MODE_2500BASEX:
> > + case PHY_INTERFACE_MODE_10GKR:
> > + fw_param = COMPHY_FW_PARAM(fw_mode, lane->port);
>
> It seems it will not work for SGMII and 2500BASEX. It is because you
> are setting max speed for all comphy modes inside COMPHY_FW_PARAM
> macro which is wrong. Did you test it?
I misread ATF implementation, I was sure the speed was derived from the
type and this parameter was useless. I'll do the modification.
> BTW can you point to some public temporary git branch where this
> patchset can be found?
Just pushed there if you want to take a look:
https://github.com/miquelraynal/linux.git marvell/5.1-rc1/cp110-comphy
Thanks for the review!
Miquèl
