RE: [PATCH v4 1/2] platform/chrome: cros_ec_typec: Configure Retimer cable type
From: Patel, Utkarsh H
Date: Tue Aug 22 2023 - 17:22:02 EST
Hi Prashant,
> -----Original Message-----
> From: Prashant Malani <pmalani@xxxxxxxxxxxx>
> Sent: Monday, August 21, 2023 4:31 PM
> To: Patel, Utkarsh H <utkarsh.h.patel@xxxxxxxxx>
> Cc: linux-kernel@xxxxxxxxxxxxxxx; linux-usb@xxxxxxxxxxxxxxx;
> heikki.krogerus@xxxxxxxxxxxxxxx; bleung@xxxxxxxxxxxx
> Subject: Re: [PATCH v4 1/2] platform/chrome: cros_ec_typec: Configure
> Retimer cable type
>
> On Mon, Aug 21, 2023 at 10:18 AM Patel, Utkarsh H
> <utkarsh.h.patel@xxxxxxxxx> wrote:
> > > > /*
> > > > * Spoof the VDOs that were likely communicated by the partner for TBT
> alt
> > > > * mode.
> > > > @@ -432,6 +453,9 @@ static int cros_typec_enable_tbt(struct
> > > > cros_typec_data *typec,
> > > >
> > > > /* Cable Discover Mode VDO */
> > > > data.cable_mode = TBT_MODE;
> > > > +
> > > > + data.cable_mode |= cros_typec_get_cable_vdo(port,
> > > > +USB_TYPEC_TBT_SID);
> > > > +
>
> Here is the first call to cros_typec_get_cable_vdo(port, TBT)....
>
> > > > data.cable_mode |= TBT_SET_CABLE_SPEED(pd_ctrl->cable_speed);
> > > >
> > > > if (pd_ctrl->control_flags & USB_PD_CTRL_OPTICAL_CABLE) @@ -
> > > 522,8
> > > > +546,10 @@ static int cros_typec_enable_usb4(struct
> > > > +cros_typec_data
> > > *typec,
> > > > /* Cable Type */
> > > > if (pd_ctrl->control_flags & USB_PD_CTRL_OPTICAL_CABLE)
> > > > data.eudo |= EUDO_CABLE_TYPE_OPTICAL <<
> > > EUDO_CABLE_TYPE_SHIFT;
> > > > - else if (pd_ctrl->control_flags & USB_PD_CTRL_ACTIVE_CABLE)
> > > > + else if (cros_typec_get_cable_vdo(port, USB_TYPEC_TBT_SID) &
> > > > +TBT_CABLE_RETIMER)
>
> And here is the 2nd.
>
> > > > data.eudo |= EUDO_CABLE_TYPE_RE_TIMER <<
> > > EUDO_CABLE_TYPE_SHIFT;
> > > We shouldn't need to call cros_typec_get_cable_vdo more than once.
> > > Either call it once earlier when you are crafting the
> > > data.cable_mode member and then re-use that variable here. Or don't call
> it there and just call it here.
> >
> > We are only calling it once depending upon which mode we enter TBT Alt or
> USB4.
>
> There should only be 1 "call site" and that should be sufficient to grab the VDO
> from the framework for all circumstances. Whether the other invocation
> doesn't get called under certain circumstances isn't as relevant.
Are you suggesting something like this?
static int cros_typec_configure_mux(struct cros_typec_data *typec, int port_num,)...
if (port->mux_flags & USB_PD_MUX_USB4_ENABLED ||
port->mux_flags & USB_PD_MUX_TBT_COMPAT_ENABLED)
cable_tbt_vdo = cros_typec_get_cable_vdo(port, USB_TYPEC_TBT_SID);
Sincerely,
Utkarsh Patel.