Re: [PATCH 1/2] typec: tcpm: Do not disconnect link for self powered devices

[Badhri Jagan Sridharan <badhri@google.com>]

On Thu, Sep 13, 2018 at 10:51 AM Guenter Roeck <linux@roeck-us.net> wrote:
>
> On Thu, Sep 13, 2018 at 04:24:08PM +0300, Heikki Krogerus wrote:
> > +Guenter
> >
> > On Wed, Sep 12, 2018 at 07:11:12PM -0700, Badhri Jagan Sridharan wrote:
> > > During HARD_RESET the data link is disconnected.
> > > For self powered device, the spec is advising against doing that.
> > >
> > > >From USB_PD_R3_0
> > > 7.1.5 Response to Hard Resets
> > > Device operation during and after a Hard Reset is defined as follows:
> > > Self-powered devices Should Not disconnect from USB during a Hard Reset
> > > (see Section 9.1.2).
> > > Bus powered devices will disconnect from USB during a Hard Reset due to the
> > > loss of their power source.
> > >
> > > Tackle this by letting TCPM know whether the device is self or bus powered.
> > >
> > > This overcomes unnecessary port disconnections from hard reset.
> > > Also, speeds up the enumeration time when connected to Type-A ports.
> > >
> > > Signed-off-by: Badhri Jagan Sridharan <Badhri@google.com>
> > > ---
> > >  drivers/usb/typec/tcpm.c | 6 +++++-
> > >  include/linux/usb/tcpm.h | 1 +
> > >  2 files changed, 6 insertions(+), 1 deletion(-)
> > >
> > > diff --git a/drivers/usb/typec/tcpm.c b/drivers/usb/typec/tcpm.c
> > > index 4f1f4215f3d6..a4e0c027a2a9 100644
> > > --- a/drivers/usb/typec/tcpm.c
> > > +++ b/drivers/usb/typec/tcpm.c
> > > @@ -3270,7 +3270,11 @@ static void run_state_machine(struct tcpm_port *port)
> > >             memset(&port->pps_data, 0, sizeof(port->pps_data));
> > >             tcpm_set_vconn(port, false);
> > >             tcpm_set_charge(port, false);
> > > -           tcpm_set_roles(port, false, TYPEC_SINK, TYPEC_DEVICE);
> > > +
> > > +           if (port->tcpc->config->self_powered)
> >
> > Add a member for that flag to the struct tcpm_port, and check that
> > here instead. I'll explain why below.
> >
> > > +                   tcpm_set_roles(port, true, TYPEC_SINK, TYPEC_DEVICE);
> > > +           else
> > > +                   tcpm_set_roles(port, false, TYPEC_SINK, TYPEC_DEVICE);
> > >
>
> Maybe I am missing something, but
>                 tcpm_set_roles(port, port->self_powered,
>                                TYPEC_SINK, TYPEC_DEVICE);
> should accomplish the same (assuming the flag is moved to tcpm_port).

Yeah Makes sense to me..
>
> Does the code at SRC_HARD_RESET_VBUS_OFF: also have to be changed,
> or only SNK_HARD_RESET_SINK_OFF: ? In other words, does this only
> apply if the port is a sink, or also if is is source ?

Good question ! I dont see any direct guidance on the spec.
All that I can see is the following:

From 2.6.1 Source Operation:
Restores the Port’s data role to DFP.

From 7.1.5 Response to Hard Resets:
"The Source Shall meet both tSafe5V and tSafe0V relative to the start
of the voltage transition as shown in Figure 7-8."

I am leaning towards to changing the link state from false to true for
SRC_HARD_RESET_VBUS_OFF
as well. This is to keep it symmetrical. What do you think ?

>
> Guenter
>
> > >             /*
> > >              * VBUS may or may not toggle, depending on the adapter.
> > >              * If it doesn't toggle, transition to SNK_HARD_RESET_SINK_ON
> > > diff --git a/include/linux/usb/tcpm.h b/include/linux/usb/tcpm.h
> > > index 7e7fbfb84e8e..50c74a77db55 100644
> > > --- a/include/linux/usb/tcpm.h
> > > +++ b/include/linux/usb/tcpm.h
> > > @@ -89,6 +89,7 @@ struct tcpc_config {
> > >     enum typec_port_data data;
> > >     enum typec_role default_role;
> > >     bool try_role_hw;       /* try.{src,snk} implemented in hardware */
> > > +   bool self_powered;      /* port belongs to a self powered device */
> >
> > I'm not sure we should add any more members to that structure, but
> > maybe it's not a problem for now. We can't quite yet get rid of that
> > structure.
> >
> > You do need to introduce a new device property already. Then read the

This would be part of usb-connector binding. right ?

> > value for that new member you added to struct tcpm_port in
> > tcpm_fw_get_caps() and in tcpm_copy_caps().
> >
> >
> > Thanks,
> >
> > --
> > heikki