> > Martin, you design has no problem for most of cases, but some hardware
> > needs special sequence for phy control. I will give an example below.
> great to hear that this should work for most devices!
> >> it would be great if you could explain the issue behind this (and
> >> thereby answer the
> >> question: why we would not want the HCD core to manage the PHY states)!
> > Eg, taking Qualcomm USB2 controller as an example, it even does not
> > allow chipidea core to manage its power operation, see
> > CI_HDRC_OVERRIDE_PHY_CONTROL at chipidea driver (usb/chipidea/core.c).
> Its phy_power_on is called after ehci controller reset has finished.
> > (usb/chipidea/ci_hdrc_msm.c).
> I see, thank you for explaining this!
> what do you think about replacing the two following fields from struct
> struct usb_phy *usb_phy;
> struct phy *phy;
> * do not manage the PHY state in the HCD core, instead let the driver
> * this (for example if the PHY can only be turned on after a specific
> bool skip_phy_initialization;
> maybe I should also do this together with my other series which adds the PHY
> wrapper to the HCD core (or even as a separate series, which would be merged
> before this and the PHY wrapper series). what do you think?
I think it is better to do this in one series.