Re: [PATCH v3 4/4] usb: typec: tcpci: add support to set connector orientation

[Heikki Krogerus <heikki.krogerus@linux.intel.com>]

Hi,

I'm sorry to keep you waiting.

On Mon, Feb 26, 2024 at 01:27:01PM +0100, Marco Felsch wrote:
> Hi,
> 
> On 24-02-26, Heikki Krogerus wrote:
> > On Thu, Feb 22, 2024 at 10:09:03PM +0100, Marco Felsch wrote:
> > > This add the support to set the optional connector orientation bit which
> > > is part of the optional CONFIG_STANDARD_OUTPUT register 0x18 [1]. This
> > > allows system designers to connect the tcpc orientation pin directly to
> > > the 2:1 ss-mux.
> > > 
> > > [1] https://www.usb.org/sites/default/files/documents/usb-port_controller_specification_rev2.0_v1.0_0.pdf
> > > 
> > > Signed-off-by: Marco Felsch <m.felsch@pengutronix.de>
> > > ---
> > > v3:
> > > - no changes
> > > v2:
> > > - Make use of fallthrough 
> > > 
> > >  drivers/usb/typec/tcpm/tcpci.c | 44 ++++++++++++++++++++++++++++++++++
> > >  include/linux/usb/tcpci.h      |  8 +++++++
> > >  2 files changed, 52 insertions(+)
> > > 
> > > diff --git a/drivers/usb/typec/tcpm/tcpci.c b/drivers/usb/typec/tcpm/tcpci.c
> > > index 7118551827f6..73a52e7f95c2 100644
> > > --- a/drivers/usb/typec/tcpm/tcpci.c
> > > +++ b/drivers/usb/typec/tcpm/tcpci.c
> > > @@ -67,6 +67,18 @@ static int tcpci_write16(struct tcpci *tcpci, unsigned int reg, u16 val)
> > >  	return regmap_raw_write(tcpci->regmap, reg, &val, sizeof(u16));
> > >  }
> > >  
> > > +static bool tcpci_check_std_output_cap(struct regmap *regmap, u8 mask)
> > > +{
> > > +	unsigned int reg;
> > > +	int ret;
> > > +
> > > +	ret = regmap_read(regmap, TCPC_STD_OUTPUT_CAP, &reg);
> > > +	if (ret < 0)
> > > +		return ret;
> > > +
> > > +	return (reg & mask) == mask;
> > > +}
> > > +
> > >  static int tcpci_set_cc(struct tcpc_dev *tcpc, enum typec_cc_status cc)
> > >  {
> > >  	struct tcpci *tcpci = tcpc_to_tcpci(tcpc);
> > > @@ -301,6 +313,28 @@ static int tcpci_set_polarity(struct tcpc_dev *tcpc,
> > >  			   TCPC_TCPC_CTRL_ORIENTATION : 0);
> > >  }
> > >  
> > > +static int tcpci_set_orientation(struct tcpc_dev *tcpc,
> > > +				 enum typec_orientation orientation)
> > > +{
> > > +	struct tcpci *tcpci = tcpc_to_tcpci(tcpc);
> > > +	unsigned int reg;
> > > +
> > > +	switch (orientation) {
> > > +	case TYPEC_ORIENTATION_NONE:
> > > +		/* We can't put a single output into high impedance */
> > > +		fallthrough;
> > > +	case TYPEC_ORIENTATION_NORMAL:
> > > +		reg = TCPC_CONFIG_STD_OUTPUT_ORIENTATION_NORMAL;
> > > +		break;
> > > +	case TYPEC_ORIENTATION_REVERSE:
> > > +		reg = TCPC_CONFIG_STD_OUTPUT_ORIENTATION_FLIPPED;
> > > +		break;
> > > +	}
> > > +
> > > +	return regmap_update_bits(tcpci->regmap, TCPC_CONFIG_STD_OUTPUT,
> > > +				  TCPC_CONFIG_STD_OUTPUT_ORIENTATION_MASK, reg);
> > > +}
> > > +
> > >  static void tcpci_set_partner_usb_comm_capable(struct tcpc_dev *tcpc, bool capable)
> > >  {
> > >  	struct tcpci *tcpci = tcpc_to_tcpci(tcpc);
> > > @@ -808,6 +842,9 @@ struct tcpci *tcpci_register_port(struct device *dev, struct tcpci_data *data)
> > >  	if (tcpci->data->vbus_vsafe0v)
> > >  		tcpci->tcpc.is_vbus_vsafe0v = tcpci_is_vbus_vsafe0v;
> > >  
> > > +	if (tcpci->data->set_orientation)
> > > +		tcpci->tcpc.set_orientation = tcpci_set_orientation;
> > 
> > I don't think that flag is needed - not yet at least. Please just call
> > tcpci_check_std_output_cap() directly from here.
> 
> The reason for having it this way was to not break exsisting user like:
> tcpci_rt1711h, tcpci_mt6370, tcpci_maxim which may or may not implement
> the TCPC_STD_OUTPUT_CAP_ORIENTATION. This way the users of
> tcpci_register_port() can decide by on its own if they do have this
> feature or not and how this is checked. I'm fine with your proposal if
> you still think that we can check this unconditional.

Ah, I failed to consider those other glue drivers. This is fine as it
is. I'm really sorry about the delay.

Br,

-- 
heikki